Converting A Williamson KT66 Amplifier To 6B4G (2A3) Class-A Triode Amplifier
HTML-код
- Опубликовано: 6 окт 2024
- I have been listening to a mono (one channel) amplifier in my shop for a few years and I prefer it to a higher end stereo amplifiers that I have. I decided to change a Williamson style amplifier from beam power tubes (KT88) to vintage 6B4G triodes. In my opinion, the detail and musical value of the triode amplifier is significantly better sounding than the KT66 amplifier
![Natanael Cano - Amor Eterno [Official Video]](http://i.ytimg.com/vi/EO4prCs6WpM/mqdefault.jpg)
Ok, I'll admit it, this video had me looking through my junk box for another set of transformers. It just so happens I have what I need. I won't be modifying my daily driver "ElPaso inspired" 6L6 amp I built last year, but maybe I'll get snowed in this winter and build an all triode amp. When it comes to audio, your excitement is truly infectious. Thanks.
Hi Bob,
I think I may have mentioned it before but I have several brand new Svetlana SV572-10 (not to be confused with the 572B used in RF amplifiers) that I really should build an amplifier with. The low mu triodes just seem to have a sound that I like and these have a gain of 10 and are absolutely beautiful. I don't have a pair of 10K primary output transformers so I will have to look around. 8K would probably work OK with a little lower voltage. I have always been particular about technical performance, and that remains, but as much as I try to stay from "how it sounds" in my YT videos, I have to admit I do have my preferences. Maybe a spectral comparison of a high-end audio amplifier compared to the lower power triodes amplifiers will shed some light on "how it sounds" although THD on both are exceptionally good. I don't know what it is.
Nice job re-engineering the supply for triode operation. I have my Dynaco ST70 set up for triode mode and find the same change to smoother dynamics with greater imaging clarity. Also saves the power transformer. Keep em glowing!
Probably one of your best builds now improved. Thanks for posting this. I'm working on building a clone of a mc240. These videos really help with the technical problems. Like eliminating hum. I'm about ready to wind the output transformers. I'm using tape wound toroids for the cores. I'd like to post the results of the finished output transformers when it's working for others to learn from my experience.
Thank you
Rory
Rory,
Indeed, it is good to post results, even those that aren't perfect as so many things can and do go wrong during equipment building and repair. I made several mistakes in this one especially in calculating the cathode resistor. Not sure exactly why but it happens when we are roaring along.
Building the MC240 will surely be a fun challenge. I wish you the best of luck and building practices. Yes, hum from an improperly built amplifier can be difficult to impossible to get rid of. I have found over and over, to keep all the PS components in one corner of the chassis and ground all power supply components, including PS electrolytic capacitors to that one point in the power supply corner of the chassis. That will go a long way in preventing hum. This doesn't include small cathode bypass capacitors on the driver stages nor does it include the input RCA connections. These types of grounds can be made right at the chassis and tube sockets. Study how McIntosh does it. It is often simpler than we think.
Please do post your results. You never know who it will help even with just some small detail maybe years later.
David
"Pure Triode Class A" Preach on Brother David.
What a beautiful amp!… 💗
Have you compared the KT66 with the 6B4G on the spectrum analyser? I think you'd reduce the THD by using two cathode resistors or a balance pot after the phase splitter in fixed bias so you can adjust each valve, to get the two sides of the PP OP balanced. It makes a big difference in my experience, and I'm sure I saw a video of yours where you mention this.
I know you like to keep things traditional, but using transistor's, for instance as a CCS for the phase splitter is always worth experimentation. Anything you can do to reduce THD and maintain good balance in the stages before the OP stage is worth a try.
Grand video as always, thanks.
I think you are right about separating the channels even if I don't use a fixed bias supply. And I certainly agree with you on separate bias adjustments for each tube for best performance. I built this amplifier two years ago primarily to see of the Acrosound transformers were as good as everyone thought. I wasn't disappointed and since posting this video I have made some measurements and both channels (within what the weaker channel can do) have a frequency response that is beautiful down to 20 Hz and below. At 20 KHz the THD goes up a bit but not bad. The KT66's did the same thing but with more power.
I do need to post some real measurements and a spectral display of this one and maybe compare it to a high end amp like the McIntosh MC40 I have. Maybe I can do those measurements tonight and compare the two.
Great video, i have some 6b4g valves, going to try to use them 😀
Your amps look great. Wish I could wire that cleanly!
At first glance at your shared cathode arrangement I immediately thought 'crosstalk', but with the bypass capacitor I can see why you don't get it, except at VLF where it is presenting an impedance. And maybe it's a good thing to common the VLF stuff anyway :-| Excellent video. It would be interesting to see a spectral analysis of the error signal. I frequently check that, to the point where I am almost ready to have it hooked up in my rig permanently.
Hey, if you dont want those 2A3s send them my way... I'm building a true loften white amp right now. loften discovered that the 2.5v filliment reduced hum drastically, especially with a directly coupled driver to power tube setup.
RCA had this thing called Dark filaments in the 40's. It was supposed to increase emissions. Might explain why the "good" tubes have darker filaments? Funny after all these years, we still don't know some things about tube amplifiers.
Isn't that the truth... so much misinformation and so hard to find answers to many questions after so many years of tube amplifiers. The most amazing (and unfortunate in come cases) is the flashing-filaments that Amperex and Mullard and Telefunken built into their 12AX7's and others that too many people think they are defective. I have a number of these tubes and use a set of them daily in a McIntosh C20 amplifier. They have been "flashing" for over 50 years and many of these old tubes have both triodes test almost exactly the same (match pair) even after all these years.
I didn't know that about Dark Filaments but those are the ones that perform best...
Thanks for the information.
Great stuff, love your videos and shop set up. I'm just starting to refurbish a pair of eico hf22 amps and you have peaked my curiosity to try the 6b4g instead of the kt66's.
From what you can see would that circuit be ok with the 6b4g other than lowering B+ a bit, changing cathode resistor and removing feedback? I believe it's a mullard 5-20 circuit.
Thanks
I think any of the Williamson variations will work well with the 6B4G. I did it for a while as you might have seen in this amplifier but I did not have a way of lowering plate voltage so I was running mine too hot and is why I went back to the KT66. The 6B4G, I believe, is good to 15 watts but it would probably be nice to the tubes to keep the plate voltage a little lower and maybe run them at a max of 12 watts per channel or so. Just my thoughts on it.
Thanks
Looks good David.
Excellent video. Would love to see the final schematic. Do you sell your designs?
I do have a LTSpice schematic and model of this amplifier that simulates beautifully but I don't know how to share the file with out emailing directly to everyone at their email address. You also have to have LTSpice installed to open the file. You don't actually have to be technically familiar with the Spice program but it is necessary to have it to open. LTSpice is a magnificent program and is 100% free from a high end company.
www.analog.com/en/design-center/design-tools-and-calculators/ltspice-simulator.html
I agree, prefer 6b4g to 2a3
it would great to build great video
Triodes do sound great! Now want to go over the top in sound fidelity? Let's eliminate the Output transformers! Say what? Yes, OTL! With pure triode goodness directly powering the speaker output. I've got some Dark Voice 6080 OTL Headphone amplifiers arriving today. 6SN7 to 6080 SE to drive 300 or 600 ohm Quality Headphones. I did bother you about a similar Headphone amplifier I made about a year ago. 6DJ8 replaced the 6SN7 and it sounds Outstanding to me. The 6080 is a dual triode and you could go PP to get more output and lower impedance required for more traditional speakers. Maybe even Parallel Push Pull with multiple 6080's to directly drive 8 or 16 ohm speakers. I know you already have some of those tubes and lots of nice Output transformers. But isn't it time to test out the OTL designs and get pure Triode driven sonic bliss? Best regards, Paul! ;^)
Ahh Darkvoice and headphones, now you got me to pay attention. ;)
I own the 'LeFigaro' which is the new brandname for Darkvoice. And indeed high impedance cans, AKG 601's + Beyer dynamic Studio.
It sound really beautiful, detailed and musical after some tube rolling.
The short version on that is as follows: Stay way clear of Russian - Soviet era Tronal .
I liked General electric's 6AS7GA best. For other tube position I had a hit with Sylvania 6SN7GTB.
But the absolute favourite is 'Realistic International' 6SN7GTB though made in Japan it's not supposed to be one audiophile tube - actually a Radio shack tube.
The clarity and details are now amazing.
@@andersforsgren3806 I want to see him test the OTL headphone amps on his test gear. Comparing technical test data with human ear perception. We know how good the DarkVoice or "LeFigaro" sounds with high-quality, high impedance headphones 300-600 ohm. It would be nice to see how his test gear evaluates what we hear as amazing clarity and details.
@@PaulinTaegu Well his videos are mostly about digging into the gear and make them work right.
Listening and evaluation bring in the subjective factor, with opinions and discussions. And before you know it the crew with cables of ultra exclusive materials and prices, the crystal healers who demand their favourite mineral should be places on top of each speaker / amplifier. Those people usually join forces on audiophile forums discussions - it's about cultist behaviour, nothing else.
Example here: www.biophotone-audio.com/Unsere-Produkte/ 'Biotechnology Sound Chips' or harmonize the sound in your room with a crystal.
With that kind of weirdos around, I fully understand that ElPaso and his colleagues Mr Carlson etc deliberately avoid listening comparisions - else their poor channel risk singing in another flamewar. So yes, nice idea, but I understand why he don't. :)
Hi David, Really interesting video.... When you had the KT88's in there was it running class A as well? I guess it was. You have me wondering why the 6B4G tubes would sound better.. Or why it is you prefer the sound of the triode over the Beam power tubes... I would think it would be something you can measure or see with all the equipment you have.. Then again maybe not... Well I still want to see a follow up video on this.. Neat stuff... Be Well, Tom
Hi Tom.
I have generally made it a point to stay as far away from how amplifiers "sound" in my RUclips videos as I can. How something "sounds" is so subjective it can never be proven or dis-proven but in this case I just like the sound of it so much better with the 6B4G's. It seems the detail of each instrument with accompanying sounds are just "there" whereas I simply do not hear them or maybe just do not notice them in other amplifiers. I have done experiments with variable amounts of NFB on a number of amplifiers and the amplifiers with little to no NFB simply sound more alive. Unfortunately, the difference in power output may be a factor of 10 lower with no NFB. Back in-the-day (1930's and 1940's) when amplifiers were being developed, the idea was to make a smaller and more power amplifier. Nowadays, we are able to play with huge amplifiers that deliver only a few watts of power per channel and weigh in at 100 pounds so we have definitely gone back in time... :-)
It is just something that strikes me in the sound of these triode amplifiers that I like. I have not tested this one from 20 to 20KHz for FR and THD. I am going to bet it is not near as perfect as the beam power tube amps. For some time, I have come to believe that we actually like the artifacts that out vacuum tube amplifiers add to the music - otherwise called distortion - but somehow we like it. Hard to believe and understand sometimes.
Thanks for all the great videos do you know where a guy can get a schematic of a Cayin 100 I really like a lot about it would love to see you do a push pull hi fi and I for one love the long videos thanks again for all your work
I'd like to try single ended 6b4g amplifier
Have you tried connecting the KT66's as triodes?
If the filaments are 6V then what a LM7806 or if 12V then what about a LM7812 regulators that might get rid of the mains hum.
You got any circuit diagrams?
