Tube PSU for a Dual Unit

[cannikin]

Guys,

This has been in my brain for a few weeks and need to know the details on the hows and whats as far as building one power supply for a dual mono unit tube unit because of familiarity (for me) lets use the Am864U for an example:

for a single unit you need:
Primary: 115V
Secondary A: 175V-0-175V @100ma (i.e. center tapped)
Secondary B: 3.15V-0-3.15V @2A (center tapped)
Secondary C: 5V @2A

So what transformer requirements would you need to build a dual unit with one PSU.

 

[Rob Flinn]

Double the current ratings for the secondaries A & B.
Check whether the rectifier valve will handle 200mA, if it does then the the 5v is ok because you only need one rectifier valve.

 

[cannikin]

Thanks Rob,

So what needs to be checked carefully is the DC output Current on the rectifier tube... correct?

5y3 Dc output Current is 125mA

5AR4 DC output Current with Cap Input is 160mA
5AR4 DC output Current with Choke Input is 220mA

5AU4 DC output current is 325mA

So it might be best to go with a 5AU4 in an Am864u type circuit.... right?

 

[pstamler]

I thought the definition of "dual mono" included a separate power supply for each unit? If you want to be strict about it, two transformers, two rectifiers, two sets of caps...

Otherwise, what everyone said. Watch out for grounding issues.

Peace,
Paul

 

[PRR]

> lets use the Am864U for an example:

Well, that particular beast shows 30mA for the HV winding. My ruff-peek says it can't be sucking a lot more than 20-25mA. Double for dual, you are still well into 5Y3 territory.

If your "example" is a decoy, study the specific unit you wanna double-up.

 

[cannikin]

[quote author="pstamler"]I thought the definition of "dual mono" included a separate power supply for each unit? If you want to be strict about it, two transformers, two rectifiers, two sets of caps...
[/quote]

Ok I was thinking of just the audio as "dual mono"... so let me me rephrase that -- two identical circuits or units using one power supply.

 

[cannikin]

[quote author="PRR"]
Well, tha particular beast shows 30mA for the HV winding. My ruff-peek says it can't be sucking a lot more than 20-25mA. Double for dual, you are still well into 5Y3 territory. [/quote]

Well here is what I don't understand PRR. If the transformer is rated for 200mA. then will it not be pumping out 200mA of current? even though the circuit only draws a total of 89mA of current (just a number). Or is a transformer current rating like a Capacitor Voltage Rating... you can safely draw up to 200mA of current from the transformer but if you don't use it, it doesn't matter because the circuit will only draw the mA is needs or uses.

I know, I know.... I need to understand Ohm's Law better.

[quote author="PRR"]
If your "example" is a decoy, study the specific unit you wanna double-up.[/quote]

No decoy.... but I wanted to stay in the realm of a tube circuit I semi-understand and know about, I know this topic could probably get philosophical or heady without an example... and when EE people start going into that theoretical place is when I'm f*ked and I no longer understand.... it ends up being the equivalent of whos on first, whats on second..... banter for me :shock: :? and I get lost

 

[bcarso]

200mA is what it could supply, not what it forces you to accept.

 

[PRR]

> Well here is what I don't understand PRR. If the transformer is rated for 200mA. then will it not be pumping out 200mA of current?

I have a 200 Amp fusebox, and presumably 200A worth of transformer out at the street. If I was really sucking 200 Amps, my electric bill would be $4,493/month.

Actually, I turn everything OFF, I use "no" power. And if I have just one 120 Watt lamp, or just 40 electric clocks, I pull just 1 Amp at 120V.

What the "200 Amps" means is: something bad will happen if I pull 201 amps. Somewhere near 500A for several hours, the street transformer overheats and goes boom. Perhaps 400A for an hour would make my feeder terminals burn off. To prevent this kind of excitement, I have 200A fuses which (are supposed to) trip-out at 400A for a few seconds or 220A for a few minutes.

But if I putter around with 1A lamps and 10A toasters and a 25A air conditioner, I never get near any 200A demand. I seem to average more like 5A.

(For a fusebox, over-sizing is fairly cheap. And some kinda safety against future demands... the A/C came 30 years after this fusebox, and we didn't have to think about it when we got the A/C.)

------------

You propose a PT rated 175V at 200mA load. If you actually connect a 201mA load, and it smokes, the maker won't listen to your complaint, altho you can probably run 250mA for years, getting something less than 175V. You can run it at zero mA, but the voltage will rise to maybe 200V. You can run it at 60mA and get something between 200V and 175V, with heat something between zero and full-load heat.

In today's market, you may "have" to take a 100mA or 200mA iron for a 60mA load. They don't make every possible spec.

> when EE people start going into that theoretical place is when I'm f*ked

Yeah, they do that too much.

The original wanted 175v-0-175V at 30ma and 6.3V at (IIRC) 1.05A. So for two, double each current: 60mA and 2.1A.

The original had one 5Y3 rectifier to make 30mA of DC. Can one 5Y3 supply 60mA? Yes. The voltage will be a hair lower, but I'm not concerned. The Am864u is not a precision limiter, you won't know the difference if B+ comes out 10V lower.

Anyway you won't find exactly 175V 60mA 6.3V 2.1A. You will probably have to buy something a bit bigger. It will be underloaded, so it will make a bit more voltage. If you are lucky, maybe just enough to counterbalance the drop from taking twice the current through the 5Y3.

While the math says 60mA and 2.1A, if you are "stuck" with a 100mA winding, the other can be nominal 2.0A. One is a bit underloaded, the other a hair over-loaded. If you don't get greedy, you can cheat a winding's rating 5% or 10% if other windings are 10% or more in excess of your needs.

