If that is your goal, just use a SS amp because a tube amp at 0.08%THD won't sound any different.
It's pretty simple to understand: why bother with using inefficient, expensive, and fragile vacuum tubes if the goal is 0.08% THD? With that low a THD number, the tube amp is going to sound no different than a much more efficient Solid Sate amp will sound.
A little bit of second order harmonic distortion is what makes a tube amp sound like a tube amp.
A little bit of second order harmonic distortion is what makes a tube amp sound like a tube amp.
you brought up a well-known myth.
the truth is different:
if such a low THD is achieved at 40-70 dB Negative feedback - then yes - it will be a "transistor" sound
but if a THD of 0.028% is achieved at 4-6 dB Negative feedback - it is an incredible tube sound!!!
but it is important to correctly configure the amplifier using a spectrum analyzer
but if you do this, you will feel with your skin what a concert in your own home is like!
I`m pretty sure that your simple but high quality DIY tube amp will outperform sonicaly way so many conventional solid state AB class amps .
Actually the TRUTH is humans can't detect 1% vs 0.1% THD in A/B listening tests. Most people can't hear even 5% THD.
And no, it's not a myth that second order harmonic distortion sounds pleasing and is why single ended tube amps like a 300B sound so good.
@Svytoslav I like the version you linked in the YT video with the regulated screens, but a LM317 Maida regulator will do a much better job of regulation than the zener diode. Often you can also regulate the power supply of the first 6H9C VAS single ended triode as well (with the same regulator), with further improvements in sound quality.
I'll tell you a secret.
Stabilizer does not stabilize anything! (because it is connected in series)
it "cuts off" the supply volts of the second grid to the level when TND become minimal (it is controlled on the screen of the spectroanalyzer).
I'll tell you more - with a resistor instead of a stabilizer in series - TND is much more !
PS: there were great sounding singing tube amplifiers - but as soon as you look spectroanalyzer - and at a frequency of 40 Hz at this LFM - 3-5-7 % TND (the video shows the whole process without secrecy).
HORROR!!!!
Simple manipulations - and the same amplifier already has 0.4% TND at 40 Hz.
PS: A current source will only worsen the sound in a tube amplifier.
Translated with DeepL.com (free version)
I know what you're talking about.
but I'll say the same thing from a slightly different angle.
the nature of sound is not fully understood. and the fact that the secret of tube sound is in the 2nd harmonic is a very simplistic attempt to explain the whole nature of tube sound.
Harmonic distortion at 0.028% is certainly not the main or primary goal.
BUT!
if in one and the same amplifier by simple manipulations it is possible to reduce distortion by times (without increasing the feedback depth!!!) - then why should it not be done?!!
Let me give you an example: the more accurately the guitar is tuned - the more beautiful the composition sounds. Yes - you can tune a mediocre master - quickly - it is one level. but when a real maestro tunes the guitar for himself - it will be a completely different level of sound.
similarly with an amplifier
PS: I suggest this interpretation of "well-tuned amplifier" - if your wife does not say "turn down the volume" or "turn off the amplifier" - it is already a well-tuned amplifier. But if your wife says "turn it up", "I want to listen to 2 more songs" - you can safely consider that you have reached the pinnacle of tube amplifier tuning 🙂)))
Translated with DeepL.com (free version)
a good output transformer is a prerequisite for quality sound
but!
if you don't make fine and precise adjustments, you will get results - but what kind of results?
1. lamps should be selected according to maximum closeness of characteristics at the current level that will be in the real circuit
2. after the phase inverter - everything should be as much identical as possible. for example - if after the phase inverter one capacitor is 0,21 µf and another 0,23 µf (instead of both 0,22 µf) - all attempts to get nonlinear distortion of 0,028% will be unsuccessful.
3. The depth of feedback should not exceed 4-6 dB. It is possible to introduce a feedback of 40-60-70% and get a distortion level of even 0.001% at a volume of 5 watts - but it will be a "dead" "transistor" sound. a small positive feedback is also useful - but this is a separate story.
4. the output transformer should be with reserve. For example - if the amplifier has an output power of 20 watts - it will be great if the output transformer will have an overall power of 80 watts. If you tune according to a spectroanalyzer - the best results are obtained with torroidal output transformers.
5. it is very desirable to use a choke to filter the anode voltage. i think it is clear that a quality tube amplifier for yourself is only in Class A (no AB or B).
6. and the most important thing - accurate tuning according to the spectroanalyzer -
If you have to replace the lamp, please be kind enough to readjust it!
It's like balancing tires at a car service station.
THE RESULT WILL SURPRISE YOU!)
Translated with DeepL.com (free version)
I recommend that anyone with a push-pull pentode (tetrode) amp hear it as a pentode for the first time with fixed screen voltage (G2). Start with the zener diode, since it is simple and cheap. The Maida regulator can be constructed very inexpensively on perf board. The C1 and C2 1.0uF mylar are very important, so keep those as they are. You can reduce or even eliminate R3 (it limits current). There are LM317 calculators online to calculate voltage. Do not run the screens at the maximum datasheet voltage. Note that R6 will dissipate some heat, so I use wire wound, which also have a more stable resistance (TCR). I use a trimmer pot parallel to R6, to trim the voltage. Distortion lowers quite a lot with fixed screen voltage, and the bias current on the output tubes can be reduced. You can use a single MOSFET as the pass device instead of the Darlington pair.
When I listen to a tube amp, it is my impression that the staging is different. It sounds like the complex analogue musical signal has been unpacked more efficiently so that the musical sources, that were separated during the recording, are still separated in the reproduction. I presume it is due to phases of waves at different frequencies.
I think it is down to the parts count of a typical tube amp more than anything else; the less that is in the way of the reproduction the better.
I think it is down to the parts count of a typical tube amp more than anything else; the less that is in the way of the reproduction the better.
Maybe something lost in translation. This is your schematic . Correct? It is good. Just not as good as LM317.
IMHO no need for any extra stabilized screen grids DC supply source including zenner diodes , what is important is to keep screen DC source supply lines internal impedances very low and clean ,
for example here`s basic schematic which for long time work very good for me supplying both B+for anode supply and half B+ for screen supply for PP pentode connected tubes , that very old Philips PSU system works very good for EL84 (up to 400V/200V) , PL36 (340V/170V) , GU50 (500V/250V) , KT88 & 6550 (600V/300V) , EL34 (up to 800V/400V) , 813 (1500V/750V) , ...
for example here`s basic schematic which for long time work very good for me supplying both B+for anode supply and half B+ for screen supply for PP pentode connected tubes , that very old Philips PSU system works very good for EL84 (up to 400V/200V) , PL36 (340V/170V) , GU50 (500V/250V) , KT88 & 6550 (600V/300V) , EL34 (up to 800V/400V) , 813 (1500V/750V) , ...
Attachments
Last edited:
LM317: low impedence, yes. Clean, yes. >60dB PSRR at 120Hz.
I have also used 6DR7 based regulator, but the LM317 is much better.
I have also used 6DR7 based regulator, but the LM317 is much better.
Last edited: