Hi All.
The purpose of this thread is to explore the different ways that tube amp power supplies can be built in order to avoid cross talk between channels and, more importantly, to minimize any electrical coupling between the input stage and the output stage. What really got me thinking about this is an old article written by Harvey Gizmo Rosenberg regarding the building of the ultimate DIY amplifier. Among other things, he extolled the virtue of having entirely separate supplies for pre and power tubes in addition to the separation of channels.
The article is located here,
PLITRON’S DREAM AMPLIFIER
His argument, if I understand it correctly, is that the sharing of supplies between the power and pre tubes causes the stages to modulate each other in an undesirable and audible way, and that building to avoid this produces an audibly superior amp.
To do what he recommends would require mono blocks with two separate power supplies in each one for a total of four power supplies in each stereo pair. If you followed the recommendation of the article, you would also build the power supplies using CLC or LCLC filters. In fact, he recommends using as many chokes as possible, including in the filament supplies. The end result would be a very large and very heavy pair of mono amps, that would be very expensive to build.
I am hoping to hear from you all with regard to other less expensive (and heavy and complicated) methods of decoupling and electrically isolating pre and power sections, as well as filament from B+ and, in one power supply stereo amps, the channels from each other.
Thanks to you all in advance for any and all suggestions, and for all your past help. It is much appreciated.
The purpose of this thread is to explore the different ways that tube amp power supplies can be built in order to avoid cross talk between channels and, more importantly, to minimize any electrical coupling between the input stage and the output stage. What really got me thinking about this is an old article written by Harvey Gizmo Rosenberg regarding the building of the ultimate DIY amplifier. Among other things, he extolled the virtue of having entirely separate supplies for pre and power tubes in addition to the separation of channels.
The article is located here,
PLITRON’S DREAM AMPLIFIER
His argument, if I understand it correctly, is that the sharing of supplies between the power and pre tubes causes the stages to modulate each other in an undesirable and audible way, and that building to avoid this produces an audibly superior amp.
To do what he recommends would require mono blocks with two separate power supplies in each one for a total of four power supplies in each stereo pair. If you followed the recommendation of the article, you would also build the power supplies using CLC or LCLC filters. In fact, he recommends using as many chokes as possible, including in the filament supplies. The end result would be a very large and very heavy pair of mono amps, that would be very expensive to build.
I am hoping to hear from you all with regard to other less expensive (and heavy and complicated) methods of decoupling and electrically isolating pre and power sections, as well as filament from B+ and, in one power supply stereo amps, the channels from each other.
Thanks to you all in advance for any and all suggestions, and for all your past help. It is much appreciated.
I picked up a pair of 40 watt Plitron push pull output transformers recently, and I may go the whole hog on an amp using them as per Dr. Gizmo's recommendations. Right now I have three amps on the table, and I was hoping for suggestions to get a similar result without building four CLCLC or LCLC supplies per stereo amp. Two of the amps I'm building, a 5998A SET amp and KT120 SET amp have one power supply, and the other one, a 45 amp, is comprised of a pair of mono blocks.
Last edited:
FWIW, I'm slowly gaining momentum on a #46 (DHT "El Cheapo") project that's been simmering for years. I will be exploiting a concept publicized by Pete Millett, which obtains 2 rails (1 "tall" and 1 "short") from a CT winding. The rectifying diodes do a superb job of blocking cross modulation. I've uploaded a "hen scratch" schematic that illustrates the concept, but the DHT "El Cheapo" will use Schottky diodes everywhere.
Look closely and you'll see that the "short" rail is FWCT and the "tall" rail is bridge.
You could exploit the idea in your monoblocks. A single power trafo per amp is all you need.
Look closely and you'll see that the "short" rail is FWCT and the "tall" rail is bridge.
You could exploit the idea in your monoblocks. A single power trafo per amp is all you need.
Attachments
Hi,
I just have to give a comment to this, in the last five power amps I have built ( I usually build 5 - 8 amps a year) I have used a well regulated powersupply for pre/driver circuits, this has in my opinion, given a real improvement when it comes to the sound quality of the amplifier. This is just to give an input.....for further thoughts.
