Regulated DC Filament Supply (Pete Millet)

[atmasphere]

In properly designed equipment, this is a moot point. Gazillions of tube type equipment have been produced and are still around being restored.

Tube rectifiers are still very much alive, and people use them every day in new builds.

Very true! But just because they are still in use is not the same as saying they are the best choice for the job as we see below:

There are more benefits to point-to-point than that. Ease of repair. I recently had some resistors fry on the input pcb of my Dynaco ST70. The board was totally fried and needed to be replaced along with all the mounted components. Took a fair amount of time and money to fix. If it had been fully point-to-point, it would have been a simple diagnosis and fix for $2 worth of resistors.

Simplicity is king. If I have a tube rectifier go bad, I open up the drawer, grab a replacement and pop it in. Five minutes max. I have only replaced one in the last ten years. I also don't need to be an electronics engineer to do so but understand the slow start benefits of my GZ32/5AR4 rectifier.

I think we would agree that proper design is essential no matter if it is old or new school. Bad designs exist in both, including poorly laid out pcbs that have ac traces running parallel to dc traces, underrated traces, etc. I coexist with both. But I am old enough to have learned, old technology may not be efficient, but in most cases it is more user friendly. Heck, I can fix my '67 gas guzzling, carbureted Vette with nothing more than a screw driver, pair of pliers and a hammer in most cases.

You can fix that 'vet, and if you spend enough time with it, you'll notice those old V8s just are not as reliable (and I'm speaking from the experience of keeping that sort of drive train on the road for 25 years and 680,000 miles- it was mandatory I bring a proper tool box with me to sort it out along they way...) even though they are lot easier to work on than the modern stuff.

And that rectifier is a good example of what I'm talking about. In the ST70 you'll find its the most likely tube to fail. Dynaco cut a corner there and really should have had dual 5AR4s to handle the current of the amp at full power. You can't really replace that tube with solid state since the output Voltage will be too high (although I've not tried some of those 'rectifier replacement' things you plug into the socket...). Triode Electronics makes a slightly larger power transformer that drops into that position in the ST70; it allows you to run dual rectifiers. With modern filter caps you can thus have a much more reliable amplifier and one that makes a slight bit more power since B+ doesn't sag at full get.

 

[John Berard]

Very true! But just because they are still in use is not the same as saying they are the best choice for the job as we see below:

You can fix that 'vet, and if you spend enough time with it, you'll notice those old V8s just are not as reliable (and I'm speaking from the experience of keeping that sort of drive train on the road for 25 years and 680,000 miles- it was mandatory I bring a proper tool box with me to sort it out along they way...) even though they are lot easier to work on than the modern stuff.

Puppy. Doesn't quite match my 64 years with millions of miles. But I understand FULLY what you are saying. The biggest culprit was points based ignition. A simple conversion to capacitor discharge fixed that. Cars didn't go as far as modern vehicles until a problem arose. These modern cars are reliable until they aren't. There are 60 to over 100 sensors in new cars. Each one will eventually fail. I have my share of MAFs, crank position sensors, oxygen sensors, etc. go bad. And you know what? The car died and no way to fix it out on the road. I also live on the coast with a salt air environment. All those connectors corrode and create quirky issues that are hard to track down.

I also have extensive experience with modern high tech 'go fast' vehicles. They are expensive and require high tech tools to work on. Heck, now we have eight coils and injectors to worry about. If I got a bad tank of gas in my old cars, they would run crappy but they did keep running. A modern car just stops in it's tracks because the computer (that will also eventually fail) won't allow anything outside the very strict parameters. Here is one of my builds. Power? Yes! Reliable? Hell no! Expensive? You don't want to know.

And that rectifier is a good example of what I'm talking about. In the ST70 you'll find its the most likely tube to fail. Dynaco cut a corner there and really should have had dual 5AR4s to handle the current of the amp at full power. You can't really replace that tube with solid state since the output Voltage will be too high (although I've not tried some of those 'rectifier replacement' things you plug into the socket...). Triode Electronics makes a slightly larger power transformer that drops into that position in the ST70; it allows you to run dual rectifiers. With modern filter caps you can thus have a much more reliable amplifier and one that makes a slight bit more power since B+ doesn't sag at full get.

