Vacuum Tube Amplifier James Audio Output Transformer & Determining Correct NFB For Williamson Design
ฝัง
- เผยแพร่เมื่อ 16 ม.ค. 2016
- A 20 watt audio (HiFi) amplifier built to a Acrosound schematic using 6SL7 and 6SN7 for voltage amplifier driving push-pull 5881 and a 8K ohm James Audio output transformer. Exercise in determining correct values of R and C for the global NFB circuit.
- บันเทิง
Great video. I was actually reading a copy of "Audio Cyclopedia Section 12 - Audio Amplifiers" while I watched this. Then you showed yours, how perfect! Don't own my copy, boo hoo, got it through inter-library loan. I agree, it is an excellent reference.
I am impressed with the power output!
+noahtheviking I agree - all the way up to the 30 and 40+ watt level. I read good things about these transformers in the DIY forums but then there is always someone ready to trash them (and everything else).
true for sure
those trafos looks actually quite impressive, and I love the Freq plot too, nice roll off at 30hz that must tightened the low end while the topend gets open and airy at 15khz 👍
great video.I'm quite envious of the equipment.My variable capacitance is from an old transmitter and once set have to use my Heathkit cap tester to tell me the value.Yours is a Gen Rad piece I think? I use the potentiometer method also and work over a lot of old theater gear.Most of the RCA tube theater amps,the true theater amps like the MI-9377's have a substantial amount of roll off dialed in from the factory to minimize hiss which was a no no in theaters.I have corrected many by raising the 27K feedback resistor and adding some parallel caps to flatten the response.they are capable of an amazing sound!Most of what
people are calling theater amps(a selling point mostly seen on the auction sites)were just for commercial sound use.I personally have removed a rack full of the MI-12182 dual 6146 amps from Macy's during a renovation back in the 70's.still have a pair.But I won't fault this equipment either as I have also seen those brutish 12188 amps with the four 807 tubes run flat out to over 30Khz.In fact RCA had issued bulletins on them warning of HF oscillation causing over dissipation of output tubes by reducing HF response with capacitance in parallel with the feedback resistance.Good Job! 73's. N2VUW
Very interesting video. Is there any way you could make a video showing how the instruments are connected?
Can you show us how you punch out holes in the sheet metal?
+The Current Source I will show the tools in some future video. But they are Greenlee chassis punches. They come in all sizes and you can find them on the Internet and Ebay. Square holes are drilled at the coners with a drill bit and cut with a Dremel tool.
Are these UL transformers. Did i miss that important info. Also 8k is slightly higher than by the book so distortion will be lower at lower power.
Hi, is that a James JS6223-HS OPT?
Thanks
To set the value of the feedback capacitor you MUST use a SQUARE WAVE not a sine wave. Feed in a 10 kHz square wave and adjust your cap box for BEST SQUARE WAVE. Probably best to do it at 1 Watt not 40W.
Thank you. You are the first person to suggest how to determine the value of the
FB capacitor. I will try it.
Well, I am glad to help..Now where did you get your James Transformers?
+ElPaso TubeAmps
Hi there e David,
Thanks for evaluating the James Audio transformers. It does shine some light on the whole thing. As much as I don't trust eBay sellers, I do trust numbers and I trust what I just saw. I almost bought a pair of them on eBay a while ago. They seemed affordable and that's what triggered the red light for me but I guess this video proves my suspicions wrong huh....
Do you know where they are manufactured?
They perform beyond well and also have a nice look to them so it wouldn't disturb the vintage look on my amps.
I really appreciated the feedback part of the video. I never liked too much feedback anyway but seeing how it works in real time on the scope really confirmed a lot of the bad assumptions I made about feedback years ago. I guess I just have to quote my father on this one, he always says "...if you wanna build a nice oscillator, just build an amp with feedback".
Now I want to experiment with little to no feedback at all on my next push-pull amp. I guess I'm gonna use that acrosound schematic driving a pair of 5881's, 6098's or even some 6L6G's I have around here. Feedback seems to be mandatory on push-pull amps, but you rarely see it on single ended amps. That's maybe one of the reasons why single ended amplifiers sound so good.
Initially, I thought I could use the same schematic but with a pair of 6080's instead of the 6b4's, than it would be triode, but the plate current is stupidly high on the 6080's and I'd have to wind a very weird output transformer for it.... so I guess I'll just stick to the pentodes for now.
