Paczeltf/Ferenc,
thanks for starting this very interesting thread. You mentioned somewhere that you posted a spreadsheet, so I registered for this forum to see it, but I still haven't found the spreadsheet. Is there a special area for files of common interest?
I'm not really planning to make any mods, at least not right away, but I found your analysis of the bias correction and over current circuit very interesting, ditto for the bridge rectifier dimensioning, and comments on various cap sizes etc.
I ordered my kit directly from Victor (shoutout to Victor, who stayed at the computer beyond midnight to answer my questions and take my order!). Very excited about getting back into soldering!
Different topic, but still about the TU-8200R, and the various mods and upgrades:
a) is capacitor microphony (I assume due to parametric amplification/modulation via vibrational changes in layer separation) really a measurable effect in amplifiers?
I found an IEEE paper which finds quite small effects for 100nF ceramic and PPS caps (larger for ceramic): "...noise currents and voltages of approximately 16 nA and 10 μV, respectively, were measured when the circuit board was vibrated, ...". (see https://ieeexplore.ieee.org/document/1032708, R. Nelson and L. Davidson, "Electrical noise generated from the microphonic effect in capacitors," 2002 IEEE International Symposium on Electromagnetic Compatibility, Minneapolis, MN, USA, 2002, pp. 855-860 vol.2, doi: 10.1109/ISEMC.2002.1032708.) That can't possibly lead to audible effects, except maybe in MM/MC input stages?
10 microVolt is 100 dB down from 1 V, which I believe is the typical order of amplifier input voltages (line level).
I do not doubt that some caps are better than others, but I'm not sure I'm ready to believe there are audio/electroacoustic specific considerations (unless maybe a phono amplifier is directly coupled to a transducer/speaker).
b) Lundahl output transformer upgrade: my guess/hunch is that most of the improvement is simply due to different physical parameters correlated with larger size and thus power handling capacity. Is that correct, or are there certain je-ne-sais-quoi effects, more diffusely related to the specific winding technique and configuration (Lundahl mentions this on their website), and overall care and quality control?
Stefan/Gruesome
thanks for starting this very interesting thread. You mentioned somewhere that you posted a spreadsheet, so I registered for this forum to see it, but I still haven't found the spreadsheet. Is there a special area for files of common interest?
I'm not really planning to make any mods, at least not right away, but I found your analysis of the bias correction and over current circuit very interesting, ditto for the bridge rectifier dimensioning, and comments on various cap sizes etc.
I ordered my kit directly from Victor (shoutout to Victor, who stayed at the computer beyond midnight to answer my questions and take my order!). Very excited about getting back into soldering!
Different topic, but still about the TU-8200R, and the various mods and upgrades:
a) is capacitor microphony (I assume due to parametric amplification/modulation via vibrational changes in layer separation) really a measurable effect in amplifiers?
I found an IEEE paper which finds quite small effects for 100nF ceramic and PPS caps (larger for ceramic): "...noise currents and voltages of approximately 16 nA and 10 μV, respectively, were measured when the circuit board was vibrated, ...". (see https://ieeexplore.ieee.org/document/1032708, R. Nelson and L. Davidson, "Electrical noise generated from the microphonic effect in capacitors," 2002 IEEE International Symposium on Electromagnetic Compatibility, Minneapolis, MN, USA, 2002, pp. 855-860 vol.2, doi: 10.1109/ISEMC.2002.1032708.) That can't possibly lead to audible effects, except maybe in MM/MC input stages?
10 microVolt is 100 dB down from 1 V, which I believe is the typical order of amplifier input voltages (line level).
I do not doubt that some caps are better than others, but I'm not sure I'm ready to believe there are audio/electroacoustic specific considerations (unless maybe a phono amplifier is directly coupled to a transducer/speaker).
b) Lundahl output transformer upgrade: my guess/hunch is that most of the improvement is simply due to different physical parameters correlated with larger size and thus power handling capacity. Is that correct, or are there certain je-ne-sais-quoi effects, more diffusely related to the specific winding technique and configuration (Lundahl mentions this on their website), and overall care and quality control?
