The Music Room
Many months have passed since I actually worked on and updated my website. Microsoft Expression is my thing. This spring, I am going to build these overly delayed pieces of Audio Nirvana. By the end of the year all of them should be de-bugged and ready for sale. The Transcendent Cathode Follower will be reasonably priced at $1380.00 USD when ready. This has to be Single Ended Bliss through some of the worlds finest output transformers, One-Electron. The Power Supply is totally upgdaded with very top end Electrolytic Capacitors, Industrial Quality. Superb choice of somponents and parts will allow this special design amplifier to over achieve against stock units. A true Tube-Lover Amplifier, Class A, Stereo, OneElectron Output Transformers, OEM Toroidal Power Transformer, Low Noise, Crisp, Clean, Musical, Smooth, Vibrant, Emotional. All that a 2A3 SE amp delivers but with a bit more slam in the low end. The Tubes will be NOS or Current Manufacture.
The CAT Clone will be priced at $1150.00 USD when tested and de-bugged. I can install premium parts and provide a quote for this extra. You can expect $500 -1000 for VCAPS, somewhat less with other quality brands. If you have a preference let us know, we can make it happen. The tubs will be standard NOS or Current Manufacture.
The final 6L6/5881 Amplifier has a special significance as the amp was a collaborative effort between my Best Friend (RIP), and his Brother. I inherited the amp as it never worked. Only one channel had sound. The schematic was of questionable repute, so be it. I stripped the amp down and plan to rebuild this to perform and make good music. I have always wanted a 6L6 Stereo Amp so thiis will be my home brew. Maybe in the future I will sell it, but for now the sentimental value makes it a long term keeper. The wood base is made of Brazilian Imbuia Wood, the very old trees that were dark wood. These were actually House Baseboards re-purposed for this amplifier. The Wood will shine with some Linseed Oil.
Transcendent Sound Cathode Follower - I purchased on of the last remaining sets of Chassis and Toroidal Transformer from Bruce Rozenblit when he cancelled this product, what a shame, but who can blame, Bruce is out there slugging away, a true innovator. But all of the parts are now at gathered in the large Zip Lock Bag, and a surprise not shown in the picture below. I will begin the assembly in late July as May has me traveling and delivering Radio Network Designs at full throttle at my day job, so free time weekends is spent taking CEU-Continuing Education Credit Courses and enjoying the company of my lovely Wife. The Transcendent Sound Cathode Follower is a "point-to-point" design, no PC boards, just how Tube Amps are best made, no PC Boards. Tube Amps and PC Boards are never really a good thing, for longevity that is. Yes they simplify assembly, but in those PC boards lurk all sorts of future issues like "cold-solder joints" and "intermittent sound" and if not properly designed, overheating from the Tubes themselves. Also when you love to roll tubes and parts, each re-heat of the PC board or Tube Pull brings you one step closer to breaking a PC board trace or creating an intermittent solder connection. In this build I have a few tricks up my sleeve, and one pleasant Power Supply surprise. The Output Transformers are the specified OneElectron UBT-1. I am open to offers when I complete the amp. But give me a chance to enjoy it before it finds a new home. The front panel is a nice Black EIA 19" wrinkle blank panel for which I also have some gorgeous black Rack Handles. The bottom panel is Hammond Aluminum and the feet will be Soft Rubber, with the tentative plan to have three rubber feet in the rear and two rubber feet in the front due to the weight distribution. Single Ended amps have "magic", and this one promises to deliver it with a better bottom end. Stay tuned as I build this for pure SE bliss. Man I would love to have the Chassis Powder Coated with a deep dark almost Blue-Black. A bargain I must say in today's world full of regular quality, run of the mill Internet Amps. Not that a Prima Luna is not the best bargain on the web. Just look at Uncle Kevin's Web Site Specials and be the lucky one to scarf a Prima Luna NOW! www.upscaleaudio.com
CAT SL-1 Kit from AnalogMetric - The actual AnalogMetric PC Board Kit was purchased in 2008-2009 when I was living in San Juan. It went dormant until I purchased the Chassis, Power Supply Kit and the Toroidal Transformer about 4 years ago. My plan includes enjoying the Preamp for a few month's, making sure that all works and it has no issues during burn-in. I then will sell it to someone who needs a high gain Preamp of Classic Design and can rely on my artisanal assembly skills. Gosh, I have lots of equipment, and my main DIY Preamp remains the Aikido 6SN7 from John Broskie, although I do stand by my AtmaSphere MP3 and M60's every so often when I need a dose of deep and cavernous sound staging with lots of air and Vinyl. This one will sell for $750.00 USD
Stereo 6L6/5881 amplifier started in the 90's - This is a project with a long story. It that began with two of my friends in San Juan, many years ago, like back in 1999. We shared the work among us. One friend painted, the other hand crafted a gorgeous solid Brazilian Imbuia base from an old baseboard molding. This is OLD Imbuia, the dark kind so rare these days. Young Imbuia trees are felled in the Amazon at alarming rates and the Imbuia wood is not allowed to reach maturity. I understand that Brazil has limited exporting Imbuia, so this makes the wooden base somewhat rare and beautiful. My part of the deal was to drill and punch the chassis, and create a wall sized plotted wiring diagram of the Schematic in a Point to Point Diagram. The interesting story here, the original attempt to build it did not work. We had chosen a "Para-Phase" Hammond 1650F design from the Principles of Power book. We could not stop this amp from motor boating, no matter what we did. Maybe there was a serious wiring error, maybe not. We even contacted the author of the Book and the creator of the schematic. He confirmed that He had never actually built the amp, so this imparts some doubt as to the design and whether it needed some empirical tweaking or corrections. "Para phase" amplifiers need very careful attention to the coupling of the multiple stages, the feedback value and also the values (poles) of the inter-stage coupling capacitors as these can cause oscillations. We tried every trick in the book and it never worked. I stripped it all down a 3 years ago and currently, I am looking for a 6L6/5881 schematic that speaks to me. What I have planned is to build a Classic Stereo 6L6/5881 amplifier with the same Blue Edcor 4.2K 50 Watt OPT's shown in the photograph below. I have to agree that the amp looks gorgeous, I am expecting the same sonically. This will likely be a Seasonal Fall build as the Leaves start to make raking a popular activity. The tubes posing for the photo below are 6550 Shuguang, 5R4GY RCA and ECC83 Tesla. It needs Power Supply Capacitors and a fresh new look at the Power Transformer that I purchased in Costa Mesa, CA when I was touring that area 15 years ago. I haven't built a Stereo Amplifier other than the Aurora Olson Amplifier, so this is bound to challenge me in several ways. Will this one be for sale? The sentimental value here may prevent me from selling it. But as we all know, everything except health, love and happiness has a price. But.... I would be open to offers starting at $1300.00 for this one-of-a-kind Stereo Beauty, hand made with a rare South American Imbuia wood base hand crafted by a real Luthier.
2016 was a slow year at ShermanAudio, but in the end, the Mullard 5-20's were resuscitated!
The year 2016 was closing out on me and so was the 20th anniversary of the Creation of my very first Hand Made DIY Tube Amps, the Mullard 5-20's.
I had spent my time fixing and upgrading existing equipment, not creating. Mind you I love to fix and troubleshoot, but my first love is to make new stuff.
My goal in 2016 was to honor my original effort and restore my personal and first made Mullard 5-20 to true operational status.
I am very happy to say they are now finally and perfectly stable, and very nice sound.
The culprit of all of my woes was the fact that original Mullard Power Supply Specifications in the Mullard manual don't transfer to current Transformers and performance characteristics.
Maybe in 1959 the original 400-0-400V, 180ma secondary was required to overcome the Copper and Iron losses in the Transformer Core at the time, who knows. But the fact remains, it took me a long time to figure out what the best off-the-shelf Power Transformer choice was.
I learned the hard was that the common 115VAC USA Primary AC Voltage standard is not necessarily what we can depend on, especially for un-regulated power supplies. There are a few DIY enemies that lurk inside of the windings of current Plate and Filament Transformers and the US power grid teams up to create voltage boosts that wreak havoc on any novice Tube amp builder and almost all pre 1970 Vintage Tube Amps.
These two enemies are today's Mains Voltages that conspire with their 115 VAC Plate/Filament Power Transformer Primary winding partners.
You see today we have at least 120 VAC as the nominal voltage in our homes, it can dip down to 120VAC and go up to 125 VAC. I have rarely measured anything close to 115VAC ever, even since my childhood days.
The co-conspirator in this electronic 60 Hz drama is the industry standard 115-117 VAC Primary Windings of Mains Transformers. Today it is high time that this standard be revised to 120VAC. But the stubborn Electrical Engineer's who work on paper only don't develop a 'feel' for this problem. It is almost the same fight between Mechanical fabricators and Mechanical Engineers. The Mechanical fabricator will have a feel for what it takes to make it happen in the real physical world. And often the Mechanical Engineer will design something on paper that 'works' but may not be a practical thing to create.
So when a 115VAC Transformer and a 122 VAC Wall Voltage meet, and both of these combine and feed an un-regulated Rectifier + Power Supply, one can easily achieve a boosted AC secondary and develop scary high DC voltages that are sometimes 30-40% over what the design maximum needs to be to operate safely. And not to mention the AC Filament voltages that can creep up and kill your precious tubes.
After 16 burned-out EL34's over the last 20 years, I am able to confirm that the 5-20's now work perfectly with a Hammond 273BX 350-0-350V, 200ma and can benefit from a lower DCR Choke than the one specified in the Mullard book that is specified for a DCR "series resistance" of 200 Ohms, wow. Can someone say poor regulation?
All of this was made possible after my years of head-scratching, red plates, boosted filament voltages and lots of $ spent on EL34 Tubes.
I mustered all of my empirical experiences and made the most out of my attempt to make them work well and not cook precious EL34 tubes.
So what did a Mullard Mains Transformer Specification and the Hammond 115VAC primary do to my tubes..... Plate and Filament Voltages and cooked the following EL34's very well-done. I recall that when I used to power-up the 5-20 amps the EL34's would make tingling noises as the Filaments and Plates suffered on warm-up.
Here are the RIP victims of my original 5-20 build.....
NOS Sovtek Fat Bottle 6CA7 made in the 1980's, very precious, very dead.
RAM Tesla EL34 Matched from Roger Modjetski, so sad.
Svetlana EL34 Winged-C from the original Svetlana Factory, wiped out.
Electro-Harmonix EL34, RIP.
At the tail end in 2006, prior to me putting them away for many years, were a quad of JJ EL34 and these also suffered.
Me said: It must be the Mullard thing that works, wrong, the Mullard EL34 Re-Issues I purchased after the JJ "red-plated".
Me tried: Every trick in the book, bypassing the AC Filaments with 10 Watt Power Resistors to load down the 6.3VAC, only to create heat and suffering.
Me tried: Installing a large series ceramic 25W power resistor in series with the DC Choke to drop the 510 VDC down to a reasonable level, bad idea.
I even used a Variac dialed down to 110 VAC and they limbered forward, with pathetic Sonics. So what was the solution ?
I ran some numbers and calculations that indicated that the Hammond 278X Power Transformer with a single 115VAC Primary as guilty as charged.
Enter a replacement Hammond 273BX. The education I received from the "school of hard knocks" earned me the best overall strategy to developed a sensible Power Supply strategy. The Hammond 273BX Power Transformer is exactly what is needed and the final solution to what is specified in the 1959 Mullard book.
The Hammond BX series provides a choice of dual primary voltage taps; 115VAC and 125VAC. Thank you, thank you, thank you Hammond Manufacturing for understanding your customers and our needs. We no longer live in a 115VAC world, today we measure 120-125 VAC RMS.
Many DIY amps have suffered and faced the same problem I did when designing with Hammond Classic Series Plate and Filament Transformers. Hammond finally listened to our pleas and created a dual primary series of Mains Transformers that are perfectly suited to 125 VAC.
One must understand that all Cathode Biased Tube Amps that implement un-regulated DC power supplies depend on a stable and proper DC power supply voltage distribution. Also, to a lesser degree, but just as important, the AC filament voltages should never be "boosted" beyond their rated nominal values by a power transformer with a low primary voltage winding.
The (V1)/(V2) = (I2)/(I1) and other formulas are the law of the land, and I am a Power Supply Whisperer....... I have been since my days at F.I.T. Electrical Engineering program. But we were not taught Vacuum Tubes... I leaned these on my own after I graduated.
Granted the BX series from Hammond Transformers still offers a 115VAC primary tap should the AC Voltage is your home EVER drop down to 115VAC.
Wishful thinking at best.
The nominal setting is the 125VAC primary tap, punto.
Over the years, with the advent of Solid State and their masses of Regulated Power Supply designs, the 115 VAC was never an issue. The Regulators simply corrected for the boost. Show me a Commercially available Tube Power Amplifier that implements a High Voltage Regulated Power Supply, I am waiting........ tick-tock.
So... after plenty of PSU modeling, paper and pencil Engineering number crunching, my volt meter and mu adjustable Power Resistors, I was able to work with the Leslie Choke I had on hand and modeled a decent current handling capacity with the measured values I needed.
The 273BX Hammond 125VAC tap places the Power Supply in the "safe region" just at or slightly below the Mullard design center values.
The Cathode Bias tubes love to play and hang out here, and the filaments are just spot on value at 6.4 VAC, and the Rectifier Filament is at exactly 5VAC.
So what was the end result: MULLARD 5-20 TONAL AND AURAL BLISS.
All of the DC voltages are now super spot on good, at or slightly below the nominal values published in the Mullard manual.
Now your choice of DC PSU Choke will have an effect on the final DC voltage distribution so choose wisely.
I suggest any good DC PSU choke with a DCR between 50 and 100 Ohms will work. But a DCR of 70-75 Ohms is ideal.
As far as the inductance, anything from 5H to 10H works. And the rating, at least 180 milliamps DC current and Hi-Pot tested to 1000 VAC.
If your chassis has the space, the ideal choice would be a Hammond 10H Choke, 193J with a DCR of 82 Ohms, this one should work perfectly, even better than the Leslie Choke that I use, 10H will lower the Power Supply ripple.
The DC voltages on my Mullard 5-20's are all perfect. Voltages around all tubes are at or slightly below the original Mullard specifications published back in the late 1950's, this my friends is Geek Heaven.
The Cathode Resistors provide a voltage drop of 28-30VDC at 122VAC rms on the primary, just perfect for this amplifier.
At 125VAC this DC Voltage would likely rise to the rated value of 32 VDC across the EL34 Cathode Resistor, perfecto mi amigo.
The Plate Voltage is also great, measured as: Vanode = 431vdc (Plate to DC Ground) - 30vdc (Cathode Resistor DC Voltage Drop) = 401 vdc, plenty good for an EL34 at close to 58 to 60ma of Plate and Grid Current.
The DC Power Supply filter capacitor voltages are now wonderful and place the EL34's bias at a quiescent 24 Watts of plate dissipation, just 1 watt below the 25 watts Mullard 5-20 manual, slightly extending tube life and creating lovely music.
Anyone thinking about a pair of Dynaco MKIV should consider building the Mullard 5-20.
I can take orders to build a pair for you, just inquire. I would use a Hammond Steel Chassis with a ventilated Steel cover and Steel bottom plate.
I prefer the Sonics of a Steel Chassis, but Aluminum will exhibit less propensity to induced chassis currents and noise.
Hammond Chassis are now available in Black Powder coat that look very slick.
The filaments are now amazingly spot-on at 6.4 VAC rms and 5VAC rms, dramatically prolonging tube life.
It is a well documented and empirically determined fact that Tubes last much longer when filament voltages are kept AT or up to 5% below their nominal values.
These are typically 6.3V and 12.6V as published in the Tube Data Manual Books.
I do consider the 115VAC primary voltage tap as "optional" and somewhat unusual.
It is available for the "rare case" where the mains at your home measures 115 VAC.
Typical applications for this tap could be Rural homes located far away from the Electrical Sub-Stations that benefit from the Voltage Drop across long Power Transmission lines after the second to final Step-Down Transformer.
A note of caution about the original Mullard 5-20 design, I have to mention here.
The original schematic specifies 470 Ohm, 3watt Cathode Bias resistors. This places the DC power dissipation through these resistors at 2/3 of maximum. At this heat level, a 3 watter will not last, it may burn out in a matter of days or weeks.
The calculations are shown below in this same section.
My recommendation here is to substitute the 3 watt 470 Ohm resistors and use 5 watt 470 Ohm resistors instead, preferably a high quality type like Mills or Kiwame. If you can avoid the White Sand Types, please do. These types are often 20% off value. You want the Cathode Resistors to be perfectly matched or not more than 2%.
I rely on my friends at PartsConnexion in Canada for quality Mills 5W 1% Non-Inductive wire-wound resistors, yet Kiwame 5 Watt Carbon-Silicon are also an excellent choice.
As far as the bypass capacitors, I use the trusty Sprague Atom 50uF/50V electrolytic capacitors but it is perfectly OK to use 47uF/50V to 47uF/100V Electrolytic cathode bypass capacitors. The 50V rated caps can act like a scary warning smoke bomb firecracker that explodes like a miniature piņata if the particular EL34 should decide to become 'soft' and run-away.
How do I know this? It happened 2x to me due to a bad EL34.
A 50V rated Cathode Bypass Electrolytic Caps will pop, informing you in a smelly smoky way that you have a bad EL34 tube.
A 100 V rated Cathode Bypass Capacitor would likely never reach a 100 V drop so the Cathode Resistor will likely burn when a EL34 fails and the current through the EL34 and voltage across the Plate increases. In any case the idea is to preserve the Output Transformer. You can also fuse each EL34 with a 80 to 100ma fuse.
I agree that a 3 watt resistor can act as a pseudo fuse, but 5 Watts is my recommendation. If you use a 100 Volt DC Electrolytic Bypass cap, the Cathode Resistor will likely fail before the bypass capacitor. In any case, Cathode Bias provides a safety for the amp and the failure of a Power Tube will cause the Resistor to smoke of the Capacitor to pop.
Worst case Scenarios where neither component in the Cathode circuit fail, melting the Primary Output Transformer wiring is pure agony.
When I find the time, I will polish up the Brass Metal plates on the amps with Brasso and some elbow grease, they do polish up very nice.
Oh and did I forget to mention that this design really works best with matched EL34 power tubes duets and 12AX7 with balanced triode sections for the absolute best performance. I plan to measure the distortion and power output when I get some bench time.
My career as a Radio Communications Engineer keeps me busy and often away from my wife and hobbies.
The Power Supply Tubes are JJ 5AR4/GZ34 and the Power Supply Capacitor is a JJ quad 500 Volt Electrolytic with 20/20/20/40uF sections.
I did place the 40uF section at the end of the voltage chain, as I want to make sure that the EF86 gets the best and lowest ripple current while the 20uF sections create a balanced CLC Pi Filter.
So for the electrically curious, last night I finally put the amps belly-up, and after a 20 minute warm-up, took measurements with my Fluke 87V VOM.
Note that this is not a high impedance VTVM so the meter does load down the circuit a bit.
At Frost Fest in Richmond VA I bought a great $15 CONAR Tube VTVM, I have always wanted one of my own.
Tube VTVM's offer the benefit of a high impedance metering circuit that does not load down the voltage measurements on tube circuitry.
Now much to my surprise, one difference I measured was that of the EF86 Screen Grid voltage, I measure much less on the Screen Grid than the values published in the Mullard manual.
It attribute this to either a publication error, (unlikely) in the Mullard manual voltage table, or my Fluke VOM could be loading down the circuit (likelier) as that specific measurement point is a floating DC voltage.
Comments, questions, etc. please send me an email to firstname.lastname@example.org, I would be more than honored to help you build your own 5-20's. If I had the free cash, I would order the Output Transformers from Sowter, Bartolucci or Hashimoto.
But the Hammond 1640 40W, 6600 Ohm Primary Ultra linear Outputs are a bargain!
You can also use the Hammond 1650F 25W, 7600 Ohm outputs as well, as the additional primary impedance has a softening effect on the sound, I know as I used the 1650F for several years on these amps. The 7600 Ohm Primary will deliver less power on the Secondary. But the sound is softer, more pleasing, less detailed.
Yet the correct output transformers are 6600 Ohm, at least 25 Watts.
Using 4300 Ohm will not produce the best sound as the EL34's operating point will be skewed.
I read so many DIY hobbyists changing the high gain EF86 input stage for something else, but the EF86 makes this amp what it is. The EF86 is an ideal companion for Passive Preamps, High Impedance output preamps such as the Dynaco PAS-2/3, and most vintage preamps that lack Cathode Follower outputs.
High impedance Preamps work best when it comes to Bass performance and crisp Mids and extended Treble on these amps that have good hefty gain.
Hammond 273BX 182VA, 350-0-350 @ 200ma DC, 5VCT @ 3A, 6.3VCT @ 5A, this is the one to use for a Mullard 5-20 Monoblock, no doubt about it.
|Mullard 5-20 Monoblock Amplifiers, Mains = 122VAC rms|
|1st&2nd Anode ECC83||325/315||317/307||VDC|
|1st & 2nd Grid ECC83||66.7/60.5||77/70||VDC|
|1st & 2nd Cathode ECC83||69.9/69.9||80/80||VDC|
|Screen Grid EF86||90.7||93.2||VDC|
|AC Rectifier Filament||5||5||VAC rms|
|AC Tube Filaments||6.4||6.4||VAC rms|
If you want to use the exact same Power Supply Choke as I did, the Leslie Choke can be had for $60 bucks. The reason I used this choke, these amps were built on original steel Leslie 147 and 122 Chassis, from old Leslie Tube amps I bought for $20 the pair, and the Chokes came along for the ride. The chassis already had the metalwork from the 60's. And side mount chokes are cool looking.
Here are the Mullard 5-10 Monoblocks in perfect working order. I just happen to be using the same old Sprague 0.22uF/600V Vitamin Q caps that I purchased in 1996 from Audio Electronic Supply, Cary NC. The main coupling Capacitors are the now out of production Auricap 0.47uF/600V, today they offer an XO version.
I am just really tempted to replace the Auricaps with some Jupiter Reds or VenHaus TFTP 0.47uf/600Vcaps V-Caps. But I better just enjoy what I have, and stay away from the temptation to change parts.
Note that I am a firm believer that parts quality is often secondary to good design. Taking a bad design and using exotic premium parts is futile. Taking a good design and using standard parts is smart. Taking a good design and using premium parts is a luxury.
The Mullard 5-20 Monoblocks are sonically stunning, their superb technical design brings out the Hi-Fi in the EL34-ECC83-EF86 combination. There is just something about this EL34 classic Push-Pull design that makes music so mesmerizing. I am sure it has to do with how well this amplifier measures on test equipment. But maybe just maybe it is the high gain of the EF86 which makes this amp so well suited to match "low-drive" high output impedance classic Preamps such as the venerable Dynaco PAS-3. Or maybe just maybe the coupling between the input stage and the Phase Splitter eliminates one mostly unnecessary phase shift inherent to Capacitor Interstage Coupling, thus improving the phase margin within the feedback loop. Or just maybe, Cathode Bias amplifiers hold such a special place in my heart and ears.
I prefer the Sonics of standing Power Tubes on their feet and not their hands.
The Mullard 5-20 design was fully "optimized" by the designers at Mullard and it is a published fact. The Mullard publication (a.k.a. The Mullard Book) presents all of the graphs and of the measurements for best power levels at the lowest distortion.