That one tube that died from old age was not in my ST70. Also, tubes still do some things better. Rad hard space satellites as an example.

 

[gadget73]

The 5ar4 is worked to death in those, they really should have had a pair.

I have 3 amps that run 5ar4s now, one has a single, one has two, one has 3, so a half dozen of the things between 3 amplifiers.

 

[Selmerdave]

Question about regulator windings - is there an advantage to fully independent transformers over two windings on a single transformer? I have two options - a pair of standard Stancor filament transformers or a nice Triad HSM with two secondaries. Ratings are adequate in both cases, is there a reason to choose one over the other?

 

[THD+N]

Question about regulator windings - is there an advantage to fully independent transformers over two windings on a single transformer? I have two options - a pair of standard Stancor filament transformers or a nice Triad HSM with two secondaries. Ratings are adequate in both cases, is there a reason to choose one over the other?

No reason other than one solution is more compact than the other.

Any project details?

 

[Selmerdave]

I'll post some photos and details once it's really underway, but it's for a DHT preamp.

 

[John Berard]

Two regulators powered by single winding for DHT filament use. Other application not an issue.

Not a Pete Millet board but here is a different scenario for a budget 6B4G build I will be doing. One winding to one DC Regulator to two valves? I also have the option of a 6.3V 5A winding.

 

[THD+N]

It will work, but your channel cross talk (left/right channel separation) will be compromised. In other words, the channel separation will not be as great as it could be.

 

[John Berard]

It will work, but your channel cross talk (left/right channel separation) will be compromised. In other words, the channel separation will not be as great as it could be.

Now that cross talk is part of my learning journey, I must ask if this was a pair of valves on one channel of a push-pull amp, it would be fine?

 

[THD+N]

Now that cross talk is part of my learning journey, I must ask if this was a pair of valves on one channel of a push-pull amp, it would be fine?

Yes.

 

[atmasphere]

That one tube that died from old age was not in my ST70. Also, tubes still do some things better. Rad hard space satellites as an example.

(emphasis added) I certainly don't argue that! But old age really shouldn't kill a rectifier tube. The ST70 simply should have run dual 5AR4s. I used to own a Knight stereo amp (KA-95) that made 35 Watts/channel and they bothered to use dual rectifiers in that.

If you have a situation where the use of semiconductors will get you a proper plate Voltage, there really isn't a good argument to not use them unless you don't care to snub them properly. HEXFRED rectifiers are extremely fast and extremely soft recovery and in most situations I've used them have no need of snubbing. These days the best use of tube rectifiers is either for looks , a blues guitar amp or a vintage hifi amp where you really have no choice.

 

[Selmerdave]

(emphasis added) I certainly don't argue that! But old age really shouldn't kill a rectifier tube. The ST70 simply should have run dual 5AR4s. I used to own a Knight stereo amp (KA-95) that made 35 Watts/channel and they bothered to use dual rectifiers in that.

If you have a situation where the use of semiconductors will get you a proper plate Voltage, there really isn't a good argument to not use them unless you don't care to snub them properly. HEXFRED rectifiers are extremely fast and extremely soft recovery and in most situations I've used them have no need of snubbing. These days the best use of tube rectifiers is either for looks , a blues guitar amp or a vintage hifi amp where you really have no choice.

In that sense there isn't a good argument to use tubes at all. Not everybody shares that dogma.

 

[THD+N]

The atmosphere must be real thin up there…

 

[atmasphere]

In that sense there isn't a good argument to use tubes at all. Not everybody shares that dogma.

?? How did you get to that conclusion??

There's a right tool for every job. If you want to hear tubes at their best, you give them every opportunity to strut their stuff.. That starts with a good power supply. If the amp is class A, the rectifier may not make such a difference since it will have a nearly constant load. But if the amp is AB or a larger amp, solid state rectifiers makes more sense. You can get less distortion using them, since they won't have a Voltage drop when more power is drawn.

Of course if you don't pay attention to their care and feeding (proper neutralization) then you might like the tube rectifier for its lower noise.