Tube DIY Asylum

Do It Yourself (DIY) paradise for tube and SET project builders.

Return to Tube DIY Asylum

Message Sort: Post Order or Asylum Reverse Threaded

Why Build a Regulated Tube Power Supply?

173.51.68.193
Posted on July 10, 2016 at 16:29:37
megasat16
Audiophile

Posts: 207
Location: SoCal
Joined: April 15, 2015
 When you can use a simple LC / CR / LCR type power supply for a DHT preamp supply?

Won't the additional tubes in the regulation circuit adds additional tube noise?

My take on this is a well designed LCR or LC or CR power supply with very good regulation will have a minimal ripple and voltage will not sag under load (for tube preamp). I understand the benefits of regulated power supply for the Amps but for preamps, is there any reasons to add additional tube noise to the B+ supply?
.
.
.Thou shall not stand where I type for I carry a bottle of Certified Audiophile Air and a Pure Silver Whip.

 

Hide full thread outline!
    ...
There is no reason anymore to have tube regulated supply or tube rectified supply, posted on July 10, 2016 at 19:17:38
Michael Samra
Dealer

Posts: 36118
Location: saginaw michigan
Joined: January 30, 2005
 in this day and age for audio.. As far as tube rush getting into the audio path at 120cps,that isn't going to happen,at least where you would hear it.
I do use tube regulation and rectification in a my larger AM transmitters that I run on 40m and 75m phone.
"For every complex problem there is an answer that is clear, simple, and wrong" H. L. Mencken

 

RE: Why Build a Regulated Tube Power Supply?, posted on July 10, 2016 at 20:44:06
Chip647
Audiophile

Posts: 2652
Location: The South
Joined: December 24, 2012
 You design to an objective. If that objective cannot be met with simple solutions, they then get more complicated. A stiff power supply on an Class AB amp can yield benefits and there are more than one way to achieve a stiff power supply. Big high capacity chokes are heavy and costly. Super quiet power supplies for preamps are much easier to achieve with conventional approaches as there are high value, high voltage caps available. Regulators are problem solvers, that is all. The biggest problem they solve is manufacturing cost.

 

Is it a question of tubes or a question of regulation per se?, posted on July 11, 2016 at 07:33:07
Lew
Audiophile

Posts: 10912
Location: Bethesda, Maryland
Joined: December 11, 2000
 There does seem to be some sentiment here lately to the effect that regulation itself is to be avoided when possible and practical. For example, in the output tube power supply. However, lately I have also seen posts to the effect that VR tubes are not a "nice" way to regulate, when regulation is beneficial. There are some very good solid state based regulation schemes, if it comes to that.

 

VR Tubes, posted on July 11, 2016 at 08:24:46
Triode_Kingdom
Audiophile

Posts: 10047
Location: Central Texas
Joined: September 24, 2006
 "I have also seen posts to the effect that VR tubes are not a 'nice' way to regulate"

I would be very interested to see articles or posts describing measured degradation resulting from the use of VR tubes. These tubes were used for years as regulators in receivers, transmitters and transmitter modulators, often in situations that demanded excellent signal-to-noise ratios. It has always been my belief that their absence from pentode-design consumer audio amplifiers is solely the result of cost cutting.



 

RE: VR Tubes, posted on July 11, 2016 at 08:38:46
Lee of Omaha
Dealer

Posts: 1800
Location: Omaha NE
Joined: September 8, 2006
 When they were used, VR tubes were the ONLY way to achieve regulation (in the absolute sense). Compared to zeners, they absolutely suck. The lowest noise way to achieve a high-voltage reference is to put together a string of 5.6 volt zeners to whatever total voltage you need.

Except for the nice color, I can see no reason to use VR tubes in a new design.

 

RE: VR Tubes, posted on July 11, 2016 at 09:24:31
coffee-phil
Audiophile

Posts: 1444
Location: Shingle Springs CA
Joined: January 7, 2010
 Hi Triode Kingdom,

The VR tube like a zener diode is a shunt regulator with limited current capability. By itself and a current limiting resistor it can provide voltage regulation for small loads but for something like the output stage in a power amp additional amplification will be required. The VR tube would be the reference and there would be be a loop amplifier such as a small pentode or high mu triode and a big pass tube such as a 6080. All that adds considerable cost, size, and heat.