Basically this is just good design, and this design helps to dispel beliefs or religions around the "detrimental effects" of global feedback affecting music and sound, to my ears, nonsense.
Furthermore, Cathode Biased amps by not being traditionally adjustable the "best" performance is only achieved by matched EL34 tubes.
Tube matching assures even current sharing and lowest IMD/THD.
Also I must mention that the ECC83/12AX7A Long-Tail Phase Splitter also benefits from perfect triode balance for the lowest overall IMD/THD.
After years of listening to many EL34 Push Pull amps, if I was to choose a "classic design" and invest real money in a solid build, the Mullard 5-20 IS the amp to DIY.
I have taken ALL of the guesswork out of this build.
Rest assured that your only difficulty will be choosing the best Output Transformer you can find.
The Hammond Output Transformers with their Inter-Leaved Winding Design work amazingly well, yet there are others out there like Sowter, Hashimoto, Bartolucci, Plitron and even Electra-Print that will impart their tonal shade in the sunny world of audio.
If you ShermanAudio to build a pair for you, I am able to do so with the highest level of confidence. Heck I can even duplicate the exact chassis layout shown in the Mullard Manual and willing to have them Sheet-Metal fabricated.
Having tweaked and worked on this particular build since 1996, there is just no substitute for time and sweat.
This amp also allows the lucky owners to roll rectifier tubes.
My preference is the standard GZ34/5AR4 as it delivers the lowest series impedance and tightest Bass, but you can easily use a GZ37, 5U4G or 5R4GY.
One word of caution, the 5Y3 will not work. You would need to install two in parallel and up the 5VAC from 3A to 4A.
A single 5Y3 just lacks the current carrying capacity this amp demands.
The 5V4... maybe, I haven't tried this one, so be advised not to until I have a second look and test one out slowly with a Variac.
My Custom Version of the Classic Mullard 5-20 Hi-Fidelity Amplifiers
Year 2005, 8 years later.....
Note the wiring spaghetti in the last pictures, sure looks like my first project and what a learning experience. The second to last photo shows the 'charred' phenolic board where the original 3 Watt cathode resistors cooked and died leaving their scorched carbon black fingerprint. I have had my share of problems with the 5-20 design, and they still haunt me today. If one studies the Mullard book and runs a few simple bias calculations, it becomes evident that the EL34's that were used back when this amp was designed were, and I assume this... very, very robust (just look at the price of Mullard El34 NOS on Ebay).
These amps bias the at 25 watts of standing current plate dissipation, ouch.
The Mullard 5-20 designers bias the EL34's pairs right up to their full 25 watts plate dissipation, and this is without any music being played.
This left me wondering, where does the extra margin comes from when music starts to play? My wallet was the immediate answer.
It has become evident, after 6 dead and roasted quads of new EL34's of various vendors that the only way I ever enjoyed these amps was with a Variac.
This equated to about 110 VAC on my Variac and I will explain why.
The 5-20 design requires a unique mains transformer configuration and a PSU choke with a DC resistance of 200 Ohms. These were parts made in the 1950's.
Today's Transformers have lower losses and perform better due to improvements in materials.
Also my hidden enemy was lack of experience and the Hammond 278X Transformer. This made 115VAC connected to 125VAC a dangerously high secondary boost so the tube filaments and the DC voltages are scary high, filaments measured 7.5-7.8 VAC rms, not good.
This was all part of my steep learning curve and the death of several matched quads of EL34's.
The Hammond 278X Transformer had 115V Primary and 400-0-400 Volts @ 200 milliamps HV secondary, 6.3 VCT Filament section with 6Amps capacity and a 5 VAC Rectifier filament section with 3 amps capacity.
Slamming a 115 VAC primary with 125 VAC was my hidden enemy, and recipe for cooking EL34's into submission.
One can run 5U4G, GZ37 and other 3 amp rectifiers no problem, but each will have a different voltage drop. One can also convert the power supply to solid state and save on the costs of the rectifier and separate transformer, including the heat given off. Purists may cringe, yet there is nothing wrong with diode rectification as long as one controls the inrush current with a thermistor of soft-start solution.
Judging the variety of killer diodes available today, one is definitely challenged to use them...... tube rectifiers relegated to the boutique aspect of what is at the heart of this tube madness.... the purists would agree.
Yet despite this, the 5-20 amp always ran HOT and eventually, the future would hopefully allow me to get them working normally.
Lets do some math:
So at 470 Ohms and 32 Volts DC drop
- P = V2/R
- P = (32v)2/ 470 (DC from the Mullard Book)
- P = 1024 / 470
- P = 2.17 Watts, out of a maximum 3 watts as specified in the Mullard tube book too close for comfort, use a 470 Ohm/5 watter please...
This means that (2.17 watts/ 3 watts ) * 100 % = 72.3% of max power dissipation.
Good design practice is to use 50% or less dissipation capacity, so I would use a 5 watter instead, but .... but the 3 watter 'acts' as an EL34 surrogate safety fuse, with a smell that you won't believe! If it cooks, this strange odor varies, depending on what the resistor is made of.
The cathode resistor dissipation decision was probably based on safety for the output tubes, better to lose a resistor than a tube. If the EL34 somehow 'ran away the resistor would fry, open circuit and save the EL34.
A quick calculation will show that the DC current passing through the EL34 will be;
I = V/R
I = 32/470
I = 0.068 Amperes (a.k.a. 68 Milliamps!) now this is strictly Cathode current the sume of Grid and Plate Current.
Calculating the 'Plate Dissipation' will be then be; Power = Voltage x Current
Pd = 390 Volts DC * 0.068 Amperes = 26.5 Watts (Max Pd for EL34 is 25 Watts!) so if I assume that the 'extra' 1.5 watts
comes from the current flowing through the Grid (G2) then I can assume that Mullard 'parked' their EL34's in the 5-20
at the maximum level possible, just prior to the main entry door into 'Red-Plate' land; 25 Watts Plate Dissipation.
Any way you look at this late 50's design, it was obvious that the approach was to put lot's of global feedback and
bias the output tubes hard. This is no longer good design practice, and at that time, the EL34's manufactured were
in my humble opinion probably capable of 30 watts plate dissipation, owing to the dimensions of the internal structure
and physical size of NOS original Mullard EL34's. No current production EL34 is the same physically, the older EL34's
from those days fetch lots of money, now I know why, they can take abuse and work well longer than most
current production EL34 and their variant tubes such as the E34L and also the fat bottle 6CA7's which are Beam Tetrode's and
not true Pentodes. As you can see, in the 5-20, if one does not control the mains DC power supply rail, there will be trouble-ahead,
just like the Grateful Dead's "Casey Jones".
My conclusion is that the 5-20 was designed back in the Golden Years of Hi-Fi where designers used high plate current's and lots of feedback
to lower distortion and improve the amplifier's full power bandwidth performance, which on the 5-20 is guaranteed to be 20 Watts @ 30 Hz.
Tube amps idle like automobiles, you either get lots of torque, or speed, or a middle combination of both.
The recipe for good sound is where to park the tubes and find the ideal transfer characteristics where Power, Distortion and Frequency Response
all somehow come together at a specific Plate to Plate impedance value, a Push pull sweet spot in other words.
It's difficult to optimize all possible tube parameters but different designs will produce varying combinations of each circuit's merit or problems.
Note that 68 milliamps idle current is close to "no man's land" for any current production EL34. The one's that have proven to work so far are the vintage Tesla EL34, JJ E34L, and Svetlana EL34.
The Sovtek EL34WXT a.k.a. Continental slimmies turn cherry red in my 5-20 after about 70 seconds.
The nipple top Ei EL34 are probably the most dangerous of them all, these tubes will turn red in less than 25 seconds after warm-up.
So many worn-out quads, so many memories of great sound. Now I realize how sweet those dying EL34's sounded, this was before I really knew what
Red Plates were all about.
I suggest that anyone wishing to build a 5-20 Mullard based EL34 amp, consider the use of a 350-0-350 VAC, 180 to 200 ma Secondary with a 120VAC rms primary, 125VAC rms even better.
What to do ?
Well, I plan to find a mains transformer that works at 125VAC rms.
Unfortunately, the 'closest' I could get cheaply were Hammond 278X. This mains transformer has the filament and HV capacity a bit over dimensioned, especially with a primary wound for 115 VAC, not the 120 to 123 VAC typical in today's power grid.
I will have to either spend some real bucks on a Sowter or Electra-Print or ...... re-build on a different chassis and use the Angela universal transformers I purchased some moons ago.
You can follow this link to interesting Mullard 5-20 Schematics plus other SE and Push-Pull amps.
Why Tubes, ... and why D.I.Y. ?
Way back in the late 60's when I was 8 years old and jamming to the Beatles and the Monkey's, little did I realize that most of the music that surrounded me was in fact coming from Vacuum Tube amplifier technology. When I got around to owning a hand-me down Dictograph Mono System in the mid 70's as a curious Teen-Ager my destiny was defined for me to one day return back to my audio roots as an Electrical Engineer.
Even back then, Hum was always the main problem. Now I know that the Electrolytic Capacitor technology was still in it's infancy at that time, one reason why large chokes and small caps in the Power Supply was the de-facto design standard.
It all began one sunny afternoon in central Virginia looking through some Stereophile magazines that I placed an order for the 'Mullard Tube Circuit for Audio Amplifiers' book. It took me a while to understand the technical details, not having a real tube background. My degree in Engineering was based on Solid State NPN, PNP, FET, MOSFET and OP-AMPS. Tubes were a similar but different area I had not previously 'academically' traversed and in the 80's were history book technology in college curriculums.
Determined as I was to build the Mullard 5-20 El34 amplifier from the Mullard Book 5-20 Schematic, I began my quest for the ultimate in parts 'scrounging' and minimal investment for pieces, parts, sockets, etc. The nomenclature stands for 5 tubes - 20 Watts. These amps are heavily biased at 68 Milliamps each EL-34 so it's imperative that the EL-34 tubes are able to withstand plate dissipations at their maximum 25 Watt value or else one will eventually see the plates turn a dark red color like a piece of hot charcoal in a BBQ, NOT good.
I recalled that an acquaintance of mine from a radio project in Volusia County FL, had offered but still not sent to me a paid-for pair of tube amps. After a friendly phone call the mystery $40 amps finally arrived. They happened to be an original pair of completely taken apart and parted-out Leslie Tube amps from Leslie Tone cabinets. One was a Leslie 145 amp, the other a 122 model. Both chassis looked just right for a different kind of amplifier. So onwards I delved into what was to become my first pair ever of EL-34 Mono-block amplifiers.
This pair of amps took me about 4 month's to build and about 5 years to hone to perfection (or almost there.) These amps are built of original Steel Leslie amp chassis, they did a number on my hole-punches. I finished the pair with some Oak sides and Brass Sheet Metal to cover-up the already punched unused Leslie holes on the chassis.
After extensive listening by the Bros' (my tube buddies) and myself these amps were named 'The Thoroughbreds' by my best friend. After several iterations, tubes and a jump to true GZ-37 rectification from Diodes, their name still stands today as a testament to my first attempt at building some nice mono-block tube amps.
The Mullard schematic on which these were based has some really amazing sound. There are some modifications and extra things that I added to compensate for the less than ideal power transformer I used like an overrated 6.3 Volt filament supply. Each Monoblock uses 2xEL34 Svetlana, 1 x CV378 British Military Rectifier, 1 x ECC83 Groove Tube and 1 x EF86 Svetlana. I boldly took the first two power supply caps to 100 uF each and the CV-378 hasn't complained yet (despite the numerous warnings on spec sheets and tube news sites.
Last night while listening to some music something dawned on me. When one is working or building amplifiers that use Power (Mains) transformers that are rated at 400-0-400 VAC, it's a whole different world out there for the first and second capacitors of the power supply section.
