Happy new year David. Hope 2025 is a great year although it seems we've already had a rough start... Anyway, I wasn't sure where to post this since it touches a few area but THD is currently what need some assistance with. It does cover the HP 8903 and Fisher amp and I've watched other videos covering this content. Anyway my first question is do you use any capacitor coupling when you connect from the output of the audio amp under test to your test equipment? Currently I use the HP 8903a and Pete Millett software to perform my testing and I'm wondering if I'm doing something wrong with the way I'm connecting to the output. What I'm doing is connecting my amp (Fisher 400) left and right output to two 100 watt 8 ohm resistors as a dummy load for each channel. I'm connecting the HP 8903a input across the resistor. I have the banana connectors configured to provide test points to connect my Sencore lo capacity test leads to (from my Sencore LC53 Capacitor analyzer). For awareness the output of the HP in sent into the aux input of the Fisher 400 I've been restoring for quite some time. My issue is that I've done a lot of work to this fisher using mods and Ideas I've found from yourself and some other youtube masters. Currently, I've successfully replaced all the Filter can caps with Hayseed Hamfest, Installed new Bias caps along with an individual bias adjustment circuit and the necessary coupling caps and resistors required to measure at the cathode of power tubes. There are additional electrolytic and resistor replacements made as needed. Everything has been done and I'm happy to say everything works however this nasty THD still exists at my output tubes. And the freq response from this Fisher 400 seems concerning dropping off starting at 10K at the output while the preamp seems to perform better. Right Side: To be clear this THD issue is concerning since its not linear across the channels. The right channel is much worse than the left. So this is what I know. At near 2 watts output I have a THD+N on the right that rises to 1.5% at 30 hz dips a bit down to 1.25 between 100hz and 1Khz and rises to 2% between 3 and 8K then starts to drop off at 10K and falls off by 30 K. Note that the freq response of this amp also rolls off at about 10K and is 5db down by 30K. As you can imagine the THD gets worse as the power is increased. Left Side: This side runs much better at the same 2 watts and starting at 20hz to 500hz I'm at .2%. Then from 500hz to 1K the rise starts slow and gets to about .5%. From 1K to 10K its a steady climb to 2% before it rolls off and follows the rite side down as the freq response drops off also. Preamp: Both channels are linear at the input to the volume control. At about 1 volt and 20Hz they both start at about .25% and start to drop to .15% by 40 hz to 8K. There is a strange ringing between 27 to 37HZ. Also at 8K to 30K the THD+N rise steady to .3%. To me this is great for the preamp of this period but I really don't have the experience to judge. I know its a lot of words that a couple of pictures could have displayed but I could use some insight. I've done too much work to not understand what is happening here. This THD was there when I started but I really thought once all the significant caps were replaced this would be resolved. I have moved the output tubes around. I have some additional 7868s and moving the tube pairs around does make changes but the essence of these curves is the same. Currently the best I can get it with the orientation of the tubes I have now. I'm not sure what the best way to couple the HP 8903 is. Also, I've tried to use my o-scope FFT and I can see the harmonics but I don't have as nice a setup as you with the HP 3580 showing those beautiful peaks... I did recently purchase at a fair price an HP 3561a dynamic analyzer but I'm not confident on how to insert this into the test. I'd hate to damage it. I don't think it will display as nice as the 3580. Anyway, please let me know your thoughts and any ideas you might have. I must be missing something and I don't understand why the L and R are not consistent. I do notice that the L and R are very slightly out of phase. I mean very slightly. When I put what appears to be perfectly symmetrical nice looking sinewaves over each other the left channel line is unmeasurably ahead of the right. Have a happy new year and look forward to more content.

The Original Westinghouse 7591 was a remarkably advanced design, and the complex plate material will likely never be accurately duplicated.

Thanks for the demo of the Collins R-390. 73 Paul AA1SU

I haven't heard what the theory of your procedure is. The comment below, by John Sitter, an experienced high vacuum worker does sound right. You will never get rid of any gas molecules / positive ions inside the tube. You would have to pump it down to a high vacuum 10E-12 to 10E-13 Torr. The issue with positive ions is they are attracted to the cathode. Audiophile people complain about the excess noise created by these positive ions on the cathode. A friend, who is a long time amateur radio operator should me a method he uses to rejuvenate low power tubes he uses in the QRP radios he builds. He applies +12V to the cathode, wrt to the plate. Yes, the tube has a low reverse voltage. With the filament on, he lets his tubes cook like that for several days. Theoretically this drives the positive ions off the cathode. Where they go is anyone's guess, maybe they get absorbed by the plate or, if the tube has a better flash, they go there. I have not tried this myself.

David: I restored a non-working Tektronix 106 a couple of years ago and after replacing the failed components I was pleasantly surprised to see that the 7189A output tubes were all fine and reasonably well matched. It calibrated within specification and i still use it on the bench today. My unit has ~ 900pS square wave fast rise time. Gotta love this old gear! Thanks for the video.

Hello again, I ended up ordering a Tazcam US 1+2HR audio USB Interface. Seems the specs on it might be a little better than the Focusrite. Lower noise floor. So the program you are using here is the Arta? Did you say you like the TrueRTA better? Which do you prefer? I appreciate your knowledge. Where is the signal going to the amp coming from?

I just did a conversion of a Packard Bell to guitar. I loved the sound! Thank you!

Thank you much for sharing your time and expertise. I'm comfortable building amplifiers from a good schematic and troubleshooting any issues, your shared knowledge really helps with the specifics of how it works and how and when to make changes to the published designs!

Merry Christmas! I know you have done other videos on your test equipment. I'm interested in purchasing some type of audio analyser so I can get accurate measurements of THD and frequency response. I would really like to get a Quant asylum QA403, but really can't afford something like that right now. In your opinion, what do you think would be the best software based device on a budget to get those measurements. I was thinking an Analog Discovery 2. I know there is a AD3 available now. It's $100 more. I didn't know if there was anything just as good but more affordable. I recall you using an audio interface with some software on one of your videos. Appreciate any advice! Thanks!

Hi

👍

Are you sure that's distilled water? The reason I ask is because there are no minerals in distilled water, which means it's not conductive. If you're on a well and used tap water, I can understand how it would work. But anything that comes out of a still has no minerals, ethanol itself isn't even conductive. Stick the probes from a digital ohm meter in a shot of vodka and see for yourself.

Always learn something new watching you work. Thanks and have a great Christmas and New Year!

You know what I found when I have been tweaking my amplifiers is that if we try to perfect the square wave that they produce, playing with the feedback and getting rid of that little bit of ringing makes them sound lifeless and sterile if that’s the word… like in my head I picture an all white room, white speakers just plain and colourless. I actually prefer a bit of overshoot and ringing personally and have left most of mine like that. I would say I play around and tune out maybe 60-76% of it and it sounds great to me!

It just dawned on me what is going on - I was driving the amplifier with a 2 KHz pure sine wave. Harmonics are integer multiples of the fundamental so they are going to be harmonics of 4, 6, 8, 10, etc. - all even. What I failed to realize is 6 and 10 ,for example, is the 3rd harmonic and 5th (odd) harmonic of 2 KHz. I was described as a kid in a candy store, and I have had too much candy... Holy Cow... the thing about it however, is the driving signal is a very pure sine wave so the amplifier is indeed creating all these even harmonics as artifacts. I kept shaking my head last night thinking I surely must be confused - and I was - when I put in the 1 KHz square wave which is the fundamental and all odd harmonics - that is what I got out. This has been banging around in my head and just did not seem right - and it wasn't. Sorry for the confusion.

Apparently, the reason why we like even harmonics is because the harmonics are octaves of the fundamental, so they tend to sound more harmonious and musical to the human ear, octaves tend to sound consonant, whereas odd harmonics sound dissonant.

Would it be possible for you to upload the schematic with some voltages written on it? I seen the schematic in an earlier post but it has no voltages written on it. For those of us who would like to try this Amp.

It is true, musically even harmonics can't even be called distortion. Musically even harmonics are not distortion at all, it is simply a musical octave. In Western civilization music, the music we all know, octaves are simply a doubling of frequency, and in music playing a note in any octave is always "still in tune". Therefore even harmonics can't be considered as distortion at all. When a musician doubles octaves, it thickens and warms the timbre, it is never dissonant. Odd harmonics however always generate additional notes in the music that are never anywhere near being "in tune". I would love to build this Amp now, oh boy just what I need, another project.

Something you might find interesting from ChatGPT on SE amplifiers. ___________________________________________________________________ You said: I tested a high-end single-ended (SE) amplifier last night and with an ultra low distortion signal generator (Tektronix SG505) I only got even order harmonics out of the amplifier. I was running the output at 3W into an 8 ohm load. The OPT was rated at 10W, 5K primary to 8 ohm secondary. The secondary across non-inductive load resistors. I can't explain no odd order harmonics or is how it performed a characteristic of a SE amplifier? ChatGPT said: ChatGPT The performance of your single-ended (SE) amplifier showing only even order harmonics is indeed characteristic of this type of amplifier, especially under the conditions you've described. Here’s a breakdown of why this happens: Even Order Harmonics Dominance: SE amplifiers, particularly those that use triode vacuum tubes, typically generate predominantly even order harmonics (such as 2nd and 4th harmonics). This characteristic stems from the inherent asymmetry in the amplification process of SE designs, where the waveform's positive and negative halves are not amplified equally. The even harmonics often provide a type of distortion perceived as more musically pleasant or "warmer" compared to odd harmonics. Lack of Odd Order Harmonics: The absence of odd order harmonics (like the 3rd or 5th) that you observed can be influenced by several factors: Tube Selection and Biasing: The specific tube used and its bias point significantly affect harmonic content. Some tubes, when biased correctly, can suppress certain harmonics due to their smooth clipping characteristics. Output Transformer and Load Characteristics: The interaction between the output transformer and the load can also influence the harmonic distortion profile. Transformers can, to some extent, affect which harmonics are more pronounced based on their design and how they handle saturation. Circuit Design: SE amplifiers typically do not have negative feedback (or very minimal), which allows the natural characteristics of the tube and transformer to dominate the sound, including the harmonic content. Negative feedback, when used, can help reduce distortion but also tends to make the output more linear and suppresses harmonics, both even and odd. Given that you were using a high-quality signal generator and a non-inductive load, your measurements are likely accurate, reflecting the natural behavior of your SE amplifier under the conditions tested. If odd harmonics were completely absent, it's a testament to the linearity and quality of your setup at the tested power level, particularly the effectiveness of the output transformer in maintaining waveform integrity. For further exploration, you might consider varying the load or adjusting the bias of the tube to see how the harmonic content shifts. These types of experiments can help confirm the behavior and characteristics of different components within the amplifier circuit.

Great video, as always. 73s and hope you have a very blessed Christmas!

Can these machines do broadcast? I would like to know if its worth messing with or not. What would it take to get some useful range out of these??? Idk what the [emf] function is but it goes up to 2.99v which would be significant compared to specifications at 1.5vpp. It may just need the right antenna but im uneducated in that realm. Im sure alot of these machines are very low use, squeaky clean and babied.

Be interesting to put a switch on that feedback capacitor to see how it affects the sound of the amp. Merry Christmas to you as well, I always enjoy your videos.

You applied injected a 2Khz square waveform but why would an amplifier has to be 10X times the fundamental? 2Khz X 10 = 20khz , I'm don't understand what makes the guitar amplifiers circuit design is 10X times the fundamental. The NFB does what to the Squarewave testing measurement?

If I remember correctly you have over 100uf cap directly connected to rectifier tube no current limit resistance.isnt that going to hard on the rectifier and hv winding of the transformer? Otherwise an interesting circuit.