Stefan/Gruesome
Last edited:
Hi Stefan,
The spreadsheet was attached to the thread initially, but early versions of TU-8200R manuals had incorrect part designation with a couple of duplicates, which got fixed later. One member got confused and made big noise so Victor deleted it. Those who are interested can contact me directly.
a) It is the other way around. DC voltage might be affected by resonances and generate noise.
DC levels in preamps are not significantly lower than those in power amps, but input signal levels can be -60-80 dB lower especially at low frequencies and are more sensitive to noise contamination.
b) The 8200 Lundahls are not that much bigger than the stock OPTs due to space limitations. They have different core structure though with winding difference.
The spreadsheet was attached to the thread initially, but early versions of TU-8200R manuals had incorrect part designation with a couple of duplicates, which got fixed later. One member got confused and made big noise so Victor deleted it. Those who are interested can contact me directly.
a) It is the other way around. DC voltage might be affected by resonances and generate noise.
DC levels in preamps are not significantly lower than those in power amps, but input signal levels can be -60-80 dB lower especially at low frequencies and are more sensitive to noise contamination.
b) The 8200 Lundahls are not that much bigger than the stock OPTs due to space limitations. They have different core structure though with winding difference.
Hi Ferenc (and everybody else),
I sent you a PM regarding your modification spreadsheet. Could you send it to me, please? You also mentioned rebuilding the driver stage of this amp; is that part of the spreadsheet? I would be curious about those mods as well.
And thanks for answering above. I don't understand though what you mean by 'the other way around'? My question was whether capacitor microphony effects could be measurable in this amp at all, given the low voltages (microvolt) involved. Do you mean they can, but specifically not (as I was suggesting) in phono input stages (which this amp doesn't have)? Or what is 'the other way around'? Mechanical resonances would be excited electrically? But that would show up in the capacitance vs frequency behavior of the capacitor, no? I.e. it would be a generally bad capacitor if it behaved like that, not just for audio purposes.
And some different questions: While comparing the original TU-8200 and the TU-8200R schematics, and also looking at http://www.valveradio.net/audio/ultra-low-distortion-valve-hybrid-amplifier.html, a few questions popped into my mind.
Related to what I asked Victor earlier on the TU-8200R thread about grounding and the AC input, would there be a potential benefit to replacing the IEC mains socket with one with built-in switch and EMI filter? I was thinking https://www.digikey.com/en/products/detail/schaffner-emc-inc/fn283e-6-06/1997114 or similar. This would avoid running unfiltered AC through the amplifier chassis, and minimize AC wire/antenna length inside.
The ultra low distortion tube amp web page above suggests putting a large bypass capacitor between the output tube screen grid and cathode. Could that be done in the TU-8200(R), and would there be any potential benefit to it?
Would there be any point in cutting the PCB AC heater traces HAC1 and HAC2 for the output tubes near their origin, and replace them with twisted wire pairs (same goal of minimizing AC wire length inside the chassis) ?
Have you had a look at the thyristor bias protection circuit in the TU-8200R (at least that is what I think it is)? Does this now remedy the concerns you mentioned about the response time of the protection circuit?
And last, but not least, has anybody measured noise and distortion for this amplifier? Or is this a complete no-no in this forum, or for this type of tube amp? (Like, better not to know, to not spoil the enjoyment....).
I sent you a PM regarding your modification spreadsheet. Could you send it to me, please? You also mentioned rebuilding the driver stage of this amp; is that part of the spreadsheet? I would be curious about those mods as well.
And thanks for answering above. I don't understand though what you mean by 'the other way around'? My question was whether capacitor microphony effects could be measurable in this amp at all, given the low voltages (microvolt) involved. Do you mean they can, but specifically not (as I was suggesting) in phono input stages (which this amp doesn't have)? Or what is 'the other way around'? Mechanical resonances would be excited electrically? But that would show up in the capacitance vs frequency behavior of the capacitor, no? I.e. it would be a generally bad capacitor if it behaved like that, not just for audio purposes.
And some different questions: While comparing the original TU-8200 and the TU-8200R schematics, and also looking at http://www.valveradio.net/audio/ultra-low-distortion-valve-hybrid-amplifier.html, a few questions popped into my mind.
Related to what I asked Victor earlier on the TU-8200R thread about grounding and the AC input, would there be a potential benefit to replacing the IEC mains socket with one with built-in switch and EMI filter? I was thinking https://www.digikey.com/en/products/detail/schaffner-emc-inc/fn283e-6-06/1997114 or similar. This would avoid running unfiltered AC through the amplifier chassis, and minimize AC wire/antenna length inside.
The ultra low distortion tube amp web page above suggests putting a large bypass capacitor between the output tube screen grid and cathode. Could that be done in the TU-8200(R), and would there be any potential benefit to it?
Would there be any point in cutting the PCB AC heater traces HAC1 and HAC2 for the output tubes near their origin, and replace them with twisted wire pairs (same goal of minimizing AC wire length inside the chassis) ?
Have you had a look at the thyristor bias protection circuit in the TU-8200R (at least that is what I think it is)? Does this now remedy the concerns you mentioned about the response time of the protection circuit?
And last, but not least, has anybody measured noise and distortion for this amplifier? Or is this a complete no-no in this forum, or for this type of tube amp? (Like, better not to know, to not spoil the enjoyment....).
This part is probably too deep (43 mm) for the available space between rear wall and output transformer in the TU-8200R. In any case, it's better to measure the standard performance first, before embarking on 'improvements'.
Answering my own question: probably only in pentode mode, otherwise that cap would short part of the output signal (mitigated by the screen grid PTC resistor).
I have sent you the spreadsheets for both amps so you can compare them. They do not include the driver update, but it was explained in this thread earlier.
Your assumption was that preamps are less sensitive to capacitor microphony. I doubt this effect can be measured in any tube amp though, because tube microphony is magnitudes higher and will mask the results.
I don't like the unshielded wire connected power toggle switch either. A pushbutton mechanically connected to the back would be a better solution.
Not in this amp. I have concerns with that approach, which works in pentode mode only anyway.
I do not like irreversible modifications with little potential benefit.
It is certainly much faster, but I can't confirm results without detailed testing.
There are measurements available for TU-8200 on one website, which is mostly about photography. Typical SET performance, why would it be any different? The bandwidth is reasonably good though. The "R" version has the same core design, you can expect very similar results.
Thanks for sending the spreadsheets, Ferenc!
That was a misunderstanding, regarding the microphony. You understood it exactly the opposite way from what I meant. What I was trying to say is that given the very small voltages generated by microphony in capacitors, this effect would more likely be visible in (high sensitivity) preamps. If it is at all visible. So you agree that capacitor microphony is not a reason to install a Mundorf coupling cap.
I found Ken Rockwell's measurements, and also tried to measure frequency response and distortion acoustically, with a speaker and a Umik-2 microphone, myself. (Posted at ASR. I'm not sure I should cross post here.) I can see that the TU-8200R adds a bit (order half a percent) of distortion (2nd harmonic mostly) when driven to produce above 80 dB SPL with an 86 dB sensitivity speaker, so at about 0.5 W.
I would like to push these measurements a bit further, but I assume one has to be careful not to ask for too much output power with a sine wave at the input (or in general)? Will the bias supply protection kick in, or will something in the output tubes or the bias supply section burn out?
The power inlet/switch/filter part I posted will unfortunately not fit into the TU-8200R, at least not in the same position and orientation as the existing IEC socket. There is not enough space between the rear wall and the output transformer.
Regarding measurements: Since this is the DIY forum, maybe this is a good place to ask about instrumentation: I'm thinking about buying a scope, and came across the Digilent 'Analog Discovery' series of all-in-one USB scopes/signal generators/logic analyzers/network analyzers. Low bandwidth (25 MHz), low bit depth (at most 14, I think), but they can do a little bit of everything, and Digilent is supposedly not a bad company for measuring equipment. Has anybody looked at that? There are of course cheaper USB scopes, or even standalone scopes, or one could could buy used higher bandwidth (though not needed for audio) Tektronix etc. stuff. But in general even the higher end used devices will be nowhere near the 24 bit SNR one would need to make competitive noise and distortion measurements on modern audio amplifiers. So, maybe a 25 MHz 14 bit scope/analyzer is good enough?
Does anybody here have experience with the 'Analog Discovery' devices? Or maybe some other recommendation?