Wanted to ask your experience regarding using tube shields in the preamp section. Any input would be appreciated
Tube shields are usually not necessary although I have used them. If you can place your fingers around the highest gain tube, i.e. 12AX7, etc, and not hear any hum increase out of the speakers or detect any additional hum on your instruments, I figure I the shield is not needed. However, recently in the case of the headphone amplifier I built for my grandson, I did use tube shields for the purpose of safety. If the circuit is RF it might be a good idea to prevent anomalies either from radiation from the tube or stray RF pickup from other circuits by the tube. An example of that might be the Collins 32S-1 transmitter where there is a shield around the 6CL6 (which makes it run really hot...) because it is in the same compartment as the two 6146's and the plate tank circuit. In this case I think it would be used to prevent the 6CL6 (the driver tube) from picking up RF from the final amplifier stage. If a oscillator has a shield around it, then it is probably a good idea to keep it there as I would expect the shield to provide some capacitance to ground that might be necessary for oscillator accuracy. Just my thoughts on it.
@@ElPasoTubeAmps Thank You, Its a quicksilver mid mono amp and they didn't come with shields, but my Sherwood s 8000 had them so I wondered why the difference. I'll try your fingers technique on the first input 12AX7 to see if I can hear anything
Do you have a schematic for this amp?
This is a great source for both the KT66 and 6B4G schematics: www.tubebooks.org/file_downloads/acrosound.pdf
@@ElPasoTubeAmps Thank you, Great channel!
What brand are those orange capacitors?
They are, Sprague 715P type.
@@ElPasoTubeAmps Nice thanks! Probably a nice vintage sound!
Why not just dc heat them??
Your suggestion of DC filaments is a good and common approach but I have never found it necessary. If the filaments are run on AC it is necessary to provide an adjustment between the filament wires to ground to minimize hum and also a cathode DC return path to ground. A 50 ohm or even a 100 ohm pot placed between the filament connections with the wiper to ground will reduce the filament hum to zero. You could still have a little power supply hum but it will eliminate filament hum. A lot simpler than converting to DC. Some of the best hum-balance pots I have ever had were hum-balance pots that were used in the older Fender guitar amplifiers, 50 or 100 ohms. I bought several of them from a fellow on Ebay.
wouldn't it be better to have two power Transformers for each channel it's always better this way then just beeping it up, put them on a separate chassis with a multiple din plugs. The best plugs are the ones that the army use for tanks rhodium coated. might be able to buy from army surplus in Germany Britain or America Canada. or Russia army Navy airforce, surplus. talking about Air Force you might be able to buy them from Old dismantled commercial or military planes that are scrapped.
or what about using a switching power supply heavily filtered with noise compression bank of capacitors running like a bank of batteries.
Hi John,
My big RF amplifiers have power supplies built into the lower part of the cabinet so I can experiment without rebuilding the PS each time but I have not taken the time to build a separate PS for my audio experimenting. It would be a good idea and as you mentioned, a separate PS for each channel would be the very best, kind of like having a pair of McIntosh MC75's. I have always wanted a pair of those beauties and several years ago I found some for sale and agreed to his price, I think about $1500 for the pair (that would be a good price for a pair in nice shape) and then he backed out. Oh well - that means I just have to build more of my own... :-) Also, one of the reasons I continue to build the PS on the chassis of my audio amps is I like to use them to listen to music and two chassis just takes up so much more room. Oh well...
I do have a box full of matching military plugs as you described but I have never used any of them. There are definitely good stuff and, I think, even water proof around the seal. Maybe I will build up a versatile PS for audio amps and use those plugs.
I am very concerned with high frequency noise in SMPT power supplies. No doubt the are so much smaller and efficient but that high frequency switching could be a real problem. The old massive linear supplies may still be the best for audio work.
You say your ears like the sound 1: check the second harmonic see if that has become dominant? , at the same time the rest have lowered?. 2: just for a temporary experiment to lower distortion, use a transformer three times larger than your math says what the acceptable rating is. Retake your distortion reading.