Very interesting. I've testing quite a few valve amps both SE & PP & have never seen a FFT result that hasn't had any 3H content. It shouldn't be a test gear anomaly. I'd be interested in seeing the result with a 100hz or lower signal. Also interested in seeing the OP prior to the OP valve. Thanks, a very thought provoking video. Edit: Re sq wave test @ 1khz, looks like there's some LF rolloff. I usually set my NFB capacitor with a 10khz sq wave, then sw too 100hz sq wave & dial in the C. I usually have an LC/Zobel across the speaker OP teminals & from anode to ground set to the same F 3dB point. The increased instability/ringing at 2khz from 1khz isn't good though, could it be down to the OPT? Have you done any no load & capacitive load testing? A 100n is a good value, it show up any instability issue's - no load LF instability, cap - HT instability. Lastly, had a thought about the no 3H, the amps I've tested have been stereo amps with a common power supply, I've had issues with Xtalk, I wonder how much of the 3H content I saw was a consequence of Xtalk & or IM? Lastly I I agree, don't matter what it tests like, if it sounds good, leave it be.

Cracks me up! You are like a kid in a candy store with all the tweaking and testing you are doing on you SE amp. I can relate too! We are never satisfied or happy with the results... ALWAYS modifying and testing and testing! Have a GREAT Holiday and stay safe.

The generation of even harmonics in a square wave is dependent on the duty cycle, 50 percent duty cycle produces no even harmonics. As to the inductive ring in the waveform, most of that is being produced by the inductive loop of the leads to RC substitution box. Bigger the loop, more the ringing.

For years, I have enjoyed the technical detail of your videos. Well done again and great explanation. Keep up the 'audio' projects and videos.

thank you for the videos and happy hollidays

Thanks for another great video, it’s so inspiring. Keep up the great work you’ve been doing since the first video.

I’m glad to see you’re experimenting with some lovely single ended amps. For a long time, I would go for performance and I would geek over push pull amps. Until I decided to build a simple Single ended amp. The difference was night and day. They sound so much better and natural. After over 20 years of experimenting with tube amps, I guess I finally found the amp for me. An ultra linear single ended el34 amp. I’m using a 4800 / 8 ohm opt I wound myself using the Delatsch Calculator app for iPhone and it’s great. I drive the el34 with half a 5670 and I’m getting 5w rms per ch from 18-19.000Hz at less than 1%thd using a tube regulated power supply (low impedance is critical for bass performance and régulation helps with hum reduction in SE amps) . If I bump b+ all the way up to 285V, I can squeeze an extra watt and a half per ch below 1% thd. But it sounds like a million bucks and I compared it to stuff like the Dynaco ST70 and even some great solid state amps. The transformers have to be oversized or you won’t be able to get down to 18Hz at -3dB. Right now I’m using a 250v power supply and a 250 ohm cathode resistor.

Even though most of this theory is way over my head I always enjoy your explanations and learn a little more. Know you’re working on a stereo version if I read correctly, oh boy. I’d love to pay for that schematic of that when it’s ready I have a few 807’s matched and a variant 6B something pair as well I believe

The reason, according to Floyd Toole and researchers at Harman and elsewhere is Even Harmonics are what occur in nature. Whether a church bell, or banging on a pipe with a hammer, or a trumpet or violin, sounds in nature just naturally amplify Even Harmonics. Maybe there is something in nature that evolutionarily led to a dislike Odd Order Harmonics. For example, I wonder what something like a Big Cat's growl looks like on a frequency spectrum? We have evolved for that sound to really catch our attention. And probably have a visceral negative reaction to it. Now, in the modern day, when we hear brakes screeching, it is not much different than a Big Cat's growl, with the same reaction. So my suspicion is that this preference for Even Harmonics is an Evolutionary trait, and that is something that could be easily researched. But in the meantime, thank Satan for Tube Electronics. They don't just reproduce natural harmonics beautifully. Also, when done well, they create incredible Soundstage, Imaging, a sense of Air and Spaciousness that I have yet to hear equalled by Solid State when comparing at the highest levels. I started in high and HiFi back in the 80s. Only recently have I heard the jaw-dropping, just astounding performance of some of the pure Class A high-powered Transmitter Tube amplifiers. Some of the ones that you have made are a great example. The 805s, 845s, and even some of the massive tubes used by NAT Audio that are as hot and bright as the Sun - you could get a tan - they are utterly amazing. They put the performers in the room, and recreate the room that they're in, like no other amps. I've also started to fall into the rabbit hole of some of the Preamps employing the big signal tubes, and again, there is Soundstage and Imaging, as well as inner detail and resolution, without any harshness or fatigue, that's on a whole other level. So the incredible ability of tubes seems to be never ending. And I'm totally down for that. Bring it on!

At first, I thought this was in reference to a Grommes Little Jewel. I'm not sure which I like better now, haha.

Every video you do is educational Dave. That is great

So 6080 is single or duble triode ? Only one in amp is mono ? Cheers !

I see two exotics FETs on the table close to the desk lamp (FET=field effect tube). Utilizing plate current to represent microphone output intensity is a brilliant way. Merry Christmas 2024 and happy new year 2025, sir.

hi interesting i have a question about negative feedback ,if i put a signal with my generator and i take the square wave at 1khz and if i turn the 10k pot to have the best square wave ,is it a good way to tune the negative feedback .thanks

Wouldn't you want to test with both channels driven for max power - as per McIntosh specs - as opposed to testing each channel individually?

hypothetically - What if you shot a stream of electrons into a Rectifier? or maybe just positive Ions? Wouldn't that instantly change it from electrons to electricity? Like for example if you were to build a massive pyramid and vibrate the heck out of it so it would produce electrons or ions? and then had a Rectifier at the top... Ions have a Pos & a Neg so.....

I have one of these! But it's the LK-72 version. Fully restored, all JJ tubes. It sounds amazing! Anyone that says they don't like the sound of tube amps, or that tube amps do not have enough base, or not tight enough, etc. They have never heard one of these with some good speakers. I can crank this thing up full volume, no hiss, no hum, no audible distortion. But at just over half volume, it will run you out of the room and shake the walls. Great sound. Thanks for the video.

do you ever sell your amps, Kurt AA4KH

Whats the watt difference between the two different tubes 4-400 to 3-500 tubes

Nice

David, thank you for all your informative videos! I realize this video is 8 years old now, but the mystery of output transformer DC resistance discrepancies is explained in this video: kzbin.info/www/bejne/lZ7VlpWtob2ZnMk Executive summary: in transformers, the turns ratio is important. Consider the primary side windings (ignore the ultra linear taps for now) - the two windings on either side of the center tap should have the same number of turns. The one wound on the inside will be a shorter wire than the one wound on the outside, so the one on the inside will have a lower DC resistance than the one wound on the outside. Solution: test the windings with AC! The video explains a safe method using commonly available test equipment. Thanks again!

Thank you David for the many informative videos. I have a quibble with this one. The title says Williamson, but the circuit is not a Williamson. Note that the input stage tube cathodes are tied together with only a 200 ohm resistor in between -- not Williamson. The schematic you show in this video is labeled simply triode circuit. Other schematics in the Acrosound catalog that are Williamson are labeled Williamson, but this one is not. Acrosound used a similar input circuit in the commercial product power amps Ultra-Linear II and St 120, but there are several differences. This arrangement reminds me of a tube differential amplifier. I cannot say I fully understand this circuit. Is this circuit worth exploring further? Thanks again!

Sir, thank you. I love these videos.

Make a *.pdf file for the schematic Hommie. ❤❤. Screen shots & GIFs. Are for pud knockers, Amigo

hi, id like to know if is it normal to have clipping on positive part before negative part with a single ended amp or they are supposed comes equal in the same time on the scope.thanks