Transformers that provide at least +400 Volts on the secondary will tend to settle down to a rectified B+ DC voltage of 500 volts or higher on the first capacitor. This of course depends on how loaded down the High Voltage section truly is from published Full Load specifications. Therefore I would conclude that capacitors rated at 500 Volts DC are NOT able to safely filter the raw rectifier 120 Hz from 60Hz (or 100 Hz from 50 Hz) without the chance of failure.
Always be safety-aware when designing with 500 Volt tube power supplies that any transformer that provides +400 Volts AC or more on the secondary is an immediate candidate for +600 WV (Working Volts) Electrolytic or Film capacitors as minimum. What I have yet to try is to stack Electrolytic Capacitors for higher voltage handling capabilities. Something about the voltage equalization resistors that doesn't play well into my engineering techy mind. Maybe it's just high voltage phobia playing with good design practices, or the Cabernet I was sipping :)
Today I installed the new power supply parts..... and the amps are once again operational (w/variac). The fundamental problem remains. The Hammond 278X Power Transformer is too much for these Mono-Blocks.
>Hammond 278X has 200ma, when the original Mullard specification calls for 180 ma. The amp actually draws a steady 147 ma so the mains transformer should not be too large, or else the power supply may need another PI section (Cap and Resistor) to drop the B+ to the acceptable value of 465 VDC. The 278X should be fine, but watch the filament voltage.
>Hammond 278X filament section delivers 6 Amperes AC at 6.3 Volts, this is a bit overkill for a Mono-Block amp that requires 3.5 amps AC. So... the only option was to install some power resistors in parallel with the Filament section to load the darn voltage down to 6.3 VAC from 7.0 VAC. Or light up the amp with enough 0.25A - 6.3 VAC bulbs!
Unfortunately custom wound examples of the required transformer in the USA are as expensive or more than OEM current production models. It would probably be about $300 USD each to have these mains transformers wound. Some people claim that in Europe, custom wound transformers can be found much, much cheaper than in the USA (as stated by another angry rec.audio.tubes forum member), yet I doubt this claim until I can see for myself. Any savings on the product cost is probably offset by the shipping and taxes.
So..... what do I end up doing with my Mullard amps. Running them always on a Variac is not what I really want or the answer to a Home-Brew project.
After some web surfing I do find that Sowter makes a replacement 5-20 Mains clone to the 'exact' specifications of the Mullard manual. I mean complete with the second 6.3 Volt, 2.5 Ampere winding (for an FM tuner which I do not have.)
I sent a mail to Sowter to inquire about these transformers and the US dollar weakness makes these about a $500 dollar investment, too much for me at the moment.
Finally after much soul and web surfing, Steve and www.angela.com has his Angela Universal transformers that will work well. These should run me about $240 USD the pair including shipping. As soon as these arrive and the amps are upgraded I will let you know what the results are!
These transformers don't really fit these chassis, so I am going to first try to squeeze them onto them and test. If all goes well I will then re-do a new pair of these amps of some bigger chassis and leave these as-is with the Hammonds and the variac. After all these were my first tube amps and the sentimental value is still there. With the Variac I can enjoy the crystal clarity of the Mullard design. Anyone who wants to hear an EL-34 produce some Bass, build a pair, you'll be surprised about common misconceptions on EL-34 and bass reproduction capacities.
The Variac solution works so well that these amps will remain on that device for now. I have other more important projects now and the transformer mod will have to wait.
But I can say that these amps sound totally lush and full with great tone and frequency extension. Despite the feedback level they employ, the sound remains clear-transparent without a hint of grain or veiling. I am using now a quad of new Russian made Mullard EL-34's. These tubes sound marvelous at 62 milliamps each. The JJ E34L's don't sound as good, they are much darker, probably close to being run too hard at 62 milliamps. The Mullard design is taking the output tubes real close to the maximum, yet the new Russian Mullard tubes get the job done in a way that other's have failed, even the Svetlana EL34's. So my recommendation for new production EL34's lies with the Mullard tubes, I purchased at the Tubestore.com. One of these developed an intermitted filament connection. Once again these current production tubes require careful testing and screening beyond what most typical vendors use.
Still using these amp with the Variac. I simply dial-in the correct AC input voltage and put the DC VOM on the first power supply capacitor and match the schematic voltage value. The rest of the amp is 99.9% as the original and the currents and voltage settle-in nicely. At the moment the Mullard re-issues tubes continue to sound totally amazing. If I do anything to these amps, it will be way down the road. But for now my only advice would be to spend the money on a pair of SOWTER Mains Transformers wound to the exact specification of the original Mullard amps in the Mullard Tube Amp book. I can't afford them now but who knows down the road :)
The mains transformer, choke and OPT's are available from Sowter and VVT Transformers. I would highly recommend that at least the mains transformer be sourced from any one of these two places. The mullard amp draws about 148 milliamps each. Therefore it's imperative that the High Voltage mains section of the transformer, the 410-0-410 VAC be capable of delivering 180 ma or thereabouts. If one uses an after-market mains transformer with the Mullard 5-20, the compromise will be that the DC resistance of the power supply choke will require the mains voltage to be less than 410-0-410VAC. One can probably get away with a choke that has 78 Ohms DCR @ 10H and a mains transformer like the Hammond 374BX. What ends-up happening is that the under-loaded 6.3 VAC @ 6 A section will offset the voltage loss from the HV section. The end result with provide enough voltage and current to meet the requirements for one 5-20 Mono-Block. With the Hammond 374BX one can get away with using a choke that has less DCR than the original design. This improves regulation and also may have some incremental effects of the sound quality vs. a 200 OHM DCR choke as specified in the Mullard Book.
Amazing how so many people are interested in this amplifier design. I plan to re-do these amps as they do sound amazing when working. Unfortunately I will have to modify the original design to 'slow-down' the rate of consumption of the EL-34's. These amps cathode bias the EL-34's at such a high level that a quad will last at most 3-5 month's before burning-out or frying the cathode resistor.
MerryXmas - Happy Hannukah - December 2016 - Happy New Year
I'm back again after a long absence from my website. New Years Eve a fitting day for my return.
My 20th Anniversary Mullard 5-20 Monoblock Amps are playing again.
These are my very first DIY tube amps and as a fact, I had once thought of taking them apart.
My dear Wife told me not to, she told me that I would regret it some day.
Well words of Female wisdom and here they are, revived on Xmas Eve and now playing on New Years Eve into 2017 and beyond. Classic Vintage EL34 High Fidelity from the 1950's; Sweet!
On the 24th I had to order a quad of new Cathode Bias bypass capacitors.
The solution at hand was to swap the original Hammond 278X power transformers out for a new Hammond model that made these amps come together like the Beatles! I still have the pair of 278X, e-mail me for a price, I will sell the pair. Note that 115V Primary limits the design center.
I ordered a pair of Hammond 273BX this year (summer) and the project just sat there looking at me every time I walked by, a reminder of how procrastination just sits there in the open, it never hides, it just nags.
There were two separate sessions prior to Dec 24th. One was to take apart all of the small things that I had done to the amps and also to fit the new parts.
NO MORE PROCRASTINATION
On Dec 24th I got some coffee and grabbed the amps after some prior work sessions fitting in the new changed parts and fitting the new power transformers.
A NEW TWIST
The new Hammond Transformers offers an electrical means to select between two AC Mains Voltage primary taps; 115V and 125V.
Being able to select between two primary voltage windings allows for amplifier power correction and the stability over long term use, especially on Cathode Biased amplifiers that have ne means for voltage control.
There are two ways to choose the setting of the AC Primary Switch.
One simple way is to actually measuring the AC mains voltage of the outlet the amps are connected to and setting the primary to match the closest value to the voltage measured. If you measure 120V then both settings work, while one becomes Lo and the other Hi.
If you measure 120VAC this is right in the middle. Setting to 115V will increase the voltages while the 125V setting will lower the voltages in the amp. The default setting should be safe one; 125VAC. Make sure the positions are properly labeled on the switch itself.
The other method to determine the best primary setting is to measure the DC voltage across the 470 Ohms Cathode Bias Resistors. The maximum DC voltage is 32 Volts DC, this is from the Mullard manual. So make sure that you set the primary voltage switch to not exceed 32Volts DC, choose a setting that drops the voltage down below 32 Volts; 27 to 30 Volts and you will be fine.
If you measure 32 Volts perfect you can leave the amp alone. But just for curiosity flip the switch re-measure the effect of this change. The normal effect should be to have 32 Volts DC drop down a % and this would make it the Lo setting at 125V and the Hi setting st 115V. This is the typical scenario. But some people may have mains levels closer to 117 Volts AC or maybe as high as 124 Volts AC.
If the DC voltage across the 470 ohm resistor increases past 32V on the 125V setting this means that the AC mains voltage is very, very high at your Outlet, so if you cannot get anything lower than 32 Volts DC on the 125 Volt AC posiion I would recommend getting a Variac.
I would never again operate the Mullard 5-20's with anything greater than 32V DC across the Cathode Bias Resistor and more than 25 Watts on the Plate. I prefer standing each EL34 plate dissipation at just 22 watts, not 25 watts.
Leaving a few watts of headroom on the plate and extending tube life. This does somewhat change the output impedance of the EL34 tube pair but the effects are mostly and almost inaudible. Yes one can measure shift of the Mullard operating point but as long as they are less than the maximums we are in the safe zone.
If the DC voltage drops below 32 Volts DC this means that your AC Mains voltage is at or below 125V. I like to save the tubes so I will almost always operate these Mullard 5-20 amps at the 125VAC setting and be done.
Note that running the amp with 115V Mains on the 125V setting would create 'dirty sound', not enough juice from the Power Supply, it starves the amp, but will still play, just lacking fidelity at higher power levels..
The whole amp just relaxes on the 125V tap, like it went from Cardio to Yoga. I will provide the final voltages in a Table below so we can see that using a 350-0-350 VAC vs. 400-0-400 VAC was more about several things than just a high AC voltage.
The 115V Primary of the original Power Transformers made almost 600 VDC with no load!
I needed to tame this down and tested the amp with some spare transfomers and found that the best compromise was the Hammond 273BX was the best center design.
Today's transformers are more efficient due to improved wire and materials quality.
ONE LAST POOF!
As I was upgrading and tesing the amps on Dec 24th, one of the 47uF/50V Cathode Bypass Capacitors exploded in a dark smelly smoke with lots of fluffy paper pieces. The part was 10 yars old so it surprised me to have it fail.
I ordered (4) Nichicon Muse 47uF/100V Electrolytic Axial Capaciotrs that arrived in record time, 2 days.
MULLARD 5-20 OBSERVATIONS AND CONCLUSIONS
I highly recommend the Hammond 5-20 in stock form despite all of the articles and suggestions to the contrary.
First I like to respect a designers work, so sticking to the original schematic makes this possible, of course with a few small substitutions of equivalent parts for a design from the 50's.
The EF86 input gives the amp tremendous musicality and gain. Yes this is not the ideal to work with Preamps that have more than 15 dB of gain.
Make sure that you use a quality 1 Megohm Audio Taper Volume control to make this another nice feature of adding a Volume Control so a Preamp is not always needed.
These amps are perfect for iPods and Smart Phones, given the correct cable, the EF86 tube recovers the goods and delivers them in clear perfect form. No need for a Preamp here if you want to use a Passive.
The EL34 tube has a sound character that appeals to me, especially on voices and instruments. The design of the 5-20 offers great damping.
You can make a pair of these from parts all available today, and now that I figured out the perfect pair of Hammond Power Transformer and Output Transformer, the build is easy enough. You will need to create the tag boards that Mullard describes in their manual. This makes assembly much easier than the random tag-board point to point method (I also like).
Note that I built these on real Steel Leslie 147 Chassis. Unless you are able to fnd a pair of gutted Leslie amplfiers, I suggest that you use one of the many powder coated Hammond Steel Chassis. You will need metal hole punches for the Octal, 9-Pin, Capacitor holes and Drilling skills. Two metal reamers help tale each metal hole to the ideal size manually, by fitting the part and reaming until snug.
Although I like the sound of an amp in a steel chassis, aluminum is much easier to work with.
Note that on my 5-20's the power supply choke is the original 5H, 75 Ohms Leslie 147 Choke. But you can easily order any one of Hammond Chokes from 5H to 10H and 50 Ohms to 100 Ohms. The series resistance of the choke will slightly skew the resting (quiescent) Hi DC+ voltage of the 5-20. Yet with the ability to choose between 115V and 125V taps on the transformer primary, I call the new feature Hi and Lo mode.
ABOUT THE 115V/125V SELECTOR SWITCH
Reviewing what I have previously stated above:
Hi mode set to 115VAC and this will give the highest secondary voltages
Lo mode set to 125V will provide the lower secondary voltages (Default setting)
The new Hammond taps empowered Primary SPDT switch ON-ON with AC Power Cord Hot (Black) wire through a 2A Fuse then to the center lug on the switch. The 115V and 125V taps solder onto the outer lugs.
A 20 YEAR STRUGGLE
This my fellow DIY lovers was the source of my 20 year struggle with these amps, I had started with a Power Transformer that was highly over dimensioned at 115VAC with no 125V tap. Even though the secondary voltages and currents were okay, a combination of using the original specs in the Mullard manual and the suffering of the tubes using this Transformer with 125 VAC on a 115V Primary.
My memory has me recalling quirks like EL34's making physical noises when warming up. Not to forget the many Red Plates on 80% of the EL34's I tried. I killed a few good quads during my steep learning curve.
My troubles are finally over. These 5-20's are playing through the new year, as I sit here at 10:10am on December 31, 2016 mission accomplished.
I set my self the goal and managed to get these amps going on their 20th Anniversary after 10 years asleep.
As I remember, building these amps in my basement and filming the process (video wil be posted soon) back in late 1996 at Connecticut Ave. in cool Lynchburg, VA. Yes I actually filmed the build, it brings back lots of good memories when I was learning about Tubes and Amps on my own.
The 5-20 amp, although simple, has lots of complexity. The circuit design is superb and the specifications for music enjoyment are about all one needs with a pair of efficient speakers, an iPod/DAC and a bottle of nice wine.
First generation Kismet - still sounding fantastic - Audio Note OPT's are SE-2.6K-20W - VTV Top rated value
Second generation Kismet 2 - James OPT's Model 6115-HS, sounds.... very nice, but not as refined as the Audio Notes !
It's been 1-1/2 years from the last report.....
KISMET-2 is working!, last night I completed the project. One thing about living in the tropics, everything here takes more time than when I was in quiet Lynchburg VA. Project's here in Puerto Rico take-on a life and price of their own, and the main topic is a combination of Heat, Humidity, Procrastination, Lazyness and the high costs of shipping. Also I lack a workshop, sort of like a Formula 1 Driver without a pit area. Every time I want to work on something, my apartment goes belly-up in wire, parts and cusswords as I step over things. This amp is a keeper for now until my buddy decides to re-prioritize some pending issues. I can knock-em out in a matter of weeks, but tweaking the amps to perfection takes month's and even years. Small projects take me only days to assemble and with a dedicated shop this will be the norm.
Note the James transformers and the gorgeous "Yaeger Audio" chassis that look very good together. I highly recommend http://www.yaegeraudio.com the Yaeger chassis for your next DIY amp. John can even have the chassis pre-punched for convenient assembly. These black Powder-Coated chassis are A#1 in quality, they save lots of time and make lot's of sense for first time DIY enthusiasts. Not even Hammond makes such nice chassis, or a complete solution at that. This chassis comes complete with IEC Cord socket, fuse and removable top and bottom covers. Also the bottom cover is ventilated!
Never say never to company travel. Now I have another workshop in Central America. It's back from a three country trip only to take off this Sunday to another country. But.... instead of pouting ....last night after reading a RAT post from John Yaeger, it pushed me out of inertia. The post got me going into the assembly mode away from the whining mode.
Yup, I was able to 'masking-tape' the top cover and began the process of marking the panel for drilling, filing, and reaming, .... heck whatever one calls going at sheet metal with home tools. The pics coming soon after I get my camera charger from the lost and found. I surely hope to have this Kismet 2 amp ready before the end of July. I will focus on this goal and at the same time complete my Dyna-Clones.
Yup I'm back again after travel, vacation and catching-up at work. The process of cutting and punching the holes on the Yaeger Audio chassis has commenced. The chassis aluminum great to work with and very thick, enough to support heavy components. The 2A3 sockets have been a real challenge to mount from below, the hole punch with a proper diameter is hard to find locally. I bought a metal punch locally sourced for 1" conduit (1-11/32" hole), this happened yesterday at Roger Electric in Carolina, PR. The hole results about 2/64" smaller than ideal, but this is easily machined with my Dremel tool into a glove fit for the 2A3 sockets. My other amp was made by using a caveman hole cutting drill bit approach, what resulted required the 2A3 sockets to mount from the top to cover the mess in my original Kismet.
The through the chassis transformer square cut was amazingly simple to make. The metal cutting jigsaw blade sawed right through the aluminum chassis with amazing precision and a simple wood like feel. Wow! the Yaeger Chassis is a real deviation from all Steel Chassis that are a real nightmare to cut and drill.
I am almost thinking that aluminum may be a better material for Tube Amps. My only aluminum amp is coincidentally the Kismet first generation and it sound smoooothhhhh.
With some hi-temp flat black to match the James Transformers, the Kismet 2 amp should be working in no less than 2 weeks time or less, that is if I am not sent away on business travel again. Stay tuned, I am full steam ahead on this project.
Surely was a productive weekend. On Friday evening I almost completed all of the chassis drilling. The Dremel tool was invaluable for shaving in 64ths of an inch holes that were just a fit too small. Now the 2A3 sockets are mounted below the chassis, and the punched holes were taken to a glove-fit.
On Saturday I took the Power Transformer and Choke over to my best friends house for a paint job. The Power Transformer is an old Chicago Transformer purchased about 7 years ago from www.angela.com, the same as on the original Kismet.
The Choke is a Hammond 8H/150ma and also can use the same paint job. Most of today's Hammond transformers look pretty ugly as the painting and finishing quality is at best 80% there. Unlike a pair of Hammonds purchased 10 years ago that had a much better finish. Well enough with the Hammond bashing, they have good products and a heck of a range of models to choose from.
As soon as the Transformer and Choke are ready, I will post the cosmetic external amplifier snapshots. The internal wiring will take me several days as I plan to make this amp an example for a production model later on in the near future.
It keeps raining, and raining and raining here in Puerto Rico. At over 85% relative humidity getting anything painted is futile. As soon as the weather breaks, the transformers will receive the final coat of paint and new photos will be posted.
A day has 24 hours, a week has 168 hours and a year has 8760 hours. Once the replacement Hammond Choke arrives I will build the amp and tweak it until I can say that it sounds like the Kismet 1. I am very curious as to the James Transformer performance, yet something tells me that the Audio Note's probably sound a bit richer in the sonics department. My plans are to complete this project no later than end of May-06. My friend is unable to produce a painted choke despite my patience and waiting now that has turned into month's. I will not allow this project to become a white elephant like my 6550/KT-88 amps that took years to complete because I was depending on other people. One common factor on all of my project delay's was my company travel and my dependence on other's to get work done. It's the "Hasta-Maņana" syndrome taken to the next level..... :-)
As soon as the choke arrives, this amps will be sounding in less than 4 hours time. I will pre-wire as much as I can and await my order from Angela Instruments.
The Original Kismet Amplifier
Here is the latest Kismet article from Fabio Camorani. I will soon post the schematic for my version above. The topology was based on the original Kismet but the circuit was re-worked to have a 6N1P driver tube and also a different Power Supply scheme. We find the new schematic with the 6N1P Russian input tube to offer much better control in driving the 2A3 than a 6922.
Here's a link to the Kismet amp and the original designer.
Back in '97-'98 I was reading many things on the web. Hence one of these was the SE amp phenomenon which was destined to become part of my future. My curiosity about those 'flea-powered' SE amplifier web sites drove me to build one for myself.
So.... after subscribing to Glass Audio, an article arrived in the mail.... The Kismet article suddenly popped into my life and I was very intent on making one for myself. So I decided to build a replica of Fabio's design as best as I could despite my lack of parts.
I ordered a pair of Audio Note 2.6K 20W SE Air Gapped Output Transformers and one Power Transformer from Angela Instruments that was a Horizontal mount N.O.S. Chicago Transformer model. Little was I to know the importance of the quality of the mains transformer in an amplifiers sonic signature. This transformer is still today dead quiet and marvelously black in the background. Too bad Steve and Angela sold out....
The Kismet's name was already decided directly after the article in Glass Audio by Fabio Camorani who as I see now, has already evolved the Kismet into a version 3. My Kismet also evolved, but in a different direction.
Kismet is Arabic for 'fate', and what fate for me to experience +50 pounds 3 watts RMS!
The sound of a single-ended 2A3 amplifier is very sensuous yet tonally correct. Not for the average person, but for the limited bunch of audio radicals who build 50+ pound 3 watt amps and match them to ridiculously expensive high efficiency speakers like Lowther and Fostex. These full range drivers offer the most 'idiosyncratic' sound there is. Now don't get me wrong but low powered SE amps are not for 95% of the people out there. Even though I read that Tony Robbins built one himself. But those who claim that Lowther and Fostex drivers can be a good match for 100 Watt Solid State or Tube amps are probably in the business of selling repair kits for blown and overheated voice coils.
The photo shown is the Kismet as it is today. It took me 48 hours to assemble and 3 years to really hone this SE amplifier together. The final version has a 6N1P Sovtek Input tube instead of the 6922 Sovtek. With the 6N1P I can have 170 Volts on the plate and 5 milliamps cathode current vs. the 6922 that provided much less drive. I hear the 6N1P as a definite upgrade to the sound over the 6DJ8/6922 types. A 5R4GY NOS rectifier and some Sovtek 2A3 single plate tubes completed the project. This amp will only move speakers that have sensitivities in the +94 dB and higher range. Anything lower and one really has to crank up the volume to get a satisfying soundstage but the music collapses and the definition gets lost.
I committed one sin in that I use Global Negative feedback (10K resistor) in my Kismet. The feedback resistor I believe improved the overall sound quality of the lower bass region and also improved the upper midrange as I now hear less glare.
My hats are off to the designer of the Kismet for having inspired me to pursue this wonderful tube audio hobby armed not only with a soldering gun, but my calculator and graph paper as well!
I now have in front of me (5) ECC99 tubes and the Kismet is ready for the next modification, take out the 6N1P driver and replace with a ECC99 JJ. Now that I have this in front of me, I have on order a pair of Kiwame 820R 5 Watt bias resistors soon to arrive so I can once and for all upgrade, tie-wrap and consolidate this amplifier into the reference 2A3 that it always wanted to be. One of the next mods is to add a Headphone Jack for private single-ended bliss and Balanced Connections on the rear panel for testing with a Ralph Karsten preamp!!!