Compared to zeners they are not very good. Sometimes folks got cheap (Even HP. Look at the vertical amp in their model 150 oscilloscope.) and used little neon bulbs as VR tubes. Now those were real crap.

Phil

 

RE: Why Build a Regulated Tube Power Supply?, posted on July 11, 2016 at 09:36:00
coffee-phil
Audiophile

Posts: 1444
Location: Shingle Springs CA
Joined: January 7, 2010
 Hi Megastat 16,

I can give you a good reason. I was using an HK Citation 4 in my bi-amped system with a subwoofer. Every time the fridge would come on it drove my subwoofer and it's amp nuts. I regulated the B+ supply of the Citation 4 and the issue was solved.

It is not trivial to to get power supply rejection in vacuum tube circuits.

Phil

 

RE: VR Tubes, posted on July 11, 2016 at 09:53:03
coffee-phil
Audiophile

Posts: 1444
Location: Shingle Springs CA
Joined: January 7, 2010
 Hi Lee of Omaha,

Better than that, use a TL 430 which is a programmable shunt regulator. They have a rating of about 30 Volts, but you can put them in the cathode circuit of a medium mu triode with the grid grounded and get whatever voltage you need. Chose a mu low enough to get the cathode voltage high enough to operate the TL 430 (~ 3 or 4 Volts should be OK).

By the way I used a TL 430 to regulate the B+ in my Citation 4. I used a photo isolator powered from DC heater supply to couple to the power MOSFET pass device.

Phil

 

RE: Is it a question of tubes or a question of regulation per se?, posted on July 11, 2016 at 11:12:33
megasat16
Audiophile

Posts: 207
Location: SoCal
Joined: April 15, 2015
 Hi Lew and everyone,

This is not a question of tubes. It's a question about why a designer would want to add a regulation circuit for Tube Preamp when it's not necessary in my opinion.

Kind Regards,
James
.
.
.Thou shall not stand where I type for I carry a bottle of Certified Audiophile Air and a Pure Silver Whip.

 

VR Tubes are quieter than Zeners., posted on July 11, 2016 at 12:28:18
Paul Joppa
Industry Professional

Posts: 7296
Location: Seattle, WA
Joined: April 23, 2001
 Zeners are avalanche devices, and thus quite noisy.

Not a problem if you use current-source plate loads.

 

Because it sounds better :^), posted on July 11, 2016 at 12:36:39
Paul Joppa
Industry Professional

Posts: 7296
Location: Seattle, WA
Joined: April 23, 2001
 In my experience, shunt regulated amps, including preamps and phono preamps, sound better if they use shunt regulation. My experience includes several occasions where a shunt regulator was added to an existing design and they could be compared directly.

I don't yet have enough experience with series regulators to have an opinion.

There several theories about why, but not enough science to support or reject most of them.

 

RE: Because it sounds better :^), posted on July 11, 2016 at 12:49:31
Triode_Kingdom
Audiophile

Posts: 10047
Location: Central Texas
Joined: September 24, 2006
 Paul, which types were you using that produced improved sonics? Were they VR tubes or something else? I have only a little experience with regulation of tube preamps.



 

RE: VR Tubes, posted on July 11, 2016 at 13:04:22
Eli Duttman
Audiophile

Posts: 10455
Location: Monroe Township, NJ
Joined: March 31, 2000
 "It has always been my belief that their absence from pentode-design consumer audio amplifiers is solely the result of cost cutting."

I'm inclined to agree. A stack of 2X 0A2s/pair of 7189s, for g2 B+, and a fairly tall anode B+ rail would make for a heck of a nice PP amp. VR tube noise is easily suppressed by bypassing the stack with 0.068 muF. of capacitance. Maybe a single regulator stack would be sufficient in a stereoblock.


Eli D.

 

RE: Because it sounds better :^), posted on July 11, 2016 at 16:31:39
Tre'
Industry Professional

Posts: 17302
Location: So. Cal.
Joined: February 9, 2002

TK, I know first hand that the Bottlehead Foreplay line stage is VR tube shunt regulated.

I assembled one for a friend.

I think both the one with plate resistors and the upgrade (with CCS plate loads) both used the VR tube shunt regulated B+ supply.

I use VR tube shunt regulation in both my phone pre and my driver stages. (no line stage, I use a AVC)

In both instances the VR tube (stacks) are CCS feed.

Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"

 

Because, posted on July 12, 2016 at 09:09:34
drlowmu
Manufacturer

Posts: 9730
Location: East of Kansas City
Joined: January 10, 2005
 We don't know how to build a superb UN regulated supply !!

Jeff Medwin

 

A regulated power supply has to be linear, posted on July 12, 2016 at 10:21:55
Ralph
Manufacturer

Posts: 4774
Location: Minnesota
Joined: April 24, 2002
 and it has to have bandwidth.

This is a real trick.

You can actually regulate better with solid state if you know what you are doing; the big deal is the simple fact of bandwidth, which tubes don't do so well in a regulator. For this reason if you are doing things correctly, at the point where the tube regulation rolls off you have to bypass the output with a capacitor of the right value. The idea is not unlike that of a crossover in a loudspeaker.

Solid state, because of improved bandwidth, means that the cap to do the crossover is likely going to be smaller.

The problem is that the crossover point has a sonic footprint. If its occurring in the audio passband, then the advantage that tubes have is that their output impedance is higher, which will be closer to the impedance of the capacitor. There will be a pretty big difference between the output impedance of a solid state regulator and its attendant bypass, making the crossover and points above more audible. But if you can place that point well outside of the audio passband then there is an advantage.

So like many other things in this sport, its not an easy answer. What is most important is that you know what you are doing with the regulator rather than just shooting the dark!

 

RE: Because it sounds better :^), posted on July 12, 2016 at 12:32:25
megasat16
Audiophile

Posts: 207
Location: SoCal
Joined: April 15, 2015
 Thanks Paul for a simple and straight answer! Sonic improvement or not has been the main subject of this thread.

.
.
.Thou shall not stand where I type for I carry a bottle of Certified Audiophile Air and a Pure Silver Whip.

 

RE: Why Build a Regulated Tube Power Supply?, posted on July 12, 2016 at 12:51:27
gusser
Audiophile

Posts: 3649
Location: So. California
Joined: September 6, 2006
 Regulation in general seems to have gotten a bad reputation here but there's a lot more to the story.

First, linear pass style tube regulators are big, hot, and expensive. Some circuits need two or more isolated filament windings. And in addition to all that they are not very stable. Morgan Jones has a few paragraphs on this subject and he advises to use SS for high voltage regulation. The only reason the engineers of the golden audio era didn't use MOSFETS is because they didn't exist yet! IMO, if you can't do it with an 0Bx tube due to current limitations, use a SS approach.

Next, it is generally agreed that regulation is not desirable on any power amp, be it tube or solid state. I have confirmed that myself with regulated power supplies on PP tube amp in the 40-60w watt range. I went back to good old CLC supplies with SS rectification.

Then we come to small signal circuits. These are typically class A, especially in tube circuits. Here regulation is a good thing if nothing more than to lower the ripple to virtually immeasurable levels.

While there is legitimate engineering evidence that the impedance varies across the audio band a lot of that talk is taken out of context in audio circuits. Just how audible is it with standard LM78xx and LM317/377 parts?

Case in point, I am restoring two broadcast video tape recorders made in 1980 (Ampex VPR-2). The analog video circuits go form DC to 4.2mhz. The FM circuits go from 7-10mhz. Yet this machine is chock full of 78xx regulators and the main power supply is a classic 723 design with a stack of pass transistors. And even in 1980, the Walter Jung regulation papers were alive and well.

In my early career, all broadcast video, audio and RF equipment used regulated power supplies. That includes recording studio equipment. Ditto that for test equipment across the entire electronics industry.

Audio is not high bandwidth!

And when you consider the the above examples I question the sub performance audibility of reasonable regulation circuits in line level audio circuits.

 

RE: VR Tubes, posted on July 12, 2016 at 13:14:09
Triode_Kingdom
Audiophile

Posts: 10047
Location: Central Texas
Joined: September 24, 2006
 I'm updating a number of HK pentode-mode amps in the near future. I'm thinking along the same lines, Eli. Will probably add VR tube screen regulation in all of them.

 

RE: Because it sounds better :^), posted on July 12, 2016 at 15:01:13
Paul Joppa
Industry Professional

Posts: 7296
Location: Seattle, WA
Joined: April 23, 2001
 We (Bottlehead) mostly use hybrid regulators, with a triode cascoded to a 431-type regulator chip; sometimes we've used gas regulators which are cheaper, and we have Zeners in one current product.

