Why these are not commonly used in tube amps

Although the discussion, so far, has focused on power amplifiers, tube preamplifiers can also benefit from regulated power supplies. The regulated power supply I built for my Heathkit WA-P2 is an example that can easily be adapted to a power amplifier by increasing the size of the pass transistor heat sink. See post #9 in the following thread: http://audiokarma.org/forums/index.php?threads/heathkit-preamplifier-model-wa-p2-–-power-supply-–-what-are-you-using.721788/#post-9731055
Nice design. This is your circuit:

TCdriver regulator.JPG

I put in some numbers so if anyone care to look deeper into how this regulator works. It's clever, Q1, D1 and Q6 replace a whole stack of zener and provide reference voltage with only about 1mA consumption. This is much lower than if you use zener.

People have different ideas and we can all learn if we just take a minute to study through other's designs. This make regulated supply easier. I just don't like the stack of zener and the current it takes to stabilize the voltage. This one relies of R1, R2, R3, R4 and R7 with D1 to set the output voltage.

Nice circuit. I take note on this.




One note of caution using regulated supply in general, you have to watch out power dissipation of the MOSFET, line voltage goes up and down, so is the raw rectified voltage. When you regulate the voltage, the change of raw DC might put extra power dissipation on the MOSFET. eg. if the raw DC goes up 10V, the power dissipation on the MOSFET will increase by 10V X I. For power amp, this might be very high.. That's the reason I apt for kind of pseudo regulated, more getting rid of the ripple by let the output voltage floats with the raw average voltage. I minimize the power dissipation of the MOSFET.
 
Last edited:
... I just don't like the stack of zener and the current it takes to stabilize the voltage.
I agree with you there. Zener diodes change voltage drop quite a bit with temperature which can make setting the output voltage imprecise and subject to drift. In the regulator design shown, the 1N4568 (D1) is a Zero-TC Reference Diode, not an ordinary zener.

One note of caution using regulated supply in general, you have to watch out power dissipation of the MOSFET, line voltage goes up and down, so is the raw rectified voltage. When you regulate the voltage, the change of raw DC might put extra power dissipation on the MOSFET. eg. if the raw DC goes up 10V, the power dissipation on the MOSFET will increase by 10V X I. For power amp, this might be very high.. That's the reason I apt for kind of pseudo regulated, more getting rid of the ripple by let the output voltage floats with the raw average voltage. I minimize the power dissipation of the MOSFET.
That is quite correct. Power dissipation with higher currents, used in a power amplifier, will require a well sized heatsink, but, that is the price one must pay, if you want a regulated power supply voltage. In my preamaplifier application, the current draw is typically 8mA, so, heating is not an issue.
 
Very interesting read. Too bad Mr. Kegger passed away, really like to ask him his reasoning in the feedback. Sounds like it has something to do with speaker back EMF from a member. I definitely going to try it on the wood board experimental amp.

Another thing I notice is you use pentodes as driver tube. You want to lower input capacitance as pentode literally is like a cascode stage that eliminates miller capacitance that load the input.

I notice the LED at the cathode of the driver tube, do you know what is the reason? My guess is it enhance even harmonics and make the sound sweeter.
 
Last edited:
I agree with you there. Zener diodes change voltage drop quite a bit with temperature which can make setting the output voltage imprecise and subject to drift. In the regulator design shown, the 1N4568 (D1) is a Zero-TC Reference Diode, not an ordinary zener.

That is quite correct. Power dissipation with higher currents, used in a power amplifier, will require a well sized heatsink, but, that is the price one must pay, if you want a regulated power supply voltage. In my preamaplifier application, the current draw is typically 8mA, so, heating is not an issue.
Yes, it's all about the heatsink. I am doing a SE amp, I don't need true regulated supply as it's class A. more importantly, I need one with no ripple as SE amp doesn't cancel ripple. That's the reason I came up with that circuit, I don't need heatsink, just bolt onto the chassis and it should be good enough. It's only about 5W or less, just simple, cheap and dirty!!!!

1N4568 is a nice diode, only 0.5mA. Good finding. You did some homework on this design.:thumbsup:


I really think tube amps can benefit greatly using these kind of regulated supplies. The big advantage of tube amps are low operating current. Even the highest practical power type of amps use less than 1A, it's manageable. These just not very practical for SS amp. I layout my circuit into the filter pcb of my SS amp and had to abandon it because it just got too hot. this is a match in heaven for tube amps.
 
Last edited:
I found it!!!! This is what Mr. Kegger address the local feedback directly:http://www.audiokarma.org/forums/index.php?threads/good-schematic-for-se-el34-kt88-amp.196097/page-2 . :banana::banana::banana:

It's the first post on the top of page two. I was right, he basically form a local feedback from plate of the power tube back to it's grid, but going through on the high voltage side. Maybe doing this to avoid using coupling cap, it's all DC feedback.

This is interesting circuit.
View attachment 1017405

I have been looking at this circuit, this is how I look at it.

R3 is the feedback resistor, R17 is the gain resistor. the gain is governed by R3/R17. But both the preamp tube and OPT are out of the feedback loop. It is really a single stage feedback where the plate of the KT88 feeding back to it's own grid. It's more like Miller compensation to linearize the power tube alone.

man!!! I was searching on this one.
 
Last edited:
I agree with you there. Zener diodes change voltage drop quite a bit with temperature which can make setting the output voltage imprecise and subject to drift. In the regulator design shown, the 1N4568 (D1) is a Zero-TC Reference Diode, not an ordinary zener.

That is quite correct. Power dissipation with higher currents, used in a power amplifier, will require a well sized heatsink, but, that is the price one must pay, if you want a regulated power supply voltage. In my preamaplifier application, the current draw is typically 8mA, so, heating is not an issue.

Actually, you can make this to an adjustable regulated supply by either adjusting R7 or R3 & R4. This design is current driven, voltage is developed across R3 & R4. you change the value of the resistors, you change the output voltage.

I designed my adjustable supply, not really regulated. This is what I have, I really don't care about regulated and stable voltage, this is for power scaling for guitar amps in my design. This is a working circuit, I put this in both guitar amps I designed. Nothing fancy, but it works like a champ. I even have a small inverting amp that track the B+ and produce the -ve grid bias voltage that is proportional to the B+.

Power scaling1.JPG
 
Actually, you can make this to an adjustable regulated supply by either adjusting R7 or R3 & R4. This design is current driven, voltage is developed across R3 & R4. you change the value of the resistors, you change the output voltage.
Yes, you are correct. In my last iteration I made R7 variable.

Updated power supply WAP2PS03b:
index.php

This is the WAP2PS03b. There are a few changes compared to the WAP2PS03. The SMPS is now a MP1584, the reference is a LT1004-1.2 and a variable resistor is now used to set the output voltage.
 
Back
Top Bottom