In the future I'd really like to experiment with triode push-pull. You just mentioned that you were charmed by the 6b4 amp and I read about a lot of people who really fell for it. Maybe it's going to persuade me too.
BTW, I loved that precision capacitor, boy, what a gorgeous piece of instrument, and for guys who build tube amps that's just so useful. Now you couple that to a nice resistor decade or some precision pots and you've got yourself one nice developing prop.
Thanks for another cool video,
Denis
+Denis Gouvea de Azevedo
Hi Denis,
I have had very good experiences with Ebay - except for two that I can think of in over ten years buying off Ebay. I did not get these James transformers from Ebay. They are cheaper ordering them directly from the company. Just Google James Audio or try this link: tctubes.com/james-transformers.aspx
They make three for SE and three for PP. Check them out closely and make sure you get the Z you want. I do not know where they are manufactured. I just cannot see paying $300 for an old Acrosound transformer.
The FB network has always been a challenge so I thought I would attempt to document my efforts hoping it will help others.
As for SE amps, I have never had one but I have built a large PP amplifier (80 w) with a variable NFB (basically from zero to -20 dB) and with no feedback and just a couple of watts the amp sounded really loud and seems to have a huge "sound stage" but I am not sure I really like it. All my amplifiers run with normal NFB.
I have seen schematics using the 6080's but I believe the Z is very low so it might be a challenge to find an OPT. I believe there is a design using six or eight of them that is transformer-less.
I made some more changes in the amplifier after the video (putting the 6B4G's back in) and I run the 6B4's with adjustable fixed bias and got rid of the cathode bias resistor and capacitor. It is interesting that the 6B4G and 6L6 have the same pin out for plate, grid, and filament. Of course you have to ground the filament instead of the cathode on pin 8 of the 6L6 but it allows me to use my "bias probe, break-out box" to measure the plate current in the 6B4's. I just like the sound (and look) of these small power triodes. I have compared it to McIntosh amps and maybe I like the artifacts in a less than perfect tube amp ?? There is just something in the 0.01% THD sound of a SS amp that is not satisfying. I can't explain it.
Yes, indeed, that capacitor is the "cat's meow" and it sure served me well last night. I think I mentioned in the video that it almost seems that these amps with their complicated adjustments would be much harder to mass produce and the 6L6/5881/EL34/etc put out so much more power at less THD so ... the winner is obvious.
I almost didn't make the video public as there is so much fumbling but I figure that is the way real life works.
David
+Denis Gouvea de Azevedo Actually, after a little more investigation I found a site that shows all the transformers that James Audio makes. The site I gave you is just where I bought mine. They are made in Taiwan.
ElPaso TubeAmps
Hi David,
The transformers perform amazingly well and they seem to be pretty solid too. Hard to compare with the late 70’s early 80’s cheap stuff made in Taiwan huh….
I’ve been trying to use those 6080's for ages but never got anywhere. I even thought about building an output transformer less but the topology reminds so much of SS amps and with such a low Z, I would just wind up with a tube amp that sounds just like a SS amp.
Now, there’s nothing wrong with liking the 6b4 amp better than the McIntosh in my book. When you come think of it, the McIntosh tube amps are all about feedback. You get a lot of both local and global feedback, that’s why they score so high in the specs. At the same time, that’s maybe why they sound a lot like SS amps too.
I have built a lot of SE amps over the years and I find them to have that "mojo". The kind of "mojo" you get when you run a PP amp with no feedback and low power. At low power, your PP Class AB amp is mostly behaving in Class A and you do get to hear the teeny tinny details hidden between the big notes. That’s why I like SE so much. Even if PP is a “superior” design in terms of efficiency and tube life…bla bla… I just find that the common mode rejection on the PP often kills that “mojo". The harmonic content is completely different and our brains can relate to that.
It's been so long since I only build SE amps that I'm currently forcing myself to build a PP amp for the sake of comparison.
Just as important as the capacitor, if not more-so, was the bias probe/break-out box. You could see the oscillation so much better using the meter than using the scope at low frequencies. Which, BTW, makes me think about getting a VTVM.
The video was also a great example of how shunt resistance and capacitance work within an amplifier. The French have a name for negative feedback and it's “contre-réactance” like in counteracting. I’ve lived in France for 11 years and only now I noticed that that’s one appropriate name for it.
