Mr Carlson's Lab - MAAS 50 Audio Amplifier Redesign, understand feedback, bandwidth, and THD. PART 3

I have a Bogan Champion CHB-100 70 volt 7868 4 tube amplififier for a schulmerich, Caroline Bell, that I have repaired but it has a little problem that is complex and I have recomendations But the 6878 EH still have this intermittent red plate problem and I have heard thing like 100 ohm individual filiment and plate resistors are nesassary on each tube to intermittent contact on the B plus ect. i saw you have one or the bench right now and I am hoping that you could go through it and give some tips on here that would help me with some problems

Eric Martel

2024-01-27 16:04:38 +0000 UTC

Which is why a 3db increase is equivalent to a doubling of the power level.

Timothy Gregerson

2022-12-04 14:09:27 +0000 UTC

Q Why is dB defined as 20log 10? What is so special about 20? A This convention is in place due to the original definition of Decibel (dB) as a measure of a ratio between powers. If X, X0 are power measurements( X0 could be the input or reference power, X the output ), then the dB measurement was defined as X_dB = 10log_10 (X/X0) This way, if X is 10 times X0, we read X is 10dB greater than X0. On the usefulness of utilizing the log_10: it allows to express a very large range of ratios in a range of moderate size, allowing one to clearly visualize huge changes of some quantity. Many contributions to Control Theory have come from the area of Electrical Engineering. Since in a circuit with constant resistance, the power developed is proportional to the square of the applied voltage, then if V and V0 are two voltages of interest (I/O) one finds that: V_dB= 10log_10(V^2/V0^2) = 20log_10(V/V0) which is the reason why Electrical engineers use the dB defined as 20log_10(V/V0)

gunsofsteele

2022-11-06 12:02:21 +0000 UTC

For those whose mind got stuck at wondering why 20log?

gunsofsteele

2022-11-06 12:02:00 +0000 UTC

Great Project thank you Mr. Carlson

LoSaYa

2022-10-19 19:34:39 +0000 UTC

thank you , star getting familiar and comfortable .

Jorge Perez

2022-06-28 21:08:24 +0000 UTC

Paul: Great video. i have been watching it over and over. If I were to try to order the audio output transformers and power transformers, from Edcor or Hammond for example, what specifications would I use? Do you recommend specific transformers for a person who would like to build a set of these monoblocks?

Bobcat

2022-04-21 00:33:19 +0000 UTC

Mr Carlson...in this video you talk about not exceeding the capacitance rating for rectifier tubes, but in a much later video (I can't recall exactly which one) it seems like you are saying that it's ok to add more capacitance for filtering. Which is it? Thanks, and love all your videos and inventions that you share with us!

Edward LaRose

2021-08-08 00:15:34 +0000 UTC

WOW! I really enjoyed this one. An awesome look into audio amps and I loved how small changes in filtering and feedback cause big differences in how it performed. I saved an old tube amp out of a Magnavox console years ago. Time to think about using it to power the music in my radio room. Thank you for an outstanding class.

Jared Kusner

2021-06-05 11:51:21 +0000 UTC

Very interesting and very educational for me. I'm 77 years old have have a lot of experience with tube circuits but you increased my understanding. One comment though, I think you also changed the feedback resistor from 100k to 68K, thus increasing the feedback level?

Fred Winograd

2021-03-19 18:51:29 +0000 UTC

I would still like to see how impedance vs capatience and how it affects frequency response I get alot out of building block videos like that personally and they certainly increase knowledge.

Jason Martin

2020-06-22 08:07:38 +0000 UTC

Great video. The wish list of kit for my lab is increasing and I've only just begun watching on Patreon. I might need to build myself a bigger shed!

Keith Colin

2020-05-15 01:30:15 +0000 UTC

The 100K resistor in the grid of the input 6J5 also forms a low pass filter with the Miller effect capacitance of the tube. A grid stopper resistor id good design practice, but 10K is a more suitable value to minimize the low pass filter effects.

Sam J. Breidt

2020-05-03 20:45:52 +0000 UTC

Thanks!

Mr Carlson's Lab

2020-04-15 01:59:22 +0000 UTC

That was incredible excellent

Michael Infante

2020-04-13 13:35:04 +0000 UTC

Yes (RF and below) and Yes.

Al & Cheryl Lipscomb

2020-03-12 17:41:01 +0000 UTC

Bell curve? Are you trying to be punny? LOL

David Campbell

2020-01-12 04:44:05 +0000 UTC

Originally, the two tubes you replaced had metal shields on them. Was this for RF shielding? Are metal tubes in a vacuum like glass tubes?

Terry l Fiala

2020-01-05 05:41:24 +0000 UTC

This was great. Thanks!

David Slaughter

2019-11-06 05:02:43 +0000 UTC

Would love to have heard music and voice on that modified version! Books on Tape stuff, instrumental music....

Wayne Thompson

2019-09-06 11:38:13 +0000 UTC

Hi Kevin. It all depends on the strength of the noise, and the circuit involved. Many older vibrator circuits used .0068 (and around) valued capacitors.

Mr Carlson's Lab

2019-07-17 09:31:21 +0000 UTC

Thanks Kevin!

Mr Carlson's Lab

2019-07-17 09:29:05 +0000 UTC

P.S. - luv'n the video's...

Kevin

2019-07-16 16:29:00 +0000 UTC

Mr. Carlson, I was wondering about the capacitors you used to reduce the switching noise of the diodes. Can essentially any 0.0068uf capacitor be used, is there a value range that will work fine, or should a specific value be calculated, or ? Essentially I'm wondering if one could just keep a certain value capacitor (like the 0.0068uf) around specifically to use across the power/rectifier diodes?

Kevin

2019-07-16 16:28:46 +0000 UTC

It is amazing what you can achieve with trial and error, thus, I am impressed how you can shield the amplifying tubes from the rectifying tubes. I liked how easily you can handle the Network Analyzer to record the Amplitude Frequency Response and the harmonic distortion. The downside is that such Network Analyzer are very expensive. I learnt quite a lot from this video - probably one of the best video you did, I would say!

Roland Bernold

2019-06-06 19:00:04 +0000 UTC

Mr Carlson, wow! What a video! My head hurts, but I'm understanding. By the way, I have two cats, and let me tell you, they are a real "long-tailed pair." Haha

Bruce Dow

2019-04-09 19:09:25 +0000 UTC

Hi Paul. Another great video, thank you! I notice the pre-amp and phase splitter valves don't have cathode bypass capacitors; would there be a benefit to include them?

philippe retourne

2019-03-23 13:47:35 +0000 UTC

You're welcome Mike.

Mr Carlson's Lab

2019-01-26 14:05:47 +0000 UTC

Awesome video, thank you. Such valuable information.

Mike

2019-01-24 19:49:42 +0000 UTC

Thanks Paul, a tremendous amount of insight and experience packed into this segment. I really appreciate the care that goes into making these videos and I am learning a great deal.

Joel M. Sciamma

2018-12-01 09:52:36 +0000 UTC

Very nice video. i'm new in electronic with a very basic knowledge. i was wondering if you ever consider selling those amps as a kit for beginners like me... it would be great for us to practice along. Again tks for the video...

Roger Canal

2018-11-27 23:58:15 +0000 UTC

Long but WORTH it! As a guitar amp guy, I had to smile at the hilariously low THD values. Even a notoriously "clean" Fender might typically have 8 to 10% THD at moderate volume levels. ;-)

Gee Bee

2018-11-19 20:29:12 +0000 UTC

You're welcome Donald!

Mr Carlson's Lab

2018-11-01 18:28:33 +0000 UTC

Holy smokes, this video is full to the brim with really informative amp design tips and live analysis using various components. That light bulb in my head is once again burning a lot brighter thanks to you, Paul. :) Awesome!

Don Bonin

2018-11-01 14:28:51 +0000 UTC

Wonderful video Paul, and as always with your videos I learned a lot. I have heard that using fast recovery diodes will result in less diode switching noise. Is there any truth to that? Thanks again!

David Rohrer

2018-09-14 02:40:28 +0000 UTC

You're welcome!

Mr Carlson's Lab

2018-09-06 09:39:54 +0000 UTC

Thanks Paul. That was a lot of fun!

Robert Calk Jr.

2018-09-06 07:57:39 +0000 UTC

What exactly do you want to know about clipping? This may help me narrow down a video.

Mr Carlson's Lab

2018-05-21 10:56:18 +0000 UTC

Hi Paul, A couple of times in your videos you mention clipping on the sine wave and you said you would explain what that is in further detail but I do not remember hearing your explanation. Can you pint me to a video that has a full explanation? Thanks.

Kenneth McCarthy

2018-05-18 11:36:55 +0000 UTC

thanks for the ans. the pattern on the oscilloscope was a half wave pattern and that confused me. thanks again. doug

Doug

2018-04-11 01:32:49 +0000 UTC

Hi Douglas. The rectification is full wave (same as a bridge rectifier circuit, just configured differently) The large caps are making up for the missing reactor.

Mr Carlson's Lab

2018-04-11 00:35:57 +0000 UTC

very good explanation of everything. thanks for the time u put into this. one question. why not use a full wave bridge in the power supply then u wont need the large caps??? thanks again. doug

Doug

2018-04-10 14:48:52 +0000 UTC

Hi William, all the attachments are right under the video description (in red) with the little grey paper clip beside them.

Mr Carlson's Lab

2018-04-07 22:14:04 +0000 UTC

Where are the Bluw attachments that I'm to click?

William Harris

2018-04-07 21:40:42 +0000 UTC

Hi Arnon. I got my ceramic standoffs in an estate sale, they are available here: <a href="https://www.surplussales.com/Hardware/Standoffs/ceramic.html" rel="nofollow noopener" target="_blank">https://www.surplussales.com/Hardware/Standoffs/ceramic.html</a>

Mr Carlson's Lab

2018-03-18 07:52:54 +0000 UTC

Hi Paul. How you call and where you get the small ceramic isolators you use to substitute electrolytic?

Arnon

2018-03-17 21:19:54 +0000 UTC

Thanks for taking the time to write Arnon!

Mr Carlson's Lab

2018-03-16 04:09:49 +0000 UTC

Excellent video Paul. I was deeply impressed by the distortion figures. These numbers are not usual in tube amplifiers. This video could go for hours before reaching all the potential of this amplifier. There are many points that call for improvement. Starting at the output stage one could use individual cathode resistors to obtain a perfectly balanced current across all tubes. On the phase inverter circuit any other configuration would probably perform better than the Paraphrase Inverter. Although this is not my preferred phase inverter configuration we should remember that the Quad II amplifier of the golden era of British amplifiers implemented this circuit. The Quad II is highly collectable and is a continuously celebrated amplifier. The Paraphase Inverter is the less elegant inverter solution but the results obtained by Paul are amazing. I understand that Paul didn’t want to change the “character of the MAAS design so lets leave the inverter as is. I would like to suggest a more elaborated solution for the added preamplifier if I may so. Instead of the two triodes at the input stage, possibly a single high gain pentode could be directly coupled to the phase inverter. This would minimize phase distortion and thus the feedback signal could be applied to the pentode. I didn’t check if pentode gain would allow the same amount of FB but possibly you will find it something to explore.

Arnon

2018-03-15 03:37:14 +0000 UTC

Hi Jaan. The input impedance of the DUT can vary, and will have no affect on the calculation if done correctly. 20 Log is part of the formula to calculate feedback in db.

Mr Carlson's Lab

2017-12-24 16:44:46 +0000 UTC

Very good explanation, very detailed, I like it. I have question about the feedback value to convert to dB format. By definition Detsibell is power ratio. If you measure voltages then you have to input and output resistance must be equal and then the formula 20log can be use. Or is this the case that we do not care about the power and we need only know voltage amplification. And we do not worry about the input and output resistance? And the 20log formula is fine

Jaan Viil

2017-12-24 14:53:31 +0000 UTC

That's why fella's like me are here :^)

Mr Carlson's Lab

2017-12-19 02:31:00 +0000 UTC

This is why we have a "throwaway" society. No one wants to or many times can fix things. Peoples attention span is so short. This applies not only to these old amps, but to almost everything that we touch or use today. Look at automobiles, they are so complex that not many can fix them and most people cannot afford to fix them when they do break. This drives me crazy.

Michael Lynch

2017-12-18 20:39:58 +0000 UTC

Hmm. . Tektronix 547 Oscilloscope. . thou shalt not covet; thou shalt not covet. . . :)

Frank

2017-10-29 02:49:12 +0000 UTC

Studied and appreciated the mods you made to the MAAS 50 Amp. What I found interesting was the increased bandwidth achieved by way of simply removing the three capacitors (250pF @ plate 6J5, 470pF @ Grids near 6L6). In my experience capacitors of that size in locations like that where sometimes ruthlessly employed to prevent AM reception, and therfore engineer a low noise audio amplifier rather than a boistreous broadband AM Radio. ;) Please note that even the far more modern Edwin Amplifier tended to sometimes operate like a radio. Something possibly due to HF pollution in various European Cities.

Frank

2017-10-29 00:11:22 +0000 UTC

The speakers do affect the amplifier quite a bit actually. This is a good idea for a future topic, thanks for your input Serge.

Mr Carlson's Lab

2017-07-23 03:52:50 +0000 UTC

Amazing video and thx for putting so much effort in it. I'm learning at lightspeed! I'm a noob on this amplifier analyzing and I don't have the gear to test it myself. So can you help me with the following consideration? Al tests in this video are done with a resistive load. Speakers are inductive loads. In what degree do speakers influence the output behavior of the amplifier? It would be great to see some tests with different builds of speakers attached to the ouput .

Serge De Saffel

2017-07-22 11:27:17 +0000 UTC

Take away gems for me on this one was the elegant way you found the diode ringing—which I didn't even know was a "thing"—and how utterly trivial it was to eliminate: a couple caps across the diodes. Second gem was the idea that one can put too much filter capacitance in the PS, such that it can kill the rectifiers on startup. Question on the diode ring: Can solid state amps fall victim to this, or is it pretty much just vacuume tubes the do this sort of "listening?"

Scott Morison

2017-06-10 04:20:20 +0000 UTC

Sounds like a great video idea Marc, thanks for your input!

Mr Carlson's Lab

2017-05-02 08:39:14 +0000 UTC

Thanks that makes sense now ;) Out of curiosity, I just finished building a Tube amplifier kit (a VTA ST120 which is basically a ST70 with 6550 tubes and larger transformers :) ) I got it built and it sounds great but me being me always wants to learn more.... One thing I was noticing is that you have a lot of fancy test gear, stuff that is likely out of range for most of us mere mortals :) I was doing some research and found something called the Analog Discovery 2 Pro bundle that looks like it will do most of what an amateur hobbyist, needs with regards to amplifier building and troubleshooting, was wondering if you were familiar with this device and if so what your thoughts are? (it incorporates a function generator, o-scope, logic analyzer etc etc) the pro bundle comes with what looks like a NICE array for 300 US... If you don't think this would be a good choice for a learning amatuer, would it be possible one day to make a video explaining what someone getting started would need on his/her bench? Thanks man!!!!

Marc Hering

2017-04-30 22:27:49 +0000 UTC

Hi Marc. the -3dB, or "3dB down point" is the half power point. This is the industry standard for looking at useful amplifier bandwidth. Bandwidth, or sometimes known as "frequency response" is the area of effective listening.

Mr Carlson's Lab

2017-04-30 00:40:15 +0000 UTC

Def learning a lot, couple of questions tho...When you did the initial gain measurement, why did you go down 3dB for the reading?

Marc Hering

2017-04-30 00:14:41 +0000 UTC

Thanks Barry! Much more coming.

Mr Carlson's Lab

2017-04-19 14:36:59 +0000 UTC

Hi, I was pleasently surpried at the in depth nature of this video. You explained dB and negative feedback in a simple relaxed manner. Also the diode ringing was something I thought that only I had noticed in the past... amazing! I envy your calm patient manner and am looking forward to more tube related content in the future. please keep them coming.

Barry Swanborough

2017-04-18 23:45:17 +0000 UTC

Hi,

Barry Swanborough

2017-04-18 23:35:47 +0000 UTC

Paul, I'm curious where you find so many of these old tube treasures. Are you just attending ham fests or searching flea Markets? You seem to find such interesting pieces. I've located most of my stuff on ebay but it's not the best source in my opinion. I did find two good pieces at a ham fest in Orlando recently but many pieces i would have liked to buy were over priced. I love the restoration videos of tube equipment. Keep them coming.

Mike Harden

2017-04-07 01:58:59 +0000 UTC

Hi Fred, with the post enlarged, you should be able to view all the comments. If not, go to the top of the post and click "Load More Comments" To enlarge the post, click on the large black title under the video.

Mr Carlson's Lab

2017-04-01 04:54:58 +0000 UTC

When viewing these comments I am only able to view up to 75 of 90. Is there a way to load and view beyond 75 comments?

Fred W7IPM

2017-03-31 23:30:16 +0000 UTC

One thing for sure.....the more this course gets known....the higher the price will be on old test gear as people start to realize its true potential .....oh well....lol. Still a fair trade off I guess.

Rick ross

2017-03-26 04:29:24 +0000 UTC

Those 500 series oscilloscopes were awesome instruments in the 60's and still are. There is so much high quality test gear out there for cheap money. Digital is nice, but for working on vacuum tube equipment I think a good quality analog scope will serve you better

Bill Moran

2017-03-25 16:48:36 +0000 UTC

Excellent lesson on amplifier design!

Bill Moran

2017-03-25 16:44:55 +0000 UTC

2) (sri, gotta get used to not using CR) Were you able to pick up that SR780 reasonably, say as a basket case rebuild project? A quick search showed used prices in the 3-7K $ range. My Rigol DS1102E has a basic FFT static screen function but of course it's not nearly as useful. BTW, for above I have some 5R4WGBs NOS I could use for nostalgia's sake (and winter shack heating) that were in the same box as the 829Bs. I think they are surplus airport radar tubes based on the types in the box. Also have a few big 4.4H chokes from surplus Wx station PS racks if I wanted to go whole hog. I'm open for suggestions on how to work this stuff into a project or two if something springs to mind. tnx! great vid!!

chuck yahrling

2017-03-25 13:18:48 +0000 UTC

Coupla questions: 1) I have a bunch of scavenged microwave oven diodes that give me something like 4-6 volts of ripple with minimal cap filtering, at 300-400V full wave rectified out of Triad R26A secondary delivering 200ma. Are they too "loose" to use as primary rectifiers?

chuck yahrling

2017-03-25 13:02:48 +0000 UTC

Hi Stephan. That's just a way of putting more stages within the loop. By doing this you get lower overall distortion, but at the cost of gain. If you can keep a stage or 2 out of the loop, you get the added gain. Sometimes when amps have really ugly THD specs, doing this helps as well.

Mr Carlson's Lab

2017-03-25 10:31:10 +0000 UTC

Hi Paul Loved this video. It was very informative. With regards to the feedback loop, I've seen designs which feed the feedback loop to the cathode of the pre-amp tube rather then the input of the phase inverter. Would you mind explaining the difference in both designs ?

Stefan Saliba

2017-03-25 08:48:16 +0000 UTC

Glad you're enjoying Chuck. Thanks!

Mr Carlson's Lab

2017-03-25 04:14:07 +0000 UTC

course at the right time, so tnx for being there for us, 73, chuck AB1VL

chuck yahrling

2017-03-24 23:11:30 +0000 UTC

Hi Paul, glad to see you are expanding into great new territory. I'm a fairly new ham, licensed in 2013, mostly CW and wsjt-x digital. I run a K3 I kit-built and an SPE FA-1K into aHygain explorer 14 on a 52 foot crank up, along with a home-bre w 6M moxon and other wire ants, FWIW.and have gotten some great RF elmering thru your vids - info I could not find elsewhere. Help me Obiwan, you're my only hope <g>. I picked up several RCA 829Bs NOS a few years back and have been looking for a tube amp project ; candidates range from a 2M power amp to boost the 40W from my IC7000 (for MSK441) to a redesign of a Fender Bassman circuit for live music. So I'll be interested in learning all the angles that need to be considered. I joined b/c I think your course will be the right </g>

chuck yahrling

2017-03-24 23:10:45 +0000 UTC

The Tek 547 looks nice. I don't have any experience with it. I got hooked on the Tek 7000 series. The 7A22 and 7A13 are nice diff amp plugins. I like to collect old gear with respectable specifications. <a href="http://w140.com/tekwiki/wiki/7A22" rel="nofollow noopener" target="_blank">http://w140.com/tekwiki/wiki/7A22</a> <a href="http://w140.com/tekwiki/wiki/7A13" rel="nofollow noopener" target="_blank">http://w140.com/tekwiki/wiki/7A13</a>

Nathan Haltman

2017-03-24 18:07:32 +0000 UTC

Hi Patrik. Try to go for an "ultra linear" output transformer if you can. The screen taps make life much easier, and they sound great. I prefer the Hammond 1650R (built many amps with those) If these are hard to get in your area, just use the specs and find something similar.

Mr Carlson's Lab

2017-03-24 05:36:26 +0000 UTC

Finally done with the video and the 50W at 0.19 % THD is extremely nice. I am actually considering building this version, however and since I am a n00b, I will need new mains transformer (230vac) and an OPT. Considering Lundahl for both. I guess the Voltage tags on the schematic is enough to find a replacement ? - I have both books from Morgan Jones, Valve Amplifiers and Building Valve Amplifiers that I will read before attempting to put this amp together, but your input on the transformer (issue) would be nice. Many thanks in advance.

Patrik Lövström

2017-03-23 21:24:40 +0000 UTC

Thanks for your input Patrik!

Mr Carlson's Lab

2017-03-23 17:09:04 +0000 UTC

Btw, I would buy the book. Or write chapters and release them here on Patrion for us who are very much learning and in an early stage of amplifier design.

Patrik Lövström

2017-03-23 16:30:37 +0000 UTC

I had to do it.. LOL.

Mr Carlson's Lab

2017-03-23 08:30:33 +0000 UTC

Great!

Mr Carlson's Lab

2017-03-23 08:30:10 +0000 UTC

Thanks Paul!

Mr Carlson's Lab

2017-03-22 14:25:48 +0000 UTC

It's done my friend!

Mr Carlson's Lab

2017-03-22 14:04:36 +0000 UTC

Hi Dennis. The 6V6 can not source the amount of current to act as an electronic filter. The weak point would be the bonding wire from the cathode to the lead in wire at the bulb base. To utilize electronic filtering, one would have to use a bunch of 6AS7's or 6080's..... maybe even a 6336 to handle the current draw from four 6L6's at full power. output.

Mr Carlson's Lab

2017-03-22 14:00:23 +0000 UTC

Hi Douglas. The 10K resistors on the grids are purely non inductive, and stop the grids of the tubes from seeing each other. The impedance of the plate circuit and grid circuits are very different. The plate circuit has a much lower impedance and only requires about 10 to 50 Ohms of resistance in the stopper resistors to be effective. Whats the number of the plug in for the 7000 series scope?

Mr Carlson's Lab

2017-03-22 13:55:59 +0000 UTC

If the 6v6 stabilizer was configured as a capacitor multiplier could the power supply capacitance be lowered enough to prevent the 5v4g rectifier from going boom and provide sufficient ripple rejection?

Dennis

2017-03-22 13:09:48 +0000 UTC

as the 500 series differential amp plug-in? I know you don't have space to give detailed explanations on the comments. Maybe my questions will be answered in a future video.

Douglas Tees

2017-03-22 12:09:52 +0000 UTC

Another great enjoyable video. There are gaps in my tube amp or electronics education because I am self taught. I am hoping to improve my skills here. Here are my 3 questions. 1. Why do or why would the 100k resistors on the grids stop oscillations? 2. Why are only 10 ohm resistors required in the plates on the 6l6's to stop oscillations. I have a 7000 series scope. I have seen on ebay differential amplifier plug in's for it but I never how they were used. 3. Would you consider the 7000 series differential amp plug-in to be as effective or in the same league

Douglas Tees

2017-03-22 12:07:19 +0000 UTC

Thanks for all your input!

