Vintage Vacuum Tube Amplifiers & Modifications

This is my experience with a variety of vintage vacuum tube power amplifiers and an attempt to address approaches to modern day vacuum tube design and performance. My experience has led to an understanding of parameters that we cannot change and a number of simple modifications that we can make that provide significant improvements in performance. I hope this video encourages and helps empower the audio enthusiast to realize attainable mods and improvements to their new and vintage vacuum tube amplifiers and promotes the advancements of audio technology both in the realm of HiFi reproduction and the creation of music through guitar amplification.

[@towerman75]

I am always amazed by the knowledge that some individuals hold and at how diversified they can be. I would have not known that until the last of the video. A very comprehensive course on the makes of audio amplifiers, and then to see that you have co-authored a book on dental toothbrushes. Being a retired Electrical Eng., from NASA myself, I always enjoyed my job, and at the same time, I spent 72 years as an Amateur Radio Operator, so I guess it just goes to show. Congratulations, well done.

[@towerman75]

I knew it David, with the years of experience that you have, you must be at least 150 years old. HiHi I am amazed at the knowledge of some individuals that are associated with their age, one in particular (other than yourself), is Mr. Carlson of Carlson's Lab. I can't make comparisons, but having seen Mr. Carlson in his video's, he looks to be in his late 40's, but with his knowledge (of both modern day electronics and tube type technology), just like you, he should by all rights be near th 2 century mark. Anyway, I enjoy all your videos, and his as well, keep up the good work, and the best of 73's. Jim WA5WRE

[@bobh.3705]

A couple years ago I evaluated phase inverter circuits through spice simulation. I chose the split load configuration for my use similar to the Williamson design shown in your video. It did not come without challenges. If the plate and cathode resistors are to be identical, (1) the loading on the output must be very low (high impedance), (2) to reject power supply ripple the plate and cathode resistors on the first stage must be properly sized, and finally (3) the plate voltage of the input stage (which sets the grid voltage on the second stage) must provide symmetrical operation of the split load stage. Your examples encompassed some or all of these problems. The Williamson circuit worked properly. By the way, the distorted waveform was probably due to change in bias of the input stage due to the feedback. The output tubes probably ignored this distortion since the output tube receiving this was not in conduction (for the most part). I really enjoy your videos.

[@bobh.3705]

I may have misspoke in that the output tube was not in conduction (see above). I've also seen this distorted waveform when the feedback is trying to compensate for something that has saturated (hit a rail, transformer exceeded Bmax, etc.) If you remove the feedback loop, does this lopsided distortion go away?

[@kinura26]

I realize this video is 6 years old but found it very helpful. I did the AC balance modification to my Dynaco st70. The mod works well I can null out the error in the phase splitter. One thing I did notice that made me question my scope or probes was if I connect to the grids of the output tubes then null out the error all is good but if I switch the Leeds around the error comes back. I did test my probes and they seem calibrated not sure what the issue is but I fudged the adjustment to a happy medium. Again thanks for this video

[@ElPasoTubeAmps]

Hmmm... that is interesting with the probes being different. I think I would put them together on the same point and switch between channels on the scope and see if there is a difference there. Just a thought. Yes, indeed, the AC balance does work and helps bring down THD to its lowest level. Thanks for your comments.

[@faultlessguitarsandamps1116]

Great video . I look forward to the PC based test gear episode .

[@mfr58]

Really useful ideas. Given me inspiration to tweak the performance of a couple of my amps. I also like the rule of thumb calculation for x'fmr impedance. Thanks. Matt

[@emorris272]

Great video. I've got a pair of Mark III's. Their driver boards have been upgraded for 2 tubes. I bought them that way. I didn't know AC Balance could make such a difference. Thanks so mu;ch for the video!

[@materialsguy2002]

Edward Morris i

[@olipito]

Nice video! And congratulations for the publication!

[@billmoran3812]

Congratulations on your paper! We need to keep the mind working or it will just fade. I always enjoy your experiments with amplifiers. It it through such exercises that we learn the real details of circuit design. The two channel scope method seems a great way to balance the amplifier for minimum distortion.

[@MichaelBeeny]

In the UK one of the most famous amplifier circuits for the home constructor was made by Mullard namely the 5-10 with EL84s and the 5-20 with EL34s .Anyone in home build would have one or other of these amplifiers. I have built dozens of both these amplifiers. It looks like these circuits are not at all well known in the USA for some reason. The QUAD 11 amplifier KT66 also does not feature much at all but one of the best I have ever measured for distortion and extended frequency response and sounds really clean and detailed. I guess it was probably availability in the 60/70s.