Long ago we built a test bed with five or six different regulators including Zeners and gas regulators, plus three or four hybrid arrangements. The audio circuit was a single triode cathode follower, which is the most sensitive to power supply noise and linearity. The differences were small but there were two that tied for first place - we went for the simpler and less expensive one.

The other winning circuit was one developed by John Tucker; there are probably still some of his products around that use it.

Tre', the stock Foreplay II (it's no longer in production) used a gas reg, but the upgraded version replaced it with the hybrid shunt using a 12AU7.

 

RE: Because it sounds better :^), posted on July 12, 2016 at 15:11:43
Paul Joppa
Industry Professional

Posts: 7296
Location: Seattle, WA
Joined: April 23, 2001
 FWIW, the first time I heard a shunt reg it was just regulating the driver stage of a SET amp. The driver plate load was a current source with a 5 megohm equivalent resistance isolating the driver from the power supply - I did not expect to hear any real difference. The difference was so big that I immediately turned the amps over and checked for other differences. Only one amp had the regulator, so I also insisted we swap the amps in case I was hearing a speaker difference rather than an amplifier difference. It was indeed the amps.

 

RE: Because it sounds better :^), posted on July 12, 2016 at 15:15:37
gusser
Audiophile

Posts: 3649
Location: So. California
Joined: September 6, 2006
 To all:

There are plenty of listening reports on the net regarding "regulator sound".

But I would really like to see a calibrated measurement on an AP box or equivalent of some audio distortion caused by this regulator over that one. Like the infamous LM7815/7915 versus the LM317/377.

To be clear I don't need proof that regulator impedance changes with load frequency. We know that happens. I want to see how it affects the audio band. And again, I want to see calibrated scales. Not some audio magazine where they show a waveform discrepancy but fail to tell you it's 0.01db in magnitude.

 

The driver would have more gain, posted on July 12, 2016 at 18:05:23
Chip647
Audiophile

Posts: 2652
Location: The South
Joined: December 24, 2012
 I can see that if the driver was approaching its limit driving a SET, having more gain will help ensure the driver will not distort at a given output level. Really hard to hold all else equal when actively loading a tube. It will also make the modded channel louder which is almost always perceived subjectively as better.

I took this thread to be about the simple regulation of the B+ supply. i.e. 300 volts with or without regulation.

 

Measurements, posted on July 13, 2016 at 08:18:28
Triode_Kingdom
Audiophile

Posts: 10047
Location: Central Texas
Joined: September 24, 2006
 "I would really like to see a calibrated measurement on an AP box or equivalent of some audio distortion caused by this regulator over that one."

That's fairly easy to accomplish, but common measurement techniques don't tell the whole story. We all know that two amps with vanishingly low distortion, ruler flat response, similar damping factors, etc. can sound remarkably different. That's not to say specifications are unimportant. They do tell us about amplifier performance in critical areas, and I consider basic measurements to be a necessary step in the process of optimizing an amplifier. But something else is also in the mix, something we're not measuring. So, while I agree that the measurable effects of regulation are important, scientifically controlled listening tests must always be the final arbiter.

 

Thanks Paul, that's very helpful!, posted on July 13, 2016 at 08:21:22
Triode_Kingdom
Audiophile

Posts: 10047
Location: Central Texas
Joined: September 24, 2006
 Hadn't thought about smaller tubes like the 'AU7. They're entirely adequate though for low current applications like preamps. Thanks for mentioning that, it might set me off in a different direction for some of the designs I'm contemplating.



 

Everything in moderation?, posted on July 13, 2016 at 13:32:05
megasat16
Audiophile

Posts: 207
Location: SoCal
Joined: April 15, 2015
 ....when it comes to adding an active device in the chain?

All active device suffers more thermal distortion more than most passive devices...Perhaps, that extra distortion makes your music sounds so sweet?


.
.
.Thou shall not stand where I type for I carry a bottle of Certified Audiophile Air and a Pure Silver Whip.

 

RE: Everything in moderation?, posted on July 13, 2016 at 16:08:39
Tre'
Industry Professional

Posts: 17302
Location: So. Cal.
Joined: February 9, 2002
 In what sense are you using the term "thermal distortion"?