Mr Carlson's Lab

2017-03-22 06:24:36 +0000 UTC

Wow, what a great lab... Good lord

Paul

2017-03-22 03:08:11 +0000 UTC

Bell curve ... LOL

Patrik Lövström

2017-03-21 23:46:04 +0000 UTC

And finally: i would love to hear some words from you about on how to design a nice class AB2 PushPull amp, where the output tubes are driven (well) into positive control grid voltage region. If you want to do me a favor: to see a video covering that toppic would be highly appreciated...Thanx :)

TubiCal

2017-03-21 23:43:57 +0000 UTC

Rectification problems are mostly overlooked, sometimes builders whack them in and just leave them alone, But they would do a much better job when adressing comutating problems and/or cut-out/reconduct issues. You can sort this out by adding C-L-C or even L-C filtering in order to stay in SOA for them rectifiers. And/Or if you using some tiny capacitors across any diode helps a lot....as you just showed at around 1:13:00 and ongoing. ;) Andother nice touch would always be when you shield the PreAmp tubes with a can, as you done here for that test as well....most Instrumatation amps have these on their preAmp tubes to reduce hum....but they also keep on using the entire chassis as the return path (AkA Ground) - as a great source of hum. you may talk about this in a future video as well...

TubiCal

2017-03-21 23:35:56 +0000 UTC

And a final word on these comments, please reactivate the coments on the video page, as this is, by far, much worse than the YouTube comment system. i can´t even create any form of paragraph here...

TubiCal

2017-03-21 23:07:49 +0000 UTC

And finally i would go for a fixed bias output tube design, one for each tube. The 6L6 tubes would run cooler and also providing more "uumph". I know you´re not into intrumental amplification but you could also "tune" the "sound" of the amp by variing the negative control grid bias lightly.

TubiCal

2017-03-21 23:05:00 +0000 UTC

And as soon as you presented the schematic in this video i went: uuh what ?!? ;) i´ve seen this kind of phase inverter in an old guitar amp at least one time...ugly..i converted that thing into a standard chathodyne phase splitter.

TubiCal

2017-03-21 23:01:24 +0000 UTC

I do like your complete and in depth explanation of all these things....But i also think that roughly 70% of the "sound" of a tube based Amp is made or has its origin in the output (or interstage if present) tranny. I built so many amps and also serviced a whole lot more that i can tell that everytime i needed to change an(y) audio related tranny this also done a huge change of the "sound" as well...

TubiCal

2017-03-21 22:53:39 +0000 UTC

Awesome video Paul, I learned alot from this video.

Leland Clayton

2017-03-21 22:19:59 +0000 UTC

Great to read Carl! Thanks for your comment.

Mr Carlson's Lab

2017-03-21 18:49:07 +0000 UTC

Well Paul,,, think your channel name should change to Prof Carlson's Lab. Gotta tell you I really learned a lot just on this video. This is just great. Thank you.

Carl Loar

2017-03-21 18:46:20 +0000 UTC

I don't know if you want to do more editing before releasing into the wild or not. If you do, I'd recommend overlaying the schematic at <a href="https://youtu.be/KqrKtGn1yyo?list=WL&t=394" rel="nofollow noopener" target="_blank">https://youtu.be/KqrKtGn1yyo?list=WL&t=394</a>

David McAnulty

2017-03-21 16:15:40 +0000 UTC

Correct. More capacitance = more filtering at the top end. Introducing just one capacitor on the grids of the 6l6 tubes (470pF) doubled "and some" the distortion specs. The 250pF cap was moved to the grid of only one 6L6. This was a compromise between bandwidth loss and stable operation away from parasitic oscillation.

Mr Carlson's Lab

2017-03-21 14:37:02 +0000 UTC

A quick question. There's a reference to possible distortion owing to only one of the 6J5 tubes having the 250pF capacitor from plate to ground, and it strikes me that since the grid of the second 6J5 is connected to the plate of the first (via a voltage divider), that the HF response of both tubes is being shaped by a single capacitor. Surely if there was a capacitor from the plate of the second 6J5 to ground, then it's output would have even more top-end rolloff, since it's grid signal was derived from the (already filtered) plate output from the fist tube, then you'd be applying even more filtering at the plate of the second tube?

Karl Adams

2017-03-21 14:05:48 +0000 UTC

I think feedback (the amount of) is personal choice. I have designed push-pull amplifiers with low distortion stages and no feedback, and I love the sound. I installed a feedback switch so I could compare on the fly..... over and over I found myself listening to the amp in "no feedback mode" It seems to make everything stand out more. As for the rectification, yes it's from the SS diodes only, it's known as diode ring. Whats actually happening: The rectifier tubes on the top of the chassis are acting as antennas. The diode ring is getting on the top of the chassis this way. If I were to remove the rectifier tubes, the 6SN7 wouldn't hear the noise as much. The .0068uf caps and the .1uF cap across the large can (470uF) take care of the noise.

Mr Carlson's Lab

2017-03-21 12:50:31 +0000 UTC

Thanks Chris!

Mr Carlson's Lab

2017-03-21 12:39:11 +0000 UTC

Is the switching noise a product of SS rectification or can it also appear with tube rectification ? I was surprised of how flat (if narrow) the frequency response was, event with no feedback. I expected a hump somewhere. Thanks for the demonstration on feedback, it basically looks like it cuts the head off the response curve, lowering gain but widening it, very interesting. You said that extra feedback if often used to hide design issues. Do you see any downside to it besides reduced gain ? (will a system designed for maximum linearity with no or minimal feedback work / sound better)

Alexandre Côté

2017-03-21 11:44:55 +0000 UTC

ok so that was worth the wait. you sure put alot of effort into that video. thanks so much for your efforts. so very cool. your a mad genious!!!!!

chris lee

2017-03-21 10:49:17 +0000 UTC

That would be difficult but do-able. You may need a sensitive pre-amp for your scope for noise measurements though.

Mr Carlson's Lab

2017-03-21 04:23:32 +0000 UTC

Hi Anders. I didn't measure the impedance of the output transformer, so I couldn't tell you at this point. The 250 ohm tap design was chosen just so any output could be used, and less thinking on the installer was needed. This way equals: less mistakes and a possible warranty call back due to oscillations melting the amp. The bump at the lower end of the frequency spectrum is just do to a portion of the circuit being tuned by the stretching of bandwidth through the application of feedback. The paraphase phase inversion scheme is sensitive to this (being too much FB) and prone to oscillation. FB is a Fine Balance :^)

Mr Carlson's Lab

2017-03-21 04:21:08 +0000 UTC

Hi Matteo. The "Audio Cyclopedia" by Howard M Tremaine, is great. You may also be interested in the "Audio Measurement Handbook" by Bob Metzler.

Mr Carlson's Lab

2017-03-21 04:11:30 +0000 UTC

Thanks Jon!

Mr Carlson's Lab

2017-03-21 04:04:07 +0000 UTC

Circuit impedance is important for power transfer "efficiently." You will note that plate impedance and grid impedance are sometimes vastly different, yet a capacitor is all that's between stages. This is why (for extra shielding) the outside foil was marked on the early capacitors so one could place that side of the capacitor to the lower impedance portion (plate of the previous stage) in the circuit. In audio amplifiers, the impedance in stage coupling is somewhat overlooked. The difference in the "overlook part," and proper impedance matching is a vast difference in cost. There is a catch to this as well... Proper impedance matching would require an interstage transformer, the iron in this is more likely to limit the frequency range of the amplifier over capacitive coupling option. You will note that all the early amplifier's out there all have transformer stage coupling..... Back when (doing it right) meant more that cost. Making really good (wide bandwidth) audio interstage transformers is expensive and time consuming.

Mr Carlson's Lab

2017-03-21 04:01:24 +0000 UTC

I think ebay would be your best bet. To build this from scratch would be costly.

Mr Carlson's Lab

2017-03-21 03:47:04 +0000 UTC

That's great Andre!

Mr Carlson's Lab

2017-03-21 03:46:14 +0000 UTC

Glad you enjoyed Filiberto.

Mr Carlson's Lab

2017-03-21 03:45:59 +0000 UTC

LOL, great machine, but a pain to set up. These SR-780's are still current, just the floppy is replaced with a usb. The menu requires a legend due to so many functions under functions. What you saw when I went from swept sine to FFT , was a path already created. If there was no path, you would have seen me poking buttons for 5 minutes. The user interface is ugly in this machine.

Mr Carlson's Lab

2017-03-21 03:45:23 +0000 UTC

These were all intended for "bell service" so I'm not sure if there are any with better bandwidth and low distortion.

Mr Carlson's Lab

2017-03-21 03:40:48 +0000 UTC

That's the nice thing about name brand caps. It's close, but still safe. I wouldn't trust this with a counterfeit.

Mr Carlson's Lab

2017-03-21 03:39:18 +0000 UTC

Thanks Russell.

Mr Carlson's Lab

2017-03-21 03:36:07 +0000 UTC

Thanks for your input Alexandre.

Mr Carlson's Lab

2017-03-21 03:35:54 +0000 UTC

Thanks James, glad you enjoyed.

Mr Carlson's Lab

2017-03-21 03:34:56 +0000 UTC

oops it's later here and pressed carriage return before I finished. I do not have a spectrum analyser and was thinking old school, like you did in explaining how to measure volts in and out with a meter (I have two old school (VTVMs)and an oscilloscope . Where could I find the correct log paper to draw out what your analyser does in hard form? Could it be done in a semi modern way with excel and a graph? With a bit more knowledge I should be able to do noise measurements with my meters they are very sensitive (I assume here)?

Steven Calver

2017-03-21 02:51:10 +0000 UTC

Awesome presentation!! Thanks!!

James Westhoff

2017-03-21 02:50:14 +0000 UTC

Excellent thoroughly enjoyed that video found it very useful. Thankfully I had some understanding of reading a schematic and spotted those 250 & 470 pF capacitors straight away assumed what they were used in the circuit for and awaited for my assumptions to be confirmed. I made notes as I went through and found it no trouble to skip back if I missed something whilst I was writing my notes.

Steven Calver

2017-03-21 02:45:00 +0000 UTC

I think you should post these on youtube at some point. Those videos are a lot of work and can help many more prople

Alexandre Côté

2017-03-21 02:24:23 +0000 UTC

Amazing thank you.

Russell Smith

2017-03-20 22:58:15 +0000 UTC

Another question, i noticed that you put a 68k resistor for the feedback to the 6sn7 from the "250" tap of the transformer, why did they use this 250 tap in the design?, couldent you just gotten that from the 16 or 8 ohm tap - or perhaps they just used a standard Audio transformer... very rugged design.... this one will run for a 100 years 24/7, ...do you know the input impedence of the transformer? with 6L6´´s in push - pull in parallell it cant be more then 4-5k or?

Anders "SM7HCE"

2017-03-20 21:04:37 +0000 UTC

One question - the 450V cap on the B+, isent that to "cut it a bit close" for this one with 444V B+? - poor cap :-) haha - anyway - great video, keep it up please!! / Anders

Anders "SM7HCE"

2017-03-20 20:44:39 +0000 UTC

Paul - Great video :-) now we havent had any system like that in our churches as far as i know, so to get a pair of these would be an ebay matter, and probably would involve a substatial amount of shipping dollars, but what i do have is a fair amount of 6L6 and 807´s so i´ll fix one or two up to try them out - you managed to get really nice specs on this baby :-)

Anders "SM7HCE"

2017-03-20 20:42:30 +0000 UTC

Is there an equivalently well designed&built amp that was intended for HiFi use from the beginning?

John Puma

2017-03-20 19:30:45 +0000 UTC

Excellent. Thanks.

John Puma

2017-03-20 19:28:21 +0000 UTC

Very informative, as a side note your signal analyzer reminded me of playing "scorched earth" with the sound effects and slow trace. haha

Mike Drzewiecki

2017-03-20 19:17:01 +0000 UTC

Hi Im a hobbiest in electronics but like to try building one, so much to learn I love it , thank you great job.

Filiberto Caraballo

2017-03-20 16:48:26 +0000 UTC

All I can say is WOW!, Great presentation. I had look through the original schematic and I did not pick up those faults. I learned a lot today. Thanks for sharing this informative video.

Andre Gopee

2017-03-20 16:44:55 +0000 UTC

Great video. About how much is it in parts for this? Two of these would look awesome in in the living room with low lights :)

Bengt R Kiil

2017-03-20 15:57:43 +0000 UTC

Thanks, some very useful info in this video. Not sure if you forgot or I missed it, but I would like to know the cause of the "bump" or peak at low frequency on some of the frequency response curve's. Also very interested on the subject of Z in a circuit and some of the more advanced topic's of amplifier design. Am building a 100w valve amp ( I know you don't need 100w, but am building it to learn and for fun) using 6 paralleled EL34's ( 3 each side PP UL; at first used 807's, g2 max 300v a problem). It's taken me a year, so can understand you loosing a week. Am ready to start building a chassis. Just winding the mains tfmr and ironing out issue's with the PSU, so your tips on the MAAS 50's PSU were helpful. It took me a long time to figure out how to drive the OP valves cleanly; am now getting 130w, so have plenty of gain to lose on NFB. Still have a long way to go. Thanks again.

Diabolical Artificer

2017-03-20 15:21:04 +0000 UTC

Yet another great video! My hero.:)

Jon Fredrik Våle

2017-03-20 13:39:37 +0000 UTC

Hello MC (Mr. Carlson), can you suggest us some book(s) too? I'm an engineer so no I'm not afraid of formulas or lengthy chapters. I'm kind of used to that. So, if you know of books which could support your videos (preferably amazon or ebay), I'm all ears.

Matteo Guglielmi

2017-03-20 11:16:13 +0000 UTC

Excellent presentation! A suggestion to anyone watching this presentation. It is a long presentation full of very useful information. I tried my usual rewind back and review what was said, but the presentation is so long that finding the right spot is very hard. Just treat this as a class and take notes. I would like to know, and I am asking because I am too lazy to look it up, are there active micro voltage probes that can display on a normal oscope in milli volts ie. X 1000 micro volt readings displayed as milli volt on a normal oscope? Now it is time to put on the bench my projects and look for and or apply what I have learned. Thanks. Love the minutiae, and I am looking forward to your little side videos as well.

Carl Davis

2017-03-20 10:17:43 +0000 UTC