[@tashpaug]

A great video! Also congrats on the published article. I really enjoy all of your videos. Very informative and always good.

[@paultopace1234]

Thanks mate i was just thinking today I have to understand this impedance / transformer stuff. Its very easy, thanks.

[@MichaelLloyd]

Congratulations on being published David! Personally I like the rack full of stuff but being a gadget minded person I'm interested in the PC based approach as well. I guess I'm doomed :)

[@ElPasoTubeAmps]

I don't think we are doomed yet. I downloaded two oscilloscope programs last night and their display is very bad, not even usable. These are all freeware so, maybe there is better but I'm not going to invest $$ in this experiment. I think our rack full of stuff is still safe. :-) but I will continue to look into it for a while longer.

[@PeterMilanovski]

Every man likes a good rack... Mr Carlsons lab is definitely a boob man LoL..

[@neilzspam]

Great video thanks ElPaso. I always enjoy your vids, but this one seemed to really speak to me. And please make that PC based test setup video.

[@Finom1]

Looking forward to your three PC based test equipment software videos. I would like to know what manufactures of the software & hardware you like the best. I would like to build a guitar amp. Your videos are very educational.

[@shyammohabir8283]

Nice video ElPaso.. So for the 6L6GC in a 60 Watts Push Pull configuration .. What do you normally see in the market for the Plate/Grid voltage and Cathode voltage/Current?

[@KarlAdamsAudio]

Interesting point re: resistor matching in the phase splitter, the same issue comes up with regard to the Schoeps condenser microphone head-amp circuit, where an apparently 'symmetrical' circuit actually has a degree of impedance imbalance at its outputs. The key issue is the difference in output impedance between the two sides of the phase splitter (in this case a JFET, but the principle remains the same) - the cathode (source) side always has lower impedance because it's acting as a voltage follower.

[@ElPasoTubeAmps]

Isn't it interesting how we just "buy it" and follow along because it "makes sense". I like your point about the lower Z cathode side. Stepping out of line and pointing out an error in such an established principle has its risks unless it can be proven with simple empirical evidence. Thanks for you thoughts and postings.

[@tjasont1]

It would be very nice for you to go through this pc stetup I really need thd and spectral analysis of the amps I tinker on so it would be a tremendous help to me and am sure many many others. Am a lowly hard working average Joe and don't have the money really to spend thousands in test equipment. The attenuators u use the software and setup would be quite valuable info for me am sure I could figure it but it would be nice to learn from your experience so I can avoid any costly mistakes I love your videos and appreciate greatly your insights. A man is only as good as his tools allow him to be lol thank you

[@kubockferre5532]

hi,the calculation to find the output transformer works well for push-pull ab-1 but i tried the same calculation for single ended class a with 6l6 tube and its dont works .Can you tell me how to calculate for class a single end thanks

[@kinura26]

I had a thought since I don’t have a THD meter could you disconnect the feedback loop from the output transformer then measure for the cleanest signal at the speaker terminals then reconnect the feedback loop ? Just a thought

[@ElPasoTubeAmps]

Many amplifiers have about 16 dB negative feedback. For example, a 20 watt output amplifier with 16 dB global feedback would be measured at 0.5 watts or at least no more that looks really good on an oscilloscope then reconnect the NFB. Or, keeping the output as seen on a scope at a very clean linear level (with NFB connected), adjust for the best display raising it to just starting to clip and trimming at that point.

[@goaway2174]

Just came across your video and appreciate you taking the time to educate us on these classic designs, it was very informative! Just wondering when biasing the current on the Dynamo MKIII tubes, if you were really concerned, could you just put the meter across the two bias points and adjust the DC balance until it read 0 vs. ground and each bias point individually? Thanks!

[@ElPasoTubeAmps]

Yes, if you install the DC balance pot (I took that idea from the Poseidon board design) onto the PC board as I have it in the video, you can adjust the voltage between the two bias points to zero, naturally, making sure your cathode resistors are very well matched and then with the original bias adjustment that Dynaco designed, adjust the idling cathode current to 70 mA thru each tube.

[@dannyalvens5418]

Nice video,but the mods only works if you have the equipment. If you don't have the gear,Your stuck.... I figured you were going to talk about mods as in Coupling capacitors,HV capacitors, Feed back loop,Cathode bypass caps....How about doing a video on using or sizing Filter chokes,Regulated voltage supplies.How about a Voltage drop calculator for resistors.Just a thought..

[@copperaudio9664]

I know this a 5 year old post - a really great post IMO. But I have to ask if anyone knows where to get those high quality BNC to RCA connectors @25:40 mark? All I see on the web are poorly made and ill fitting. Thanks.

[@hybriddevice]

Really enjoyed your video. Your work paralleled my own thoughts, discussions and findings. Many thanks for putting it up. Do you have a copy of the original Williamson article/circuit as printed in the British Wireless World magazine? It differs a little from a lot of the Americanised Williamson circuits. You might find it interesting. Driver tubes are all single triode Osram L63 types, basically half a 6SN7. I would be more than happy to send you a copy of the entire thing. Best wishes, Haden.

[@materialsguy2002]

Well, I really enjoyed this video, thanks. Regarding PC based audio analysis gear, consider also the Analog Discovery ( I and II) along with the PC software from Digilent and the audio analyzer software from thestuffmade. Blueglow and Tomtektest have been doing some excellent work with this relatively inexpensive setup. It is a complete audio analysis system as you have described.

[@ElPasoTubeAmps]

Thank you. That is exactly the information I am looking for.

[@ElPasoTubeAmps]

Shaun, I have been watching Tomtektest videos and he does a good job of demonstrating the Analog Discovery 2 setup. I am wondering if I even need to make a video on this type of PC setup with his videos posted. I think I will make reference to his videos for now. I would have to buy the Analog Discovery equipment and I think my viewers are interested in a setup with software that could be obtained as freeware and using the existing sound card of the PC. I will have to think about this before proceeding with this video project. Thanks again for your help.

[@materialsguy2002]

ElPaso TubeAmps: I agree Tomtektest did a great job on the AD videos. And Blueglow is setting up a similar, but more integrated system using the AD II as well. Alternative solutions to a problem are always good, so I am looking forward to see what you work out.

[@garytrautman5404]

Love your videos. Ya, they help. One of my favorite instruments on my bench is a Singer Panoramic Real-Time Spectrum Analyzer (vacuum tube based work of art). It has a dedicated SOLA "Tele Volt" voltage regulator that feeds the power supply. These ferroresonant voltage regulators are an interesting study. Subject for another video?

[@ElPasoTubeAmps]

In a few videos back there was a discussion of how amplifiers often times perform different at venues and the performance was discussed in the context of AC power. Low voltage from long extension cords, switch mode power supply noise , etc was discussed somewhat and I think the conclusion was that the Sola type ferroresonate, constant voltage transformers might be part of a solution. They are available on Ebay - some very expensive - and heavy. I doubt anyone will seriously start carrying around these devices to plug their equipment into but it was an interesting discussion. Thanks for your comments.

[@garytrautman5404]

The Tele-Volt series was a portable version.

[@nickk6109]

Interesting - you could signal generate (even multiplex) and analyse using FFT (discrete), the key is having a good enough DAC/ADC with enough bits/sec bandwidth, so perhaps a programmable chip (ARM?) to drive the data capture as USB2.0 is too slow - it would need USB3 to simply take the stream of ADC. I've used GPUs to perform realtime 2D FFT for image alignment at 17fps. It's possible.. Interesting point about the dynamic load vs static - I wonder if you could make a programmable dynamic load too using current chopping to different resistances.

[@bobh.3705]

David, I would be interested in a video showing the THD of the Focusrite (or other sound cards) output as used as a signal generator. Sound files could be generated by spice programs or other various ways and then shared between users.

[@ElPasoTubeAmps]

It has been some time since I have use Spice or PSpice but I always enjoyed it even the DOS version. I have thought of some short videos using PSpice but would have to work with it for a while to get efficient with it again. As for just testing the Focusrite and other USB sound cards, that's a great idea and it would answer the question of how well they do at 20 KHz and above. Thanks.

[@bobh.3705]

I've been using the free version of LTspice for about 18 years now. It can consume a lot of time and I miss the smell of solder. I'm always comparing your results to the simulation. I love the way you measure various pieces of real hardware. Remember what Lord Kelvin said about measuring?

[@ElPasoTubeAmps]

Yes, measure once in mm and cut twice in inches then convert to the speed of light. That't the way I do it when I work with wood. I tried to make a speaker box last night and all I can make of wood is fire wood. :-) Seriously though, and management converted it to, " if you can't measure it, you can't manage it", which I guess is true too. I need to get back into PSpice, I had some projects while working for Lockheed where I used it and I found it a lot of fun. One project was modeling a non-linear load on a half-wave rectified system and then full wave rectification, driving azimuth and elevation motors for illuminating radar antennas. I will never forget that half wave had predominately even order harmonics and full wave had predominately odd order harmonics.

[@JamesGMunn]

I will be interested in your PC based test equipment setup! Maybe add a 2 port USB scope?

[@ElPasoTubeAmps]

I know your comment is 9 months old but I have made videos on using the Analog Discovery 2, also known as AD2, that is what I had in mind as a PC based test setup. It does work and you may like it. I have posted a few videos on the AD2 and so have others. Hopefully they will help.

[@harbselectronicslab3551]

David what did you do as a daytime job.....you have an amazing passion for tube gear, and its very addictive .

[@ElPasoTubeAmps]

Hi Harb, Electronics and vacuum tubes were just my "first love" and I never could get enough while I worked. Such a beauty in them that can't be seen in solid state equipment. My job was always with solid state equipment but now that I am retired I can be a kid again and indulge myself to the point of insanity and share it with the world on KZbin. :-)

[@bucyruserie1211]

Hi David, on the 3rd amplifier you showed (the Fisher I believe) you weren't able to get the error line as straight as on the 1st amp with adjusting the 50k pot... was that because there wasn't enough range? If you changed the value of the 68k resistor in series with the pot, would you then have been able the get the line straight? It did increase the performance quite a bit on the 1st amp.. Great idea, Great video..

[@ElPasoTubeAmps]

Thank you. That error signal that comes out of the Tek AA501 is very useful but I have learned to take trying to make it perfect a losing battle. It is the sum of the errors and each amp seems to have its own harmonic profile so getting it to zero (or really low) can be frustrating. I am really pleased that so many viewers, like yourself, have found this approach useful and only through real empirical evidence we see that the "matched" resistor approach is not the best solution and making this parameter adjustable does indeed improve performance. As for the Fisher 70-A, it is far from perfect in the high end (on the test equipment) but sounds fabulous and has a really nice low end and along with its matching preamp the 50-C, a pair in good shape can sell on Ebay for big bucks. I listened to one of these Fisher amps and preamps starting at 12 years old playing out of a huge speaker system (looked much like the Klipschorn) but running the EV 15W and some University mid range and tweeter horns and I haven't gotten over it yet.

[@annomaus8386]

Can you recommend a free FFT program? And can't I use the internal sound card? It's good to 48000 @ 8 bits and decodes MSK144 and JT9/65 FB. Nice vid!

[@ElPasoTubeAmps]

You can download and use ARTA free with some small limitations that are unlocked with you purchase a license and you can download and use SpectraPlus fully functional for 30 days. They will work with your sound card.

[@gizmothewytchdoktor-419]

if transformers for example are of a set value of resistance and the two sets of windings are to de-couple the tube outputs from the transducer why not decouple with capacitors and use resistors for load value like a transistor amp?

[@ElPasoTubeAmps]

The transformers convert the high impedance side of the tube (typically about 5000 ohms) to the low impedance side of the transducer (8 ohms or so). An RC circuit wouldn't do this impedance transformation. In some solid state amplifiers where the impedance of the output transistors are very low, they can be direct coupled to the speaker or isolated through a capacitor as you mention.

[@Sparky-Tim]

Great video David. Could you use these methods to save money on buying matched pairs of tubes (or quads)?, as these can be quite expensive. Would also save on buying a tube tester to match tubes yourself.

[@johnsweda2999]

ever come across a leak TL12 12 Watts. Nothing much out there in tube land very good amplifier grandaddy of the Mackintosh 0.1 THD. got 1 for many years probably needs restoring don't know what to do with it. I have a EMI preamp as well can't find anything out about that I think it's a mic preamp mono no model number the serial numbers 0000001 maybe a test piece I'm thinking, silvery gold colour with grey knobs. where's a good website for schematics. like to get rid of them I think but not on eBay or swap!

[@God-CDXX]

I have to send you a simple resistor circuit you can use a single channel scope to do this

[@ElPasoTubeAmps]

I can already see what you are talking about. Sum the out of phase signals together and use a single channel scope. I assume that is what you mean. It just never crossed my mind but that is the beauty of these discussions to find other ways to get the same thing accomplished.

[@God-CDXX]

yes it is how I do it I use 2 20 meg and a 1 meg pot in a small metal box to calibrate the box I hooked it to a ct transformer with the ct being ground than I tweaked the pot until I had 0 reading on the scope than I locked the pot ... this is a cheep passive tool it cost me more for the 3 bnc connectors than the rest of the parts

[@God-CDXX]

from amps to tooth brushes

[@ElPasoTubeAmps]

Thank you - that's funny. I am working with some of those same men as the tooth brush on a self sanitizing mat last night at their lab. Working with Vacuum tubes can get you into all kinds of trouble :-)

[@God-CDXX]

I will stick to amps