I just have to give a comment to this, in the last five power amps I have built ( I usually build 5 - 8 amps a year) I have used a well regulated powersupply for pre/driver circuits, this has in my opinion, given a real improvement when it comes to the sound quality of the amplifier. This is just to give an input.....for further thoughts.
Last edited:
Thanks for the schematic Eli. Another method to add to the list. With respect to using diodes, do you think that it is possible to block all cross modulation through the use of diodes alone?
toriverson: Is the well regulated power supply for pre/driver stage a separate one from the power stage's supply?
toriverson: Is the well regulated power supply for pre/driver stage a separate one from the power stage's supply?
Last edited:
I did this once using a power transformer with multiple windings. There were three supplies: output, driver and C- for bias (300B). All choke loaded and all tube rectified. Since I didn't try a more conventional supply I can't say which was better but I can say it made it super easy to make changes to the driver without disturbing the output stage. It also ended up weighing a ton!
A good point leadbelly. It should have been part of my original post to ask if this is actually a problem at all, and if so if it needs solving. Do stages affect each other through their power supply connection in a manner that is detrimental to fidelity? If so, in lieu of building a minimum of four separate power supplies per stereo amplifier, what methods can be used to minimize this undesired interaction between stages through the power supply? Dr. Gizmo. Everything that I've ever read by him was extremely entertaining at a minimum. Some of it made me think about audio in a way I hadn't before, and ask questions. Like the ones I'm asking now in this post. I'm no expert, but my guess is that if channels adversely affect each other through their connection in the power supply, so do stages. I could be wrong though.
Last edited:
Proof that it doesn't happen in every amplifier that shares a power supply between stages is also appropriate. Why all the effort to create stiff supplies that are minimally affected by the fluctuating power needs of the amplifier? If the power tubes modulate the power supply in any way, how is that not going to affect the pre tubes and their handling of the signal since they will draw power from that same source? Do you even need to go back as far back as the power transformer? What about shared capacitors and chokes? Don't both store energy that is used by both power tubes (heavily) and pre tubes (lightly). While I would like everyone's opinion on if there are any fidelity gains to be had by using separate power supplies for each stage, I am mostly interested in hearing what others have done to minimize this problem without going whole hog and building four power supplies per stereo amp. I am also interested in their observations with the amps that they have built that utilize these ideas. I have heard a few ideas already, please keep them coming as they are much appreciated.
i have no proof other that what i experienced in a tube amp i repaired
that motorboated, replacing downstream psu electrolytic filter caps helped...
DF96. Dr. Gizmo was a wacky dude. Obviously. Much of that article is typical of his often wacky, shamanistic approach. He does bring up some interesting points though. Like the one that I'm hoping to get some input on in this thread. Oversized power transformer, inputs next to the input tube, ac filter, outputs next to the output transformer, stepped attenuator, and power switch in the back next to where the power comes in instead of the front panel? Yes, these all make sense. DC filaments? Maybe for DHTs. Lots of chokes instead of resistors? It seems that consensus is with using chokes instead of resistors generally, despite their price. Silver wire? Maybe not. Separate power supplies for pre and power stages? I never even thought about it until I read the article. Maybe it's orthodoxy and not religion that is the enemy. Unfortunately, sometimes science is also affected by orthodoxy.
Last edited:
His views might get more attention if they were presented in a more sober and readable fashion. This could be a mixed blessing: any good ideas would be noticed, but so would all the bad ideas or exaggeration too.
One of the standard methods used by the better 'gurus' is to identify a real circuit issue yet one which is several orders of magnitude too small to make any audible difference. This allows them to claim the support of physics for the circuit (so awkward people like me find it hard to criticise), while at the same time knowing that claims for audibility are hard to prove or disprove. Lower grade 'gurus' just talk nonsense.
One of the standard methods used by the better 'gurus' is to identify a real circuit issue yet one which is several orders of magnitude too small to make any audible difference. This allows them to claim the support of physics for the circuit (so awkward people like me find it hard to criticise), while at the same time knowing that claims for audibility are hard to prove or disprove. Lower grade 'gurus' just talk nonsense.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.