Last night I opened up the Kismet and replaced the series 500 and 300 (800 Total) Ohm Cathode combo as seen in the 'belly' photo a 'Brown Ceramic Tubular' in series with a '5 watt - 300 Ohm Sand' Resistor. The Kiwame 820 Ohm, 5 Watt precision resistors from Parts Connexion look real 'beefy' and to me, promise to lower any distortion artifacts that the other resistors may have introduced in the cathode circuit. The extra 20 Ohms (800 vs. 820) probably made improved the bias level for a softer (more negative) voltage drop across the Cathode for a Vintage 2A3. I will measure to check the value of the bias and also the plate voltage as well to fix the operating point on graph paper. Some tube rolling also caused me to re-visit my Sovtek 2A3's again. It was obvious after 15 minutes that the Sovtek 2A3's were crying for the 5AR4 rectifier in place of the GE 5R4GY Brown Bases. The 5AR4 radically changes the DC voltage distribution causing the amp to run hotter and the tubes to draw more current. I have designed the center points in this amp to be able to accept any suitable rectifier. My Power Transformer has un-loaded taps so the 5 Volt, 2 ampere rectifier section easily handles 2 and 3 amp current draws. All I can say now with the 5AR4 is that my next report will be the ECC99 mod with the 5AR4. The sound of the Sovtek 2A3's confirmed that these tubes are a bit more robust than standard Vintage 2A3 tubes in both bi and single plate configurations. If I run the Sovtek 2A3 at the normal operating point for a standard 2A3 they tend to sound just a bit glassy and edgy, as confirmed by a tube that is over biased (i.e. running a bit cold). To next drop the ECC99 in place of the 6922 basing (6N1P in my case) I need to re-wire only the heater circuit. There is a need to jumper pins 4 and 5 together and move one wire from either pin 4 or 5 to pin 9 to make the socket a ECC99/12AU7 base. Tonight, I will drop-in the ECC99 biased at 6N1P level to do a first check. Maybe I will need to change Plate and Cathode resistors to optimize for the ECC99, the sound and my VOM meter will indicate just where the ECC99 has settled. I will undertake the mod prior to my first cold Beck's and report back mid next week. I will also post the photographs of the upgrades and just maybe.... the ECC99 will be there as a permanent improvement for a Kismet in it's ever evolving state. As I acquire experience, so do my home made amps also benefit. The more I know, the better they sound.
Oh well, not all that glows glimmers! The ECC99 'upgrade' was attempted without positive results. I adjusted everything from the Cathode, Plate and series dropping resistors to attempt to 'squeeze' the correct conditions for the ECC99 driving a 2A3. If failed due to distortion, noise and a terrible buzzing sound when connected to my system. The only resistor that I did not play with was the 100K 2A3 Grid resistor. I just may eventually place a 500K rheostat linear potentiometer in series with a 47K resistor and tweak the tube into submission. But this is only an assumption on my part that adjusting this resistor should tame the ECC99.
Well to turn a frog into a prince I paid attention to optimizing the 6N1P taking the plate voltage up a notch and replacing the cathode and plate resistors. In the plate I used 24K and the cathode I used a 390 Ohm (old carbon 1 watter that actually measures 500 Ohms) and the results were nothing short of a real nice upgrade. I also replaced the Cathode bypass capacitors for 100 uF Nichicon Muse for a smoother more relaxed presentation. The 6N1P was a 120 uF and the 2A3 was at 220 uF, and they were 'cheapo' electrolytics for transistor stuff.
How does the Kismet sound now, like heaven. The Kiwame bias resistors now come through and it is evident that in a simple design as the Kismet, parts quality is immediately audible. Now that I am on a roll, I will re-do the Kismet front panel in a tasteful way. I want to simplify the panel, remove the VU meters and take it down to bare essentials. I just may add a Cathode bias control (or take the design to Fixed Bias?)
Pictures during the re-build of the Kismet. My last mod will be to change out the front Panel altogether without the VU meters. I want to install a nice Power Switch and a Painted Logo that says "Kismet" in some nice color, probably not the white it came with... or maybe I will just leave it as-is...
I am going to build another set of Kismets. One will be using the new James 8 Watt Orient Hi-B OPT's from Euphonia Audio. The rumor is that the other one will use Tamura or Magnequest. The one with the James Transformers is under way but the other one is pending parts ordering and conformation from the buyer. With not too many days at hand and all my travel its a real challenge to build amps and have time to relax.
Stay tuned as this 2A3 phase promises to be an interesting ride. We will be able to compare the sound of the Audio Note's and the James as well as whatever the other buyer intends to pay for the output iron!!
This Kismet will sport a new front plate and a simple look. New pic's coming soon
Kismet taken from it's original state of parts bin assembly ready to be upgraded with better components and some different value electrolytic capacitors. Much smoother sound was obtained.
Original Kismet during the rebuild phase.
Guess what..... I am building a new Kismet using James OPT's. The OPT's are the small James that promise to produce some interesting sound. Coupled to this is the http://www.yaegeraudio.com chassis that are 90% complete, all I need is to punch some holes and get the metal work completed end may. These should be ready soon. Pictures will follow. Another friend of mine wants a Kismet was well. When he gets around to buying the parts I will reciprocate with third version, maybe different than what I currently have and the Kismet 2. It's may very well be a duplicate of Fabio Camorani's Kismet 3 that employs fixed bias and a number of critical improvements that promise to create Single-Ended magic once again. To date no one has ever complained about the sound of the amplifier pictured above. Even when it's not even warmed-up the Kismet has the ability to pull one right into the music, the 2A3 is a tonally correct tube. I have not yet heard a 300B amp that sounds as good as the 2A3, then again I know there has to be an amazing 300B amp out there, I just haven't been lucky enough to listen to one. If one can live without deep bass, the 2A3 SE amps are very sweet sounding.
I replaced the feedback resistors in the Kismet. Instead of 15Kohm I installed 22Kohm. This lowers the level of applied negative feedback and opens-up the sound of the Kismet. I will install a switch for the purist types that believe it's a sin to apply negative feedback to any SE amp design. Yet to my ear a little feedback goes a very long way. I notice the Kismet looses control of the Bass at high volumes without feedback. With the feedback the lower and upper frequency extremes remain very stable until clipping which then makes the Kismet amp sound loose and hard. After comparing the sound of the 300B to the 2A3, I have to say that I prefer the 2A3 sound over the 300B, it's clearer. After the summer, I am may have to build a Stereo 45 amp to have another toy to drive my Lowther speakers with. No name for this amp yet, but It shall be a product of my own design using SE CAD.
Original Aurora with Vintage NOS Magnetek-Triad S-42A output transformers, good iron, but not quite Hi-Fi.
Improved Aurora with Hammond 1645 output transformers, better sound, musical and detailed.
Aurora Power Supply sketch:
Update Feb 3, 2019
After many years of enjoyment, I re-visited the design while looking at the ratings of the Russian 6F6 Tubes. I determined that I really could drop the series resistance of the Power Supply Chokes down and squeeze a few more volts out of the DC supply. I went from Hammond 159M to 156M that meant a reduction from 256 Ohms down to 86 Ohms DCR. This allows more headroom to the amplifier circuitry. The fidelity improved. I have not measured the output power but my last measurements were 7 Watts per channel into a 100 Watt 8 Ohm Dummy Load.
Now enter the original Aurora build experience >>>>>
Now for my third home brew project I had purchased a clean empty aluminum "aluminium for those from the UK" onto which I wanted to blend art and science into what was to become one of the most interesting projects in my life.
Back in 1995 I visited my mother back in San Juan, Puerto Rico. Due to the fact that vintage parts and tube stores were already disappearing in '95 I decided to visit trusty old 'Robert's TV Parts' on De Diego Ave (still open for business) in the Puerto Nuevo Metro area. I managed to buy a NOS pair of Magnetek-Triad S42A 4.5K 50 Watt 4-8-16 Non-Ultra linear output transformers triode 'three wire primary.' Now that I had the iron, I needed a schematic. What was I to build with this iron ? after spending 120 dollars for the pair!
These transformers went totally dormant at that time as I was building the Mullard amp as I needed 6.6K Primary OPT, so the 4.5K did not inspire me to experiment so I saved them for a rainy day. Plus the Mullard amp needed an ultra-linear transformer (five vs. three wire primary.) So these transformers traveled all the way back to Virginia where they remained and returned back to Puerto Rico in December of 1996.
Eventually, in late 1998 sometime after Hurricane George I surfed across a schematic on the Claudio Bonavolta web site called the 'Olson Amplifier'. And guess what, it required a 4.5K primary, non ultra linear OPT, just like my Magnetek. At first I was taken back by the use of 4 x 6F6 tubes in the output, knowing quite well that the Glass Version 6F6G and 6F6GT were hard to source and not sure whether I wanted to build a metal tube amp.
But I endeavored and voila, I built the Aurora Amplifier and it was aptly named the 'Aurora' after a classified High-Tech Pulse Propulsion Plane that some say does not exist. Well the my Aurora does exist and here it is in all its 14 tubes of glory.
The amplifier uses Hammond Transformers and Chokes, the S-42A Magnetek Output Transformers and the following tube lineup:
8 x 6F6G (Russian 6F6 glass version), 2 x 6SN7 Tung-Sol's, 2 x 6J5 Fivre Italy and to finish things off 2 x GZ34 German RSD.
How does Aurora sound, well I can only say that going from 6F6 Metal Tubes to the Russian 6F6G was a revelation in clarity, the glass tubes taking the upper hand. The last upgrade was to install dual 15H, 100 ma, 262 Ohm DCR chokes to really smooth things out. These chokes added a level of dimensionality to the sound that I find hard to describe in words.
Last night I decided to fire-up the Aurora again after a period of calm dormancy. The first thing that I did different was to replace the GZ34 rectifiers with 5Y3. My original design was based on dual 5Y3 so out came the GZ's and in went the 5Y's.
I did not measure the voltages, but I can surely tell you that the 5Y3 rectifiers in this amp provide a far more textured and layered sonic presentation. With GZ34's installed the rectifier section runs at 50% capacity. But.... with 5Y3 the rectifiers are working at their designed capacity of about 110 milliamps each, max is I believe 125 ma for 1 tube. With two 5Y3 in parallel the Voltage Drop must be about 30 volts across the rectifiers instead of the usual 60 Volts for one tube.
So here is the Aurora. As you can see I spared no expense in using top quality parts for a real no-costs-barred approach. All that is left is to have a custom cage built and complete the project by installing the Bass Contour and Volume Controls that were on the original schematic.
Well time to improve the Aurora. After extensive listening over the past several years I have decided to improve the Aurora. These Magnetek-Triad S-42A output transformers are beefy, yet only provide 30Hz to 15 KHz performance. I know what you are thinking, that it should not make a difference to put a 30 KHz capable transformer, but I think it does. So I ordered a pair of Hammond 1645 5000 Ohm 30 Watt transformers. The S-42A transformers may go up for sale to interested parties who want to buy a pair of "Greenies" and re-paint them to suit the application.
Basically in two separate sessions, one mechanical and the other electrical, I completed the upgrade to the Aurora last night at 11:30 pm. Initial listening tests confirm a bit less output power from the new 5000 Ohms Center Taped Hammond', yet this was fully expected. But the improvement of sonic inner detail and pure 'listen-ability' was simply superb, the noise floor went straight down to the basement.
Opening cut of Keiko Matsui's 'A drop of water' was as dynamic as I have ever heard, and the lushness of the opening of Acoustic Alchemy's new CD release American/English made me smile like an idiot at 11:45 pm after a session with the soldering iron and a cup of warm Earl Grey and Honey.
I am very satisfied with the Hammond output transformer line, but cannot say the same about their Plate/Filament products.
Hammond OPT's are consistent products and perform very well. It's a real pity that their line of standard, classic and plate only power transformers do not have practical secondary voltages and capacities for Stereo amps, but very odd combinations of Plate and Filament sections that either require one to compensate or design around them. Extra's such as filament loading resistors and voltage dropping PSU sections may be required to pad down the typical High Voltages that Hammonds provide.
Furthermore the Hammond power transformer line does not sell anything that can be used to create a practical Stereo EL34, 6550 or 5881 Tube amp on a single chassis. Either the H.V. is way under or over-dimensioned for mono-block operation, or the filament current capacity is way over dimensioned. One can end up cooking tubes for as an un-suspecting DIY enthusiast.
