Tayda has some anti-log 20k pots if you don't mind having a counter-clockwise volume control 🙂
https://www.taydaelectronics.com/c20k-ohm-anti-log-dual-taper-potentiometer.html
https://www.taydaelectronics.com/c20k-ohm-anti-log-dual-taper-potentiometer.html
50k Log will also work well enough in the NFB version but losing 0.7dB gain and some extension (F3 from 520kHz down to 470kHz) vs the ~20k Log class potentiometers. No big deal.
That "NFB ver" is pretty much the same as what I drew up for post #4,072, except yours employs more NFB than mine.
The opposite-facing zener diodes are a simple voltage clamp to protect the inputs of solid state devices downstream (I'm enjoying a pair of JBL active speakers these days).
I used a 51k series resistor to the grid anticipating a 20k pot as well. Or a 10k pot. My sources are either solid state or have buffered outputs. No phono preamps with the output taken from a 12AX7 plate.
If higher Zin is required, the R1/R2 voltage divider resistance values can be scaled up. The penalty there is if R2 is a very high value it might add noise, and the high frequency response may droop a little due to the tube's input capacitance.
It's all a balancing act of offsetting compromises. That's the fun of it, right?
The opposite-facing zener diodes are a simple voltage clamp to protect the inputs of solid state devices downstream (I'm enjoying a pair of JBL active speakers these days).
I used a 51k series resistor to the grid anticipating a 20k pot as well. Or a 10k pot. My sources are either solid state or have buffered outputs. No phono preamps with the output taken from a 12AX7 plate.
If higher Zin is required, the R1/R2 voltage divider resistance values can be scaled up. The penalty there is if R2 is a very high value it might add noise, and the high frequency response may droop a little due to the tube's input capacitance.
It's all a balancing act of offsetting compromises. That's the fun of it, right?
Goofing around with the NFB version in LTspice, I see that the THD isn't reduced all that much by the NFB.
The version without NFB makes about 0.125% THD with 1V out into 100k ohms.
The version with R2 = 51k and R1 = 200k makes 0.047% THD.
That's because there's not a lot of gain to apply as NFB, so even with the closed loop gain knocked down to a bit more than 2X, there's still appreciable THD.
Maybe the NFB version doesn't lose the magic sound?
I don't know. I'll have to build it.
The version without NFB makes about 0.125% THD with 1V out into 100k ohms.
The version with R2 = 51k and R1 = 200k makes 0.047% THD.
That's because there's not a lot of gain to apply as NFB, so even with the closed loop gain knocked down to a bit more than 2X, there's still appreciable THD.
Maybe the NFB version doesn't lose the magic sound?
I don't know. I'll have to build it.
Depending on synchronization of heaters coming up with the B+ ramp up you may experience large longish near DC cycles on the output during on/off. As the plate voltage bias slowly goes up until ready, or down until completely off, so the output capacitor. The Zener clamps may prove a practical enough solution. If not, I suggest a preamp output delay relay solution with fast off release.
Without such measures no power stage without OPT or DC output protection should be switched on before or turned off after switching this preamp.
It is moderate NFB application to roughly cut Zout and THD in half. Not extreme, but still a good deal. Should remain palatable. There isn't much of boiler steam available to spend anyway. As you say.
It was figure of speech when trying to find the 20 kOhm Alps RK27. Plastic axle low cost stuff is below the minimum standard and 24 steps won't cut the cake 🙂
Then I tried to find 18x20 mm 4700 µF 16V of a good brand and series. Oh boy, give it a few months and we can't buy socks and underwear anymore.
Last edited:
If you don't mind spending I suggest the TKD over the Alps. Hi-Fi Collective carries it in many values and in a motorised version for remote control. That's the better VFM type as well.
Hi I recall a story by Bruno Putzeys that TKD potentiometers distorted. Then someone that used them told me the pin to resistance pad connections go bad. I haven't tried them because of that or at least I don't recall using them. Once I owned a Khozmo and that one was superb. At a price but quality pays off. Many stepped attenuators of high price also scratch after a while and one really needs a 48 step version. Relay based attenuators really are top league but they're bulky. I guess the RK27 is not so bad after all 🙂 It is now meant to replace a practically new low cost potentiometer that started to scratch... all belonging to the price one pays for going back to analog stuff.
So parts are hard(er) to find and when we find them at a too high price we should buy 2 as they have the longevity of milk. It all irritates as DIY already was expensive but today it is becoming insane to DIY (costwise).
http://www.brunoputzeys.be/r4.php?item=25
So parts are hard(er) to find and when we find them at a too high price we should buy 2 as they have the longevity of milk. It all irritates as DIY already was expensive but today it is becoming insane to DIY (costwise).
http://www.brunoputzeys.be/r4.php?item=25
Last edited:
I once measured a friend's DCG3 FET preamp with 20k Log TKD CP-2511 in it and it could photofinish a MOTU M2 audio interface's self loop for THD so it should be good enough for tubes.
DCG3 does three zeroes after decimal through higher resolution test gear but I just had the M2 when the particular build was checked. Will find that FFT to post it here too.
DCG3 does three zeroes after decimal through higher resolution test gear but I just had the M2 when the particular build was checked. Will find that FFT to post it here too.
No adjustments for employimg a 10k potentiometer, just your sources not to be taxed by it. If they have modern low output impedance no problem.
I've decided that every source should have either a cathode follower or a source follower buffer on its output, so it can be connected to all the cool new toys like powered speakers and class D amps. A 10k pot also makes a very kind grid leak for the first stage of a tube amp.
The only downside would be that the output cap on the source may have to be a high value to pass well below 20Hz signals.
For a phono preamp that's not a big deal, as there's nothing useful down there anyway.
But for a DAC, you might want to pass 20Hz with no attenuation. You would need at least 2.2uF into that 10k pot, right?
The only downside would be that the output cap on the source may have to be a high value to pass well below 20Hz signals.
For a phono preamp that's not a big deal, as there's nothing useful down there anyway.
But for a DAC, you might want to pass 20Hz with no attenuation. You would need at least 2.2uF into that 10k pot, right?
DC coupled for the hardcore guys. No cap = the best cap. 10 kOhm is what I bought today. No issue for normal sources and an input RC filter may up that value a little if desired. Two birds with one stone.
The TKD CP-2511 seems OK (it should be at that price) so what type did Bruno refer to?! Was it in static situation or when turning it?! What value? Older series maybe?! I don't know.
The TKD CP-2511 seems OK (it should be at that price) so what type did Bruno refer to?! Was it in static situation or when turning it?! What value? Older series maybe?! I don't know.
Last edited: