Back-to-Back trannies

[bibster]

Hi,

I'm looking for some information on using relatively standard (thus easy obtainable) transformers to obtain the HV needed for tube circuits.

A bit more precise: I'm building a 300V B+ PSU, maida regulated. Spits out some 326 V.. I can live with that.
I'm (for the moment) using a 220V|30V = 25V|220V back=to-back solution, using rather BIG (Big as in '2 and a half Kilograms each') transformers. This works, and gives an acceptable input voltage.
I need to reduce the weight, and increase the maneuverability...

So: Do I just search for a 1.2 VOLTAGE ratio between 2 trannies?

Than: Power rating.... I need some 30mA out, so 03mA*330=10Watts.. times 1.5 for the circuit (rather large, but won't harm!) that's 15VA on the 'secondary' side of the transformer string.

Thus: Should a pair of (30VA 12V) and (45 VA/9.5V) do the trick?
I can get these quite cheap: Surplus in the electronics-shoppe (2nd and 3rd item on this page: http://www.electronique-diffusion.fr...th=4_1537_1555)


And: Could I pick some of this 12V to feed the heaters? 0.9A@12V ?)

Or should I just say s-o-d it, and buy a 230->110V converter, wire it upside down, and see what I'll end up with? (450? 400?)

Thanks for your input,

Paul

[wimdehaan]

Hello,

I am using this kind of back-to-back in my hybrid amplifier, so 230V in 12V out transformer than a 9V in and 230V out transformer yielding +310VDC; from the first transformer you can use the second 12V winding for the heater section. You might like to check out my site at http://www.wimdehaan.nl/hybrid/5687buz900/index.html
Wim

[URL=http://www.wimdehaan.nl/[/URL]

[bibster]

Hi Wim,

Thanks for the input! Your schematics are in Dutch, but I happen to read that quite well, so...

So I guess the 30VA 12V transformer will be large enough... I'll buy a second one for 'just in case' for the heaters.

Paul

[Tom Bavis]

You may not get away with putting 30V on a 24V winding - the extra flux in the core will probably saturate it. You CAN if the transformers are rated for 50/60 Hz, and you're running on 60. And 230V on a 115V winding is DEFINITELY out of the question. Of course you end up with twice the weight of the proper transformer - but at surplus prices... who cares!

[Shoog]

I have done this myself. I found trying to push more voltage into the secondary of the second transformer than it is rated for resulted in saturation, which looked like flattened waveforms on the scope. This may not be a universal situation, but my conclusion was that you could only get out the voltage you put into the first primary, with "maybe" the option to squeeze 5-10% extra.

Personally I would only consider this for class A preamp duty.

Shoog

[bibster]

well...

It just 'happens' to be for an SRPP preamp, so... ;-)

Got a 12V and a 9V, works okay, but I can't look at the waveforms due to a severe lack of scope. (Please tell my wife that this is UNFORGIVEABLE, not te have a scope)

Paul

[Shoog]

My scope has just failed on me. Its a simple power supply problem (I think), but they just aren't servicable, as you can't get access to the PCB. Second hand scopes don't come up often in the **** end of nowhere.

Did you give up on your 5687 preamp in the end ?

Shoog

[bibster]

Hi Shoog,

What a pity yer scope just failed on you... I just HATE this non serviceable stuff..
That's like the nec-plus-ultra of modern consumption-society.. broken? Throw it away and get a new one.

anyhow... kind to ask about my 5687 preamp!
This IS for my 5687 preamp ;-)

I turned my 'first attempt to fiddle with tubes' into an aikido with help of John's board (and a pair of 12ua7's). Adapted the PSU with a maida reg. to get some 300V (326 as a matter of fact... lousy zeners I guess), add a nice regulated heater psu... Works like a charm!

Now I had this pair of other 5687 laying around, and I owed something to a good friend.. so... I started building a little pcb with another (identical) maida reg (with FWB and 'lytics this time, no tube rectifier) and a 12V (7812 for the moment) heater supply.

I didn't like the CSS load, su I thought an SRPP could to the trick, just to see what it's like. It's sort of like "pat's ultimate preamp", but with 10uF output caps (SCR, that's french for Solen, innit?) , 1k2 resistors (about 8ma idle thus) and 220uF lytic + 470nF MKP cathode bypass.

Didn't want to buy an expensive 220-260V tranny, so I went the el-cheapo way... :-)

A bit more threat-jacking: What's this story I read somewhere saying to put a 50K volume pot BEHIND this "pat's ultimate" ? True or false?
But then, sould I s-o-d the 1M resistor between the out and Gnd ??

Regards, Paul

[Eusebius]

My experience is as follows:
I used two 25VA transformers for a preamp which has choke input to a preamp tube stage that takes 12mA

I'm in UK so here it's 230v in to 12+12, then 12+12 to 230v out (approx). In the centre I've connected each 12+12 in series to give 0-24 and then connected the 0-24 ends

In theory the trannies should be good for 100mA but in practice the problem is with the first - it overheats. A friend told me it was "magnetically coupling" which I don't know much (anything) about. The trannies are covered in a blue plastic wrap making them square on the outside, and are side by side with bodies touching. So the heat is additive - not too good. But even so - it surprises me that I'm getting overheating. The trannies are below the top plate and the top plate becomes very uncomfortable to touch after a few hours. Heats up slowly.

Can I solve this with better ventilation or do I need a bigger first trannie?

[EC8010]

Real transformers are different from perfect transformers. If a transformer was perfect, then using one transformer to step down, followed by another to step up would be fine. However...

Although the ratio by which a transformer step up (or steps down) a voltage is dictated by the turns ratio between primary and secondary, this implies that so long as you had the right ratio, you could use as many or as few turns as you liked. Imagine a transformer with a secondary not connected to anything. If the secondary isn't connected to anything, we can simply throw it away. We are now left with a choke across the mains. The choke (usually called primary inductance) has a reactance:

X_L = 2*pi*f*L

We have mains across this reactance, so a current must flow that is inversely proportional to the reactance (V/R = I), and this is generally known as the magnetising current. Inductance is proportional to the square of turns, so the more turns we have on the primary, the higher the reactance, and the lower the magnetising current. The key issue is that the higher the magnetising current, the more heat will be generated in the wire and in the core.

It might seem that the ideal would be to use as many turns as possible to maximise the inductance and thus minimise magnetising current, but thin wire has a higher resistance and would cause higher losses when a load current was drawn. Thus, the primary inductance of a transformer is a careful balance of a number of factors to maximise efficiency.

What happens when we connect two identical transformers back-to-back? The primary inductance of the second transformer is reflected through itself and seen (scaled down by the square of the turns ratio) across its low-voltage winding. We now connect that low-voltage across the secondary of our first transformer. The first transformer reflects the external inductance through itself and it appears across the primary of the first transformer. The important point is that we now have two equal primary inductances in parallel, so the magnetising current for the first transformer doubles.

Large transformers (the kind that are bigger than refrigerators) are carefully designed to be extremely efficient, but small transformers are designed to be cheap. The upshot is that a small transformer might be so inefficient and have such a high magnetising current that connecting two back to back could cause the first transformer to be overloaded by the two magnetising currents even before any external load is applied.

Summing up: Connecting transformers back-to-back might be convenient for providing HT, but it relies on the transformers having low magnetising current to avoid overloading the first transformer. A quick and dirty way to check for suitability might be to connect one unloaded transformer to the mains, and see how warm it got after five hours. If it stayed cold, it would probably be suitable for back-to-back connection.

An alternative? Use a 1:1 isolating transformer followed by a voltage doubler if necessary.