Despite this a strange amp like the Aurora was made a reality by a Hammond power transformer I picked that was almost a custom fit design, owing this to the eccentric nature of the Hammond designers and their current production line of Plate and Filament Transformers. I have sent several mails to Hammond about this but no response has ever been received.
After an extensive listening period I do conclude that the Hammond OPT was a success. I would recommend that you give this amp a spin, but try and design the amp with current production tubes, not the rare 6F6 like I did. Yet if you can get your hands on a quantity of the Russian 6F6's go ahead. This amp sounds neutral and un-colored, flat is the word I use. It works well with and without global feedback. At the moment I am using a Brown-Blue-Orange 1/2 watter bypassed by a 330pf mica cap. The feedback tames the upper frequencies that can be very bright with no feedback. But i think that I need to reduce the feedback level by increasing the resistor in the loop. Maybe I will change it to 22K vs what I have currently installed. Or maybe I will put a potentiometer and have variable feedback. The final mods will be to install the Volume and Tone Control that were on the original schematic.
Two years later and the Aurora continues to play well. I decided to completely eliminate the Global Feedback. The amp did not need it and originally, the designer did not implement this. What I can say about this amplifier is the following. Such a platform would make an interesting amp that uses 6V6 tubes instead of 6F6G which are hard to find today.
Left Channel Mono-Block
Right Channel Mono-Block
What to do next? I asked myself this question many times before engaging my next home-brew project. After having struggled with so many tag-boards and swapped parts, I decided to go all out on a pair of 6550/KT88/KT90 Monoblocks with the Triode Electronics MKIII upgrade circuit board.
The name Tornado comes from the pair of British Pilots who were shot down due to friendly fire in the recent Iraq conflict. After the Kismet these amps already have names before their launch date. I chose to honor these fallen warrior pilots as they were willing to give the ultimate sacrifice.
As these Monoblocks will be based around the Dynaco MK-III Upgrade Boards from Triode Electronics. I expect to save time in wiring by using these flexible upgrade boards. In addition, I am also using the MQ-782 Dynaco Replacement Magnequest Transformers! I expect these replacements to be superior to other equivalent transformers. There was a specific Hammond model that I could have used that sported a 50V vs. 60V Bias Tap, but I feared a restricted bias range. In a pinch I am sure that the Hammond's would have worked fine with the proper adaptations.
The center piece of these amplifiers will be the output transformers. A pair of real Potted M-Double C-Core Bartolucci Model V 4.3K 60 Watt Ultra-linear Microcristalline impregnated transformers. Wow! I can't wait to bias these up with some KT-90's and letter-rip.
As these amps are currently under construction, I will be adding pictures of the whole process from parts to my planned Chrome Plating of the chassis. If the chroming place permits I'll be taking some pictures there as well.
As you can see I started by placing all of the parts on my workbench and went straight into the Masking Tape process to simplify drilling and help guard against most scratches. If you follow the sequence of pictures, it will give you an idea of my approach to putting together some tube amps from scratch.
I decided to solder the 9-pin Ceramic Tube sockets directly to the PC boards. The original strategy was to mount standard 9-pin sockets to the chassis and to run cables to the PC boards. This was because I plan to do allot of tube rolling. These MKIII boards can accept, with a special filament switch design (worked this out already) either 6N1P, 12AU7, 6CG7 and 12BH7 tubes in the V2 position.
This requires that I install the PC board under the chassis with metal standoffs that are 0.625" long so that the board under the chassis and the sockets line-up flush with the surface of the amp base. This I believe is a better solution in the long run, keeps wiring down to a minimum as originally designed.
Visiting my friendly local New Electronics Center Store on Fernandez Juncos Ave. "BTW -The place is for Sale" also picked up a pair of Shurite VU Meters identical to the one's I used on the Kismet amp. I am real tempted to 'add' the tube based low level driver circuit shown on the Shurite VU Meter instruction sheet. This keeps the VU meter out of the secondary side of the transformer. This is still under consideration at this moment.
In addition, as I plan to install a 'dual-bias' configuration on each Mono-Block, I ordered today from Allied Electronics four Model 3307 Shurite Panel Milliammeters. This will allow me to install a DC current bias meter per KT88 (6550) so that I can read the bias current directly from the meters. It's far more 'real-time' than having to fumble with calculators and volt-meters. I can always install a meter shorting switch to take it out of the circuit, besides the Shurite specification is 1 Ohm DC Resistance for the meter. Does not hurt to leave them permanently part of the Power Tube cathode circuit.
Once the parts arrive I will continue the construction process. Thanks for visiting The Music Room!
Unfortunately 'the company' that sent me the 1-3/4" Sheet metal punch sent it Fedex 'Ground' to Puerto Rico... duhhhh! Did they think that Puerto Rico was driving distance from Kansas? Due to this, I have the project on-hold until the sheet metal punch arrives. I cannot continue the project without knowing how the (4) large metal can caps will fit prior to doing anything else. One must not indiscriminately make holes in any chassis unless the complete picture of how the parts will fit is at hand! Stay Tuned..........
Still waiting on the Hole Punch vendor!! Apparently the company moved to another state and the new 'operation' is not as efficient as the older one. So much confusion to send one Hole-Punch, geezzz. I had placed several days earlier, two orders from other vendors that were received just days later. Seems like the company that I buy my Sheet Metal punches from has lost the 'edge', let's see if they can get the part to me soon.
I will be away on travel until early September so the Tornados will face a hiatus for now. In the meantime, one piece of good news. My friend has a plasma metal cutting system. This means that I will be able to make the 'square' holes required for the Shurite 100ma meters.
As I mentioned before, I plan to install Edgewise 100 ma DC meters, one per power tube. This way I can have a visual indication of cathode current while I measure the DC voltage drop across 1 Ohm resistors in the cathodes. Note that I say 'Cathode' current as it's the sum of the Plate and Grid currents. Some folks don't realize that on Fixed Bias amps the grids will draw some current often making the amplifier operate in class AB2 vs. AB1.
I may just install the meters in 'series' with the Plate leads from the output transformers to the plates of the tubes. But.... floating milliammeter's at such high voltages does not give me the 'warm-and-fuzzy' feelings that I like to have when I build stuff. Let me investigate this and I will get back to you right here in 'The Music Room!'
Unbelievable, still waiting on the 1-3/4" Hole Punch! Heck, if Greenlee made that specific size I would have gone to Grainger weeks ago! Now I am stuck again waiting on a 'phantom' Fedex delivery. The vendor finds it impossible to send to me the Fedex Airway-bill number unless I ask repeatedly. I am about to 'cancel' this order and get the punch directly from the EOM in the U.K.! This is one example of an order that has gone seriously wrong! I am on 3 weeks travel so the Tornado delays continue. As some people believe ' Good Things Take Time.' I did snatch a quad of KT-90 Type III on Ebay in the meantime! Thanks again for visiting the 'Music Room!'
The 1-3/4" sheet metal punch finally arrived!! Seems that 'The Company' (I won't order from them again) insisted on Fedex ground from Kansas to P.R.? This meant that Fedex does 'not' do the local delivery, another company takes this role?.?.? How confusing can they make things! Anyhow, I have been away for 3 weeks now on company travel and arrive back home this Friday the 5th of September.
Current plans are: Complete the physical assembly of the chassis, then go to my buddy's "Plasma Metal Cutter." I want the square holes for the DC Milliammeters to be as sharp as possible. Whether I install VU meters is still up for debate. But probably not, as they require 2" round holes on the front of the chassis inevitably weakening the already 'more-than-twice' perforated chassis. Besides, I can use the VU meters for some other Preamp project down the road!
Will post more photos after the next round of metal work! Thanks again for checking in with the 'Music Room!'
Signs of life!!! I'm back after such a prolonged hiatus. As you may recall the 1-3/4" Hole punch fiasco took so much time to get here that it threw the whole project out of schedule. Over the holiday's I have managed to continue the chassis metal work assembly of the Tournados.
From the pictures you can see that I managed to install the Oil PSU capacitors from www.angela.com which are actually motor start capacitors converted into a DC filtering application. I have to believe that these capacitors have to be a marked improvement over the standard Electrolytic types, and over twice the size at that!
The PSU caps were so big indeed that I chose to mount two of them thru-the-chassis and the other two under the chassis. They lined-up quite well underneath. In addition, I also aligned the Octal tube sockets also from www.angela.com after having securely mounted the PSU caps. It is always better to go step-by step, from the larger to the smallest components in any amplifier project. First mount the Transformers, then the caps and the terminals and finally the Octal Tube sockets. One then finishes with the tag and turret boards located strategically for under chassis wiring layout purposes.
This time around, I believe that I hit a bulls-eye as the Power Tubes and the Rectifier Tubes have very good clearance between themselves allowing for much cooler operation and prolonged tube life.
Now I will be concentrating on getting the rectangular holes for the Shurite 0-100 ma Milliammeteres (one per power tube) machined to exact specifications on the front panel of the chassis. This amplifier will feature one meter per power tube so I can read and adjust the bias idle current on each Power Tube individually without having to fuss with Tip-Jacks or DC Voltmeter voltage drop readings across known power resistors in the Cathode. By installing permanent bias meters I can always 'see' the bias current and will immediately know if there is any problem developing as the Power Tubes age and the bias begins to drift.
I have included the latest pictures of my project. The bottom received some rubber feet and the top cages were also drilled and fitted as well. I will post these pictures and the larger versions of the newer series as soon as I get the chassis back from the machine shop.
Until the next report, keep jamming and enjoying the Music Room!
Update..... Today is a national holiday in Puerto Rico, Eugenio Maria de Hostos. I called the machine shop and they were open.... Hooray!
I took both Hammond chassis and one single Shurite 0-100 milliammeter into my favorite machine shop a bit after lunch time. The name of the place is "RO-MA Welding and Machine Shop, Inc. / (787) 767-1406."
Mr. Rodolfo Mayol's machine shop is located at the end of the 'golden-mile' in Hato Rey. Many a time I took to Mayol's raised floor tiles from communications rooms for drilling 2" to 3" round holes. If you are familiar with these types of tiles then are double heavy gauge metal with a cement filled interior. Not the job for the average drill-press!
I spent a good 30 minutes with Mr. Mayol explaining exactly what I expected. Basically the rectangular holes I need machined need to measure 1-43/64" wide x 33/64" tall (1.672" x 0.5157").
Mr. Mayol said that he would have to manually make the holes and finish them with a file. Let's see how they come-out. He promised them to be ready next Saturday the 17th. Heck I could have gone at this with my saber-saw and saved some $, but I know that the finish would have been rather rough. Ideally these types of rectangular holes should be 'punched' out by walking a rectangular punch machine across the hole and finishing the burr's off with a file. Mayol did not avail of such a machine, and Sescom is in Kansas, too far and expensive to send the chassis out for this task.
In the meantime, I plan to begin to paint the DC Chokes and Power Transformers the same 'Volvo Silver-Grey' Inboard-Outboard Motor paint to match the chassis. On the Bartolucci transformers, my plan is to simply apply some elbow grease and wax it up with some high quality wax.
There are some 'cheesy' cursive Black-Gold labels on the Bartolucci transformers that I plan to hopefully remove if I can do so without damage to the Gloss black paint finish. If not they stay put.
Enjoy the Music Room..... see you soon!
On Saturday the Machine Shop was closed....., I should have called before driving out there. Today is Martin Luther King holiday so I am not sure whether they will open or not. Let's see,,,,,
Having been raised in Puerto Rico, sometimes we may get a feeling that our personal priorities do not match other's priorities. Today I stopped by the shop, nothing, nada.
The chassis and meter are still on the table where I left them.
Patience Richard, lots of patience, 12 day's worth of patience! This is what we Puerto Ricans refer to as the "Ay Bendito!" clause, a sad story about how one is so busy that the work is late.
Well, I picked-up the chassis as nothing had been done to them. Guess it's me, my drill, files and the Dremel tool. Stay tuned until further notice on my progress!
Due to my ongoing move and living out of cardboard boxes I must say that the project is now on Hold until after I move into my new house. This will give me the extra time to have the chassis machined properly for the rectangular holes needed by the milliammeters.
Wow, I'm back from a world of distractions. All of my things were boxed-up since late February. I have since moved into my new condominium and have now started the process of un-packing. My first order of business will be to make space for a work area to continue the project. With my current travel situation, it will challenge me to finish this project by midsummer.
Heck, I have to be honest, the chassis are at my friend's house now. My buddy has a plasma cutting system. With this he plans to slice-out the rectangular holes so that the Milliammeters fit snugly. Once these holes are cut it's a simple matter if soldering and wiring time before I fire these-up!
I am a bit embarrassed by the lack of progress on this, my latest project. But nevertheless I'm back again. After going through my EL34 Monoblocks and revamping them to near perfection, my gaze is turned towards my second home brew mono-block set. The chassis are the same as before. My friend got caught-up in FEMA work (it's been a busy year for those people) and now I am focusing on getting the square meter holes machined and the chassis chromed or maybe gold anodized.
My week of vacation starts on 12/25/04. Maybe by then I will have the chassis ready for the chrome job. Once again, stay tuned.... The Music Room!
Just a bit of insomnia tonight.... (it's 4:45 am, Tuesday).
The chassis were taken to the premier machine shop in PR, TEMSCO.
The owner and two machinists spent about 15 minutes going over my requirements. All looks AOK for these chassis to be ready soon.
Also spoke to Caribbean Electroplating and once the chassis are ready to be plated, the owner promised to show me their complete line of electroplated finishes.
I think that these amps should be ready by end of February the latest.... but this has been one long ordeal. As soon as I deviate from an all home-brew project, delays are introduced. If it was not for my being so critical about the electronic meter holes, these amps may very well have been built with Tip-Jacks and some calibration resistors in the cathode.
My patience is at the very end now, but I do see some light (music) at the end of the tunnel.
I just may put this pair up for sale..... or they just may be keepers. It's coffee time!
Voila, called and they are ready to be picked-up. Friday is Pay-Day so I will wait until then as I don't know how much this job is going to cost me. Next stop is Caribbean Electroplating. My plans are to have several items electroplated as well so I can get a better cost/work/price ratio. I will post photos of the machined chassis for documentation purposes. I ordered an extra 10 MB of web space as I am already close to 19 MB/20 MB and often receive warning messages from my ISP to delete some files or PAY!
KT-88-6550-KT90-MONOBLOCKS...... COMING REAL SOON RIGHT HERE AT "THE MUSIC ROOM!"
Couldn't wait (you probably guessed that already.) I picked up the chassis this morning at 9:30 am sharp. TEMSCO did a very professional machining job. The masking tape is even intact, this must have been done with a special sheet metal cutting tool.
Anyhow, tonight I am putting the amps back together physically (as in the photos above) and am going to sketch the preliminary wiring plan. This will enable me to install the final tag boards and bias potentiometers to complete the chassis metal cutting.
My next stop will be to Caribbean Electroplating in Cataņo (Barrio Palmas), Puerto Rico. Over the phone the owner was very nice to me. He spoke at length about the electroplating process, which went pretty much over my head. Hints at several types of finishes in different colors makes me very curious as to what this would look like in the end.
At the same time I will also drill and punch some ventilation holes in the bottom covers. The metal cages I plan to paint with a matte-black BBQ paint which looks very smart and resists heat.
It's amazing why I did not take these chassis to the best Machine Shop in Puerto Rico in the first place. My gratitude is extended to Ing. Cegre for understanding the needs of an eccentric-eclectic hobbyist. I am sure that TEMSCO enjoyed working these Hammond Chassis as much as anything that they have done lately. I even had to take all 4 Milliammeters so they could custom fit each one to the chassis in case variations in the dimensions of the meters would cause mis-alignment of the mounting holes. If they ever knew how loud I can crank-up "Reise Reise" by Rammstein, they just might have sent me home.
I am on my way to Bartolucci Heaven! My friend has several Quads of KT-88's, one of these if a JJ Blue Glass type. With so much riding on these amps, it's only natural that my anticipation for how the Triode Electronics boards sound as compared to a real Dynaco MK-III is killing me!
The photo is posted above showing the Machining of the chassis for the Shurite 0-100 ma edgewise meters. Each 6550/KT-88 will be fused with a 100 ma fast acting fuse. In case the bias ever climbs past 100 ma so I don't ruin the meters. Project under way again, soon to be more photos!
As soon as I return from my next business trip, I'm diving head first into these amps!
I leave to Sweden next week so I begin this project when I get back towards mid-May.
HAPPY NEW YEAR!!!
Wow!!! I say Wow!!!
It's been waaaayy too long, +10 month's to be exact. My last update was last year. Time flies when we hit middle age!
I had lost my Web design SW so I finally have Front Page 2003 loaded and I am happy to be back.
The KT88/6550 Mono-block project is sounding! I am pleased to say that the amps are 100% complete.
Somehow over the Xmas holiday's I managed to pull myself together, plug the soldering iron in and after a series of decisions on paint and inner layout of wires, tag boards and other goodies, this set the stage for final assembly.
I left the original Hammond Grey Powder Coated finish alone, messing with this meant a questionable spray paint job in tropical humidity.
By masking taping the chassis for the drilling and holes, I inadvertently saved the original paint job. When I pulled off the masking tape, lo and behold a practically scratch free coat lie underneath. Just a bit of Goof-off and they were as good as new.
Next the decision was to lightly sand and paint all the transformers with BBQ Hi-Temp paint a flat black. I used Krylon paint. The results were OK.
The amps were to be put together with the parts at hand, no more orders (so I thought.)
Needless to mention now, when I opened-up the 4 x Shurite 0-100 ma Edgewise meters from their boxes, to my dismay the Machine Shop had somehow damaged all four of them probably through some rough handling, my best guess. One meter looked like it has been opened-up and repaired, it simply fell apart in my hands. I assume that this particular meter had fallen from a height not kind to analog meters. This is probably the point when I received a phone call to take in the other three meters as they argued that they were necessary to complete the work, huummmm... now I know why.
As all four meters were broken, it was then a matter of calling Allied Electronics with my Credit Card for another 4 pieces. The total cost of machining 4 square holes and one of the crooked was now over $250 USD. It's a sad scene here in San Juan. Lot's of talk, very poor quality.
Once I assembled and wired the amps another 3+ weeks of troubleshooting followed. There were strange noises from the speakers, Svetlana Red Plates, and those small quirks such as loose wires, cold solders, loose speaker and RCA connectors plus a 'fried' 10K AC Splitter Balance pot. I am happy to say that all these issues are now resolved.
These amps are heavy enough to limit my moving them around. Each time I have to pick one-up, I breath heavy and hope I don't throw out my back. Not sure about their weight, but they must be about 40 lbs. each.
These Mono-blocks sound absolutely amazing. It's nice to step-up to the 60 watt/ch category after years of 35 watts/ch.
The Dynaco MK3 Triode Electronics PC board, and some good quality parts the project have come to their proper conclusion. I must mention that in my amps the only electrolytic capacitors are in the dual bias supply kit, also purchased from Triode Electronics. The main P.S.U. capacitor values are identical to the Vintage Dynaco MK-3, but these are Oil types ASC 30uf-20uf-20uf-20uF motor start caps (I think they are for Motor starting) purchased from www.angela.com.
With my ear flush to the tweeter, I can't hear the stereo-typical Tube Amp hiss characteristic of my other Vintage tube amps. The only other amp I own that is quieter is my Stereo 2A3 Kismet amplifier.
The front panel layout consists of a Red Pilot light, AC ON (Radio Shack), a green LED light (Radio Shack) is connected to B+ via a 150K 2 Watt Resistor to indicate PSU cap charge and discharge. It also acts as a Bleeder for the PSU when the amps are powered down off. An internal 6.3 VAC/.25A lamp and lamp base (Radio Shack) illuminates the meters.
The two Shurite meters display direct Cathode Current. Each meter is bypassed by a toggle Switch (Radio Shack) to effectively remove the meter from the Cathode circuit when listening to music. Each small toggle switch places a short across the meter, thus connecting the Cathode's direct to DC ground.
The Bias controls are located in front of each power tube. By rotating these CCW the cathode current increases, CW decreases this current. These controls are a rather intuitive way to wire the pots, the rotation of the control follows the meter movement.
If you want to build a pair, just let me know and I can help with any advice and tips. You can save money by using all Hammond transformers. Yet I highly recommend spending extra on using the Triode Electronics Dynaco MK-3 Power Transformer and the best Output Iron you can afford.
This amp is a foul-proof unit and works very well. The PC board saves lot's of hassles and headaches.
My only regret...... not using Digital Milliammeters on the front panel. Yes the Analog meters are nostalgic, but they have some issues, especially when mounted on steel chassis. This brings me to another point, I would build these amps again on Aluminum Hammond Chassis or have Jon Yaeger build and machine some ready-to-go chassis. Aluminum material is easier to machine and is a non-magnetic. But I doubt that the Aluminun Chassis from Hammond would have the gauge to properly support the weight of the Transformers and Choke. Jon's products do have the thickness, especially the top panel, believe me they are beefy.
AIKIDO Tube Stereo Preamplifier Project
For the curious about what that toggle switch in the rear does, here's the Aikido's grounding scheme with the Pin 1 problem solved.
Power Supply Unit
WHAT.... ANOTHER AIKIDO !!!
This is what I love about the DIY audio community, there are plenty of good designs and PC boards to stuff.
With this Tube preamp project I wanted to achieve a level of success that my first attempt didn't.
My first DIY preamp project had a great Line and Phono Stage circuit but a bare bones power supply.
Coupled to that, the aluminum boxes in which they were built resulted in many compromises as they were too small.
As you can see, I rummaged through my parts bin for this project. I can start by saying that this was fun,,,, well maybe, more of a learning experience as you can see by the photos and schematics of my first foray into Preamp land....
Rozenblit SRPP married to the classic RCA RIAA Phono Circuit.... worked.. yes, noisy, yes, power supply CHEAP...
I learned my lesson, never skimp on a power supply. Below is my 2007 Preamp project. Somehow it worked, but the performance was not there.
Preamps are about the most frustrating projects as noise is everywhere. Building a preamp one learns about wiring topologies, hum, component placement and all the extras.
As you can see from these photos, I also installed a BENT AUDIO 50K Alps Stereo Remote Control unit on the Preamp. Any couch potato will vouch for this option. The Aikido is a great design.
Mr. John Broskie who designed the Aikido circuit is a very clever designer. Building on existing topologies, John will see things and is able to come up with upgrades to and positive changes on existing circuits and topologies.
Where I find the Aikido a good Preamp baseline is in it's ability to present a relatively neutral sonic character, with low noise and high performance.
I would put most of the effort into building the best Power Supply possible, as the Aikido design speaks for itself. As far as board stuffing, I feel that nothing short of best premium components is a good idea. Heck now that one is spending money on a Aikido PC board, why not put in the best stuff.
In my Aikido board I have 100% Tantalum Resistors, Mundorf Silver/Gold capacitors and top of the line NOS JAN 6SN7WGTA tubes for a marvelous sound and musical texture.
Here is the Aikido Power Supply which was mostly defined by the power transformer I had on hand, a NOS Military Encapsulated UTC.
After several weeks of tweaking, I had to install a separate filament transformer as the winding section of the NOS UTC simply did not have enough extra capacity and the filaments were running
starved at 5.5 VAC. Now that I put in a TRIAD 4amp model, the filaments are operating from 6 to 6.2 VAC.
Project Parts List