Tube Voltage Regulator, Is it worth the effort?
I’ve been experimenting with tube based voltage regulators for my next amp, which will be 2A3 or 300B (switchable) SE stereo amp.
See this link from JC Verdier more details on the tube voltage regulator tech300B
My first impression is it’s not worth the effort…. But I’m still undecided and looking for feedback.
You will see from the test data below that I never got down to the less than 1R z-out as the JC Verdier article indicates….
My B+ for the baseline of this test was based on a LC-RC PSU (10H,60uf-200R,100uf)
If I add 200-400uf to the LC stage of this psu I get roughly the same Z-out numbers as with a VReg psu
Could I hear the difference? I can’t say because I didn’t do a side by side blind test.
Here is the test setup:
The input to the amp is a sine wave at full power, from 20hz to 5Khz
100R in series with the OPT used to measure the AC voltage drop and calculate the AC current
The AC current was very constant from 20hz to 2Khz, at around 40ma
AC voltage measured at the B+ via a 30uf/1M RC filter
Z-out is calculated by = Vac/Iac
The psu for the V-reg test was: 10H,60uf – Vreg – 100uf
Here are the test results (Z-out in ohms) that compare several LC-RC power supplies to the V-Reg power supply:
The first 4 are normal LC-RC types, then a solid state test just for grins, the 2 examples of tube based voltage regulators, the first using a 5V zenner diode for the voltage reference, the 2nd using a 0C3
PSU Config
#1 10H,60uf-200R,100uf
#2 10H,260uf-200R,100uf
#3 10H,460uf-200r,100uf
#4 LM317 (ss) V-reg
#5 6AS7,6SG7,5V zenner
#6 6AS7,6SG7,0C3
20hz 50hz 100hz 200hz 500hz 1Khz 2Khz 5Khz
#1 50.0 29.5 15.6 8.0 3.1 1.5 .96 .84
#2 18.3 9.75 5.11 2.47 1.0 .52 .31 .18
#3 11.2 5.75 2.82 1.44 .58 .31 .18 .12
#4 45 19 10 6 4.2 3.4 3.3 3.3
#5 23.9 18.5 12.8 7.4 3.1 1.4 .57 .15
#6 19.2 14.5 11.8 8.7 3.1 1.3 .51 .21
I’ve been experimenting with tube based voltage regulators for my next amp, which will be 2A3 or 300B (switchable) SE stereo amp.
See this link from JC Verdier more details on the tube voltage regulator tech300B
My first impression is it’s not worth the effort…. But I’m still undecided and looking for feedback.
You will see from the test data below that I never got down to the less than 1R z-out as the JC Verdier article indicates….
My B+ for the baseline of this test was based on a LC-RC PSU (10H,60uf-200R,100uf)
If I add 200-400uf to the LC stage of this psu I get roughly the same Z-out numbers as with a VReg psu
Could I hear the difference? I can’t say because I didn’t do a side by side blind test.
Here is the test setup:
The input to the amp is a sine wave at full power, from 20hz to 5Khz
100R in series with the OPT used to measure the AC voltage drop and calculate the AC current
The AC current was very constant from 20hz to 2Khz, at around 40ma
AC voltage measured at the B+ via a 30uf/1M RC filter
Z-out is calculated by = Vac/Iac
The psu for the V-reg test was: 10H,60uf – Vreg – 100uf
Here are the test results (Z-out in ohms) that compare several LC-RC power supplies to the V-Reg power supply:
The first 4 are normal LC-RC types, then a solid state test just for grins, the 2 examples of tube based voltage regulators, the first using a 5V zenner diode for the voltage reference, the 2nd using a 0C3
PSU Config
#1 10H,60uf-200R,100uf
#2 10H,260uf-200R,100uf
#3 10H,460uf-200r,100uf
#4 LM317 (ss) V-reg
#5 6AS7,6SG7,5V zenner
#6 6AS7,6SG7,0C3
20hz 50hz 100hz 200hz 500hz 1Khz 2Khz 5Khz
#1 50.0 29.5 15.6 8.0 3.1 1.5 .96 .84
#2 18.3 9.75 5.11 2.47 1.0 .52 .31 .18
#3 11.2 5.75 2.82 1.44 .58 .31 .18 .12
#4 45 19 10 6 4.2 3.4 3.3 3.3
#5 23.9 18.5 12.8 7.4 3.1 1.4 .57 .15
#6 19.2 14.5 11.8 8.7 3.1 1.3 .51 .21
That's interesting.