Thanks for making the video public, it shows how hard it is to get the right amount of feedback. If there’s ever such thing as the right amount of feedback.
Denis
ElPaso TubeAmps
I guess I'm going to get a pair of transformers via the same site you used. If they ship to Brazil. The prices are really interesting and that's for the pair. They're worth trying.
Thanks
You're running at 40 watts from a what was it? 20 or 30W transformer. How much margin do you think is in that rating? Or alternatively how long do you reckon it's safe to overdrive it like that?
+QuadMaestro I can't say how long it would be able to withstand 40 watts continuous but considering its physical size I would think it could take at least 30 watts "forever". The FR, THD and other HiFi factors might not be as good as it is at 20 watts but I doubt it would go up in flames.
Any further developments in SE triode amps with smaller transmitter tubes? Has the El Paso storefront closed? I want to commission a project. Jim in Westport, MA.
+Jim Matthews I recently purchased some SV572-160 triodes that are supposed to be the equivalent of the 811A but with the plate connection on a base pin rather than the top of the tube. Also plate dissipation is much greater than the 811A. I originally purchased them for SE amplifiers but I just can't seem to move away from push-pull amps. When I decide to go back to a serious SE amp, I will probably start work on the 833A amp again. I am charmed by the 15 watt 6B4G amplifiers. I know the 2A3 is a favorite but that 2.5 volt filament is just a lot of trouble in my opinion. Another problem is I have no SE OPT for any tube. I am still thinking about it :-) What are your thoughts on a smaller transmitter tube that would be a good candidate?
+ElPaso TubeAmps The Russian GM70 seems cheap and cheerful. Would you consider building a parallel SE with the 811 quad you have, or are they better in a Williamson amp?
Great vids. I get interesting ads associated. Jim
+Jim Matthews An SE amp with a quad of 811A's... wow - I like that. Maybe a plate load Z of 2500 ohms or so ? I have to get some nice SE transformers first. I appreciate the idea.
Would you be willing to field a telephone call? Jim
the problems with push pull and also feedback is they both cause phase distortion witch destroys sound stage. Notice no one ever measures or talks about this. Our brains determine the source of sound by phase(not volume) differences between what our ears are picking up.
It is interesting that you use the word "sound stage". I described that word in an amplifier I built some years ago using KT90's with a variable NFB loop. With no NFB the thing sounded wonderfully "alive" and very sensitive to high volume and overdrive and distortion.... I understand what you are saying and as I cranked in NFB it got compressed sounding. However... and this is a very big, However... with no NFB the available pure and clean "alive" power is around 8 watts with this large amp. With NFB power rises to 80 watts and THD is low. You can listen to classical chamber music and violins at this level of 8 watts and just go off into heaven with its "openness" and "sound stage" (have you noticed that is the only type of music you hear on the Internet using these massive and absurd sized amplifiers?) but you sure as heck can't listen to Guns and Roses or Def Leppard at 8 watts.
You make a good point. use the correct design for the job at hand. I have a beautiful set of acurosound ultra linear 2 amps i rebuilt and used kt77s easy 60 watts of very clear clean sound but my single ended 45 amp with a watt and a half per channel no feedback has great sound stage.
Any ways of measuring THD without breaking the bank?
+TheLawnWanderer I started using a program for the PC called SpectraLab. I ran it on a 286 PC and it did a good job at 1 KHz. I did not have any way to evaluate it at 20 and 20 KHz. Performance depended on the PCI sound cards, etc. Google it and it is free for 30 days. You may like it for a while. I think the best buy for the money is the HP 8903A or B model. Don't get stuck having to have the B model. My A model is perfectly sufficient. The Pete Millett software for the PC is free.
+ElPaso TubeAmps Your videos are the most informative in everyway David.
Get yourself a Tektronix SG505 and AA501 and a TM50x mainframe to put it in. You'll be well under $1000 and you'll be in the wonderful world of the Tek TM500 system, which has all kinds of other lab plugins.
+ElPaso TubeAmps The best thing right would be those keithlley 2015s that are selling for 350 to 400 dollars and they were 4300 new.
+TheLawnWanderer www.partsconnexion.com/product26900.html
Check this out - per Mike' post above.