Please validate my PS schematic

[Hyakuya]

Hi all

I'm in the early stages of planing a 2A3 SET amp build based on the Shishido Roftin White and I'm starting at the PS with an eye on the bits and pieces I have laying around already.



I've got all of the parts in this schematic already (except R2) so it's a zero cost design. The idea being to use a pair of 6AU4GT to provide time for the mercury vapour rectifier to warm up before the HV is applied. I have read that the 83 rectifier doesn't actually need any warm up time but I'd prefer it this way.



PSUD gives me an output of about 500V so R2 is then used to drop that down to the 450V in the Shishido schematic.

I'll use separate transformer for the 2A3 filaments, the Hammond 266 have 2.5V options with 230V primary.

A couple of questions then....

1. does it look ok ?
2. For the 2A3 I see lots of schematics using separate windings / transformers for each tube, i.e. 2.5V @ 3A to each rather than a 2.5V @ 6A to both in series. Should I keep these separate ?

Thank you all in advance....

 

[gadget73]

The separate windings is likely either for ability to bias and/or hum balance the tubes individually. I'd say thats your call there, makes it slightly more flexible if you don't have reasonably matched tubes vs having to use a single common cathode resistor though.


I guess you could do it that way, personally I'd probably put a single slow warmup rectifier between the center tap and ground. Plate to ground, cathode to the center tap. Would do the same thing electrically speaking, just using one tube instead of two.

I don't know that either config is truly enough time to really get a vapor rectifier going though. You're probably adding a handful of seconds worth of delay. I suppose a better question is why use a vapor rectifier at all?

 

[Hyakuya]

That makes sense so I'll go with 2 Hammond 266M2 with one for each 2A3.

So you're suggesting like this....


The type 83 doesn't require the filaments to be pre heated but as I said, I already have all this gear so it won't hurt to treat it gently. The data on the 6AU4 says heater warmup is 11 seconds.

Why use a vapor rectifier.... I've already got it and why not ?

 

[gadget73]

yup, should work just fine.

 

[TriodeLuvr]

I'm not a fan of designs with components that serve no purpose. This would work just as well with either the 6AU4s or 83. Both aren't needed. The 83 can be sufficiently protected with an inrush limiter, although its constant-voltage characteristics would be more useful in an AB1 design. Does the 83 glow? There's one in my tube tester, but I don't recall thinking it was much to look at.

Also, I know you didn't ask about the amplifier proper, but a 12AX7 with both sections in parallel may not be the best choice. I think you'll get better performance from a circuit like the one below. The 2A6 is still available at reasonable prices.

Jack



https://www.bonavolta.ch/hobby/en/audio/2a3_1.htm

 

[Hyakuya]

Interesting schematic... I'll need to stare at it for about an hour before it makes sense to me.....


what's this:

Is it an ampmeter ?

I don't know if the 83 glows. I chose it because I already have it.

 

[AlexHempel]

@TriodeLuvr This is a cool schematic. I have an aspiration to build a directly heated pentode amp with a directly coupled driver and a similar output tube biasing scheme, so I'm going to study this. Have you built it before? It seems many people advocate DC for DHT heaters, so I'm curious if the hum is audible with this. I'm guessing the purpose of the 20 uf cap between B+ and the heaters is B+ hum cancellation?

 

[kward]

Seems to me you are providing a solution to a non-problem. Why add more circuitry that ultimately is not needed? It's just one more thing to eventually and potentially be a problem down the road. If the 83 can handle the brunt of a cold start, why fix something that isn't broken? Why over-engineer it? Why solve a problem that isn't a problem? Your amp's reliability (MTBF) will increase I think without it.

Also I've never designed with a mercury vapor rectifier, but it might be a good thought experiment to know what happens in a hot cycle situation. You will eventually hot cycle the amp, and in that case the 6AU4 will become transparent or mostly transparent in the circuit depending on the number of seconds between power cut and reapplication of power. Sometimes the power company does that for you, unexpectedly

 

[Squidward]

Mercury rectifiers have some of the coolest glows in all of tubes. They're used sometimes for their visual aesthetic, as well as for their low voltage drop compared to similar non-mercury rectifiers.

That being said, they do require special handling. Warm-up, cool-down, etc. I understand that if you treat them correctly you're fine, but there is the accidental hot-cycle, etc. that as kward says will eventually happen.

 

[TriodeLuvr]

Interesting schematic... I'll need to stare at it for about an hour before it makes sense to me.....

what's this: View attachment 2614883

Is it an ampmeter ?

Yes, it monitors cathode current.

Jack

 

[TriodeLuvr]

Mercury rectifiers have some of the coolest glows in all of tubes. They're used sometimes for their visual aesthetic, as well as for their low voltage drop compared to similar non-mercury rectifiers.

That being said, they do require special handling. Warm-up, cool-down, etc. I understand that if you treat them correctly you're fine, but there is the accidental hot-cycle, etc. that as kward says will eventually happen.

The 83 is different, and it doesn't require special treatment. The example in my decades old Hickok tester is the original, and there is no provision for pre-heating.

Jack

 

[zackthedog]

The 83 is different, and it doesn't require special treatment. The example in my decades old Hickok tester is the original, and there is no provision for pre-heating.

Jack

I'm pretty sure the Hickok activates the 83 filaments at turn-on, which gives them time to pre-heat before actual testing takes place. I can say from experience that hitting an 83 that's been off for a few days with instant HV isn't pretty.

 

[TriodeLuvr]

@TriodeLuvr This is a cool schematic. I have an aspiration to build a directly heated pentode amp with a directly coupled driver and a similar output tube biasing scheme, so I'm going to study this. Have you built it before? It seems many people advocate DC for DHT heaters, so I'm curious if the hum is audible with this. I'm guessing the purpose of the 20 uf cap between B+ and the heaters is B+ hum cancellation?

I haven't built this, so you've been warned. The 2A6 looks like an interesting choice for as a single triode that can do the job of driving the capacitance of the 2A3. It does have a grid cap, but I wouldn't consider that a deal breaker, especially considering the nice ST-shaped bottle.

Jack

 

[TriodeLuvr]

I'm pretty sure the Hickok activates the 83 filaments at turn-on, which gives them time to pre-heat before actual testing takes place. I can say from experience that hitting an 83 that's been off for a few days with instant HV isn't pretty.

You're right about that. I had thought my 6000A applied B+ immediately, but the schematic does show travel through the Test switch. Nevertheless, data sheets for this tube do not carry the same statement as its larger cousins calling for a 30 second warmup. After searching a bit, I did finally come across a statement in a combined 82 and 83 sheet regarding this.



None of the 83-specific sheets I've ever read say this, and I do believe an inrush limiter will fully resolve any potential issues. Thanks for your input on the Hickok.

Jack

 

[Hyakuya]

Thank you for all the advice and information guys... As always it's just what I need.

I'm sticking with the 83 and going to think some more about how I'm going to put the PS together.

 

[trobbins]

The 6AU4GT has limited heater-cathode voltage capability,,so the first schematic has a problem. The second schematic has the 6AU4 in a better location. However, preferably add 1N4007 in series with the 83 anodes, or add anode fusing, to manage collateral damage to the transformer from 83 arc-back.

 

[arts]

Why use a vapor rectifier.... I've already got it and why not ?

They generate a fair amount of electrical noise, so you'll probably need to incorporate some plate chokes. I have seen instances where additional mechanical shielding from surrounding circuitry has been needed. Also, the 83 really won't enjoy that first 100µFd filter capacitor, so you will need to either greatly reduce it's value, or go with a choke input filter.

While the 83 does not require filament pre-heating for general use, it does require that you pre-heat the tube when initially installed (or any time it's been physically moved, ie from a vertical to horizontal position) in order to condense the mercury throughout the tube. Otherwise, any mercury that has gotten onto the filaments or inside the plate structures may cause shorts or localized overheating. Usually, 15 minutes or more is adequate. Obviously, with no plate voltage applied.

 

[TriodeLuvr]

While the 83 does not require filament pre-heating for general use, it does require that you pre-heat the tube when initially installed (or any time it's been physically moved, ie from a vertical to horizontal position) in order to condense the mercury throughout the tube. Otherwise, any mercury that has gotten onto the filaments or inside the plate structures may cause shorts or localized overheating. Usually, 15 minutes or more is adequate. Obviously, with no plate voltage applied.

Are you sure that precaution applies to the 83? I don't see how it could be used in a portable tube tester if that was the case. There's no warning about anything like this in the Hickok manuals that I've read.

Jack

 

[EL504]

The 6AU4GT has limited heater-cathode voltage capability,,so the first schematic has a problem.

Could you please explain why?

With the heater being negative with respect to the cathode, the maximum heater to cathode voltage of the 6AU4GT is 900 V for DC and 4500 V for DC + peak. With the heater being positive with respect to the cathode, these numbers are only 100 V and 300 V.

I thought that the cathodes of the 6AU4GT in the first schematic never go negative with respect to their grounded heaters. Am I wrong about that? Or is there something else I'm missing?

 

[gadget73]

cathode would always be positive, voltage difference would be equal to the B+ voltage if the heater circuit is grounded. Seems like it ought to be fine at any sort of "normal" B+ level.