The only reference I can find for that term is an increase in harmonic distortion in a circuit due to a shift in bias in SS devices due to heat (thermal instability).

Tre'


Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: Everything in moderation?, posted on July 13, 2016 at 18:52:33
megasat16
Audiophile

Posts: 207
Location: SoCal
Joined: April 15, 2015
 I should have said it Thermal Noise. Every device has it when a signal passes through regardless of how immeasurably and small form of distortion (noise) is generated. Thermal Noise in active devices (diodes, rectifiers, transistors, valves) are much worse than most passive devices such as transformers, caps, resistors; etc.

The good thing is all that happened way outside of AF band so we won't likely hear it (or we can in some ways?).

Bias drift due to thermal instability in transistors also happened in valves (I think it's called Red Plating)?



.
.
.Thou shall not stand where I type for I carry a bottle of Certified Audiophile Air and a Pure Silver Whip.

 

RE: Everything in moderation?, posted on July 13, 2016 at 20:55:43
Tre'
Industry Professional

Posts: 17302
Location: So. Cal.
Joined: February 9, 2002
 Thermal Noise, yes all active devices add some.

In a power supply there is usually a capacitor to ground following the last active device so that noise is shunted to ground and does not continue to the amplifier circuits.

Tubes are much more thermally stable than SS devices. If a tube "red plates" there is something wrong with it or the Bias supply.

In a cathode biased push pull stage (using a common cathode bias resistor) when one tube goes bad and stops conducting the bias voltage is cut in half and that will cause the other (good) tube to red plate.

Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: Because it sounds better :^), posted on July 13, 2016 at 22:26:14
Donald North
Manufacturer

Posts: 1296
Joined: February 8, 2001
 I'm curious if the driver stage used instead a very high inductance plate choke if you would have heard the same difference with the addition of the shunt regulator.

 

RE: Everything in moderation?, posted on July 13, 2016 at 23:03:09
megasat16
Audiophile

Posts: 207
Location: SoCal
Joined: April 15, 2015
 I am sure it's a very common approach in SS circuits where high frequency noise is shunted using 0.001-0.1uF decoupling cap to Ground. But in SET and DHT preamps, I think shunting the extra noise to Ground is probably not such a great solution where ground bus noise should be kept to minimal.

As to the sonic benefits of regulator, I have to take Paul words and try and see if I too can hear a difference.

.
.
.Thou shall not stand where I type for I carry a bottle of Certified Audiophile Air and a Pure Silver Whip.

 

Comments, posted on July 14, 2016 at 09:56:29
Triode_Kingdom
Audiophile

Posts: 10047
Location: Central Texas
Joined: September 24, 2006
 Two things caught my attention in this "sub-thread." First, semiconductor diodes and rectifiers are not active devices. (I assume the reference was to semiconductors, as "valves" was a separate item.) Second, tubes are not more thermally stable than SS. No one would even think about building a free-running HF oscillator using tubes if stability was the goal. I have also seen audio tubes like the 6BL7 and others exhibit slow current drift (with constant voltages applied) requiring more than half an hour to stabilize. This would be unacceptable in its SS counterpart. Just sayin'...




 

RE: Comments, posted on July 14, 2016 at 10:53:32
Tre'
Industry Professional

Posts: 17302
Location: So. Cal.
Joined: February 9, 2002
 Don't most SS PP output stages have active offset sensing (or transistor temp sensors) and servo circuits that are continually adjusting the bias to prevent thermal runaway?

"

Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: Comments, posted on July 14, 2016 at 11:52:22
Triode_Kingdom
Audiophile

Posts: 10047
Location: Central Texas
Joined: September 24, 2006
 Self-heating and susceptibility to same are very much dependent on the specific application. Tubes and transistors each have their caveats in this regard, so I don't think any generalization is accurate.



 

RE: Comments, posted on July 17, 2016 at 13:42:24
megasat16
Audiophile

Posts: 207
Location: SoCal
Joined: April 15, 2015
 I stand corrected and thanks for correcting my mistake. It's funny how I forgot and still make the same mistakes in this past 20 years of calling diodes and rectifiers as active devices. They are passive semiconductors. But I think they can be just noisy as an active devices when it comes to thermal noise.


.
.
.Thou shall not stand where I type for I carry a bottle of Certified Audiophile Air and a Pure Silver Whip.

 

Page processed in 0.037 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]