I find that most folks who swear by tube regulated power supplies actually do it for reasons of tradition and sentiment (i.e. psychological bias) rather than for actual / pragmatic reasons. Personally I have had fine results using solid state regulation based on the Maida (augmented 3-terminal device) regulators. I have not measured the output impedance though.
Jan Didden (who usually hangs out in the Solid State and Power Supply areas of the forum) recently authored an interesting article in Elektor magazine that combined a solid state error amp and vacuum tube pass element. It has a sub 0.5 Ohm impedance across the audio band. (Elektor March 09, page 22, "T-Reg".)
I find that most folks who swear by tube regulated power supplies actually do it for reasons of tradition and sentiment (i.e. psychological bias) rather than for actual / pragmatic reasons. Personally I have had fine results using solid state regulation based on the Maida (augmented 3-terminal device) regulators. I have not measured the output impedance though.
Jan Didden (who usually hangs out in the Solid State and Power Supply areas of the forum) recently authored an interesting article in Elektor magazine that combined a solid state error amp and vacuum tube pass element. It has a sub 0.5 Ohm impedance across the audio band. (Elektor March 09, page 22, "T-Reg".)
Actually, it's sub-100 milliohms Zout and sub-300 uV hum/noise
http://www.linearaudio.nl/t-reg-1.htm
jd
Still, I like them.
If we were interested in perfection though, we wouldn't be using tubes at all for anything
Cheers!
"And they look really nice"
: )
Actually, it's sub-100 milliohms Zout and sub-300 uV hum/noise
http://www.linearaudio.nl/t-reg-1.htm
ok! thank ! I change tube 6080 by 6H13??? what do you thinK???
thank,
Check out this link.... this is what got me started on the Tube Based V regulator test.
Triode Spirit
Last edited:
Actually, it's sub-100 milliohms Zout and sub-300 uV hum/noise
http://www.linearaudio.nl/t-reg-1.htm
ok! thank ! I change tube 6080 by 6H13??? what do you thinK???
thank,
That'll work fine, looking at the specs. I haven't checked the pin-out though, so that's something you'd need to verify.
jd
Ok! And filter before your cricuit ??
C-R-C/ C=?, R+= ?? the value of C, R is enough, And is good filter??
In your cricuit DN2540 is wrong foot. What do you think??? yes or no??
See in the image
The plot on your website shows that that is predominately hum (and the same for all three plots, despite signifcantly different loading??).
How much ripple was present at the input?
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Glen,
I didn't measure the ripple but used a 100uF cap after the rectifier as a 'typical value'.
That hum is also partly caused by the aux supply for the error amp, which is constant with load of course. But at sub-mV values at the output, I wasn't going to spend anymore effort at that.
jd
Audio Research and Berning come immediately to mind. I'm sure there's more. But even if something is superior, it doesn't follow that it's fashionable, and fashion is what drives commercial equipment.
I would really like to do the the old fashon way:
6AS7 series regulator
6SJ7 or 6SG7 pentode
0C3 voltage reg (OK, maybe a zenner diode here)
Yes, It's lot more work, but it's a hobby!
I've been a proponent of voltage regulation in tube gear for close to 25yrs now, and in fact just about all of my designs use it. I generally prefer tube based regulators for reliability, but a Maida based on the LM317 or comparable LT monolithic can be reliable and work well with just a little care.
I think one of the things that scares people off is the added cost and complexity - also in so far as effective tube regulator design is concerned there is some art involved. (well actually, really good engineering)
My most successful power amplifier designs all featured regulated supplies. The business itself was overall not too successful..
Every tube component in my current system features a regulated tube based psu, in most cases on a separate chassis.
I think one of the things that scares people off is the added cost and complexity - also in so far as effective tube regulator design is concerned there is some art involved. (well actually, really good engineering)
My most successful power amplifier designs all featured regulated supplies. The business itself was overall not too successful..
Every tube component in my current system features a regulated tube based psu, in most cases on a separate chassis.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.