If you want to be more discrete, really -separate- channels, you might indulge yourself with -two- 5Y3s. Then you not only need twice the 5V current, you need -two- heater windings (because on 5Y3 the heater/filament is also the B+ output, and you get no channel isolation if both are on one winding.) But you will find few irons with two 5V windings until you get to exotics which cost more than two small irons.

 

[cannikin]

thanks PRR, that is really informative and answers my question.

its always a pleasure to read your posts because they are direct, chuck full of information, and clarify/expand my understandings, granted the forum is full of amazingly knowledgeable people ... just want to say thanx fr taking the time, much appreciated!

 

[pstamler]

Here's another rule of thumb: Many power transformers are rated at the current which makes the voltage fall by 10% from unloaded. (And you shouldn't run them at higher current, because the voltage drop dissipates as heat, and could fry the tranny.)

So let's say a transformer puts out 200V when not loaded. If you load it down such that it's now putting out 180V, and discover that it takes 100mA of curent to do that, you have a transformer officially rated at 180V/100mA.

By the way, that lets you figure out the transformer's equivalent series resistance. When you pull 100mA out of it, you're dropping 20V, so the equivalent series resistance is 20V / 0.1A = 200 ohms. Which suggests that if you pull 65mA out of that transformer, the ESR will drop 0.065 * 200 = 13V from the unloaded figure, so your loaded voltage will be 200 - 13 = 187V.

YMMV if the manufacturer computes ratings differently, but that's how it works in classic parts like Triad or Thordarson.

Peace,
Paul

 

[cannikin]

Great stuff.... pstamler

Thank you so much!

Makes me think... I have these NOS Triad filament transformers rated 2A and they get HOT too hot to touch after an hour.. so I'm wondering if I'm loading them too much.

They are powering 2 6sk7 (which is converted to DC), 6al5, 12At7, and a hum balance Pot.

Curious.... I should look into how much I'm loading them

THanks again thats awesome :thumb:

 

[emrr]

Filaments; a similar yet slightly different version of the 'problem'. If you use a filament transformer with a vastly higher current rating than needed, you will probably find your voltage above the standard window, which may or may not be acceptable for the tubes in question. Life will be shorter. Maybe not shorter than yours or mine. I think you want to land somewhat close to the filament transformer rating. In many classic designs you find a rating and a load that has room for one more tube than found; just enough 'headroom' to keep transformer temp comfortably below rating and voltage above design center, but not above design maximum. Remember, a 6.3V tube design center has minimum and maximum 'on paper' ratings of 6V to 6.6V. Most will work fine a good bit below, and somewhat above.

I consider B+ a good bit more forgiving in most cases. A limiter with specific voltage references will be more picky in terms of lab performance; a preamp may well be happy anywhere from 150 to 300 VDC, depending on what you are asking it to do.

Think in ratios where you can, and figure out where the fudge room is.

mmmm, fudge.....

 

[PRR]

> Here's another rule of thumb: Many power transformers are rated at the current which makes the voltage fall by 10% from unloaded.

Depends on size. Under 25VA it is not uncommon to rate for 20% sag. Over 1,000VA it is unlikely to find 5% sag. Very small iron won't overheat, big iron will: Surface/Volume ratio. (The electrical rating runs nearly as Weight which is same-as Volume; the thermal rating runs as Surface Area.)


> rated 2A and they get HOT too hot to touch after an hour
> 2 6sk7 (which is converted to DC), 6al5, 12At7, and a hum balance Pot.

6AL5 is 0.3A.

12AT7 wired for 6V is 0.3A.

"Hum Pot" could be anything. 100 ohms is popular. 6.3V/100R is 0.063A.

6SK7 is 0.3A ea or 0.6A for two BUT "converted to DC" is hard work and simple men like us should -double- the DC current to estimate the AC current from the transformer. So this is like 1.2 Amps right there.

0.3A 6AL5
0.3A 12AT7
0.063A humpot
1.2A two 6SK7 on DC
====
1.863A

You are essentially "AT" rating. And it is not unlikely that rated load is too hot to touch. And 1960s windings can run HOT since they had better stuff than the old varnishes. Still seems a bit hot for a mere 12VA transformer.

You know if you found a 12V 0.6A or 24V 0.3A DC power supply, you would have no hum inside the box and no need for a hum-pot. It is generally "stupid" to make just 6V DC, since the rectifiers lose a volt or two. That's small at 24V but not-small for a 6V supply.


> voltage above the standard window, .... Life will be shorter. Maybe not shorter than yours or mine.

That's my experience. If you don't run low-bid tubes in long series string, if you don't throw them in dumpster and pick them out again, heaters last forever, or longer than I care about.

A "6.3V" heater can be run above 24V for many minutes without dying. Bored, I took it past 40V and it did flash-out. Life is some kinda high power of voltage, like 13th, so if 4X voltage lasts a minute, anything near nominal voltage lasts a century.

 

[emrr]

Tacking onto this oldie, since I was already talking about filament transformer rating here. 

I found an example of a broadcast limiter power supply that changed the pass tube type in a later generation.  All versions of the supply have a 2A/6.4V filament rating on the critical winding, the 1st version put a 2A draw on the 2A/6.4V winding, the second version puts a 2.5A draw on the 2A/6.4V winding.  I can't give any further backstory on the rating, but it's clear in this case that it lasts 'long enough' in a mission critical piece sold over 60 years ago.  Now do I want to keep putting 2.5A on it?  I doubt I will, since there are multiple pass tube options that will work, and the transformer can't be replaced without a custom order.