Cheapo transformer substitution for Zen


2001-12-20 10:14 am
I apologise in advance if this is a silly question....

So I'm about to put together one channel of the Zen amp. But, being a jobless student, rather than buying a nice toroidal transformer, i've been scrounging around my house for a free replacement.

I found a pretty big 375V hammond (115 on the primary, and 375V on the secondary plus some misc. filament windings), which, to me, is better than no transformer at all. Is it legit to use the secondary winding of this transformer as the primary? (i.e.: put 115v on the secondary, and get out ~35 on the primary)?

35V is close enough to the ~25V Pass accounts for with the Zen. Is this workable/safe?

Chris McGraw


2001-06-01 4:53 pm
Ham radio folks do stuff like that all the time. Transformers are inherently reversible. As long as you derate the VA according to the amperage rating of the windings.
Transformers are bloody expensive, I try to use surplus ones too.
Grey probably just wants to see more vacuumtubeaholics around here.


2001-12-20 10:14 am
Hammond specs

The transformer I have is a hammond 274X, a supply transformer intended for tube use, cause it has filament windings. the rating is 138VA (which i'm not sure i understand--what's the relation between VA and Watts?)

While we're talking about cheaping out on the transformer, how bad would it be if I just used any old 25v (suitably rated) transformer for this project (i.e. not a toroidal one)?

Thanks for the help,
Standard 'square' transformers are commonly called EI (after the shape of the individual pieces of metal that make up the core). They will work just fine. There are advantages to both kinds of transformers, but one that will appeal to you is that EI transformers are generally cheaper than equivalent toroids. Some people go so far as to claim they sound better. I have no opinion, myself, having never tried swapping transformers.

Okay, somebody check me on this, as they're replacing the phone system here at work and there are about 10 million people running around in circles being loud and distracting.
Went to the Hammond site ( and found that the secondary on the 274X is 375-0-375 V @110 mA, which works out to 82.5 VA on the secondary (yes, the transformer is rated at 138 VA, but a fair hunk of that is on the filament windings & can't be used).
Now, the turns ratio is going to be 6.25, so if the transformer is run backwards, we're going to have 19.2VAC on the primary (which now becomes the secondary). If you use a cap input filter after the rectifier, and subtract a couple of volts for rectifier losses, you'll end up with something like 25V coming out of the power supply. That's about 9V lower than the original rail, but the amp will still work fine, albeit at reduced power.
Now, the limiting factor, current-wise, is going to be the secondary (which is now the primary). The original primary (now the secondary) is good for the full 138 VA. It then becomes a question of how much current we will pull at 25V.
Assuming that the circuit is built as published, the revised Zen drew 3A (I think...somebody will let me know if I'm wrong, I'm sure). Yes, it will still pull 3A even at the reduced rail, because the current draw is set by the sense resistor in the current source, not by the rail. That would lead to about a 75 VA load. Offhand, I'd say that would run a bit warmish, as that's a full 3A all the time.
The original Zen, however, only drew 2A (again, quoting from memory)...which would lead to about 50VA. That sounds better. More elbow room.
Okay, so here's my take--if you're going to do it, run the amp at 2A. Mind you, you can still build the newer version, but you'll need to substitute the older sense resistor (.33 ohms, yes?) for the newer value, which is, I think, .22 ohms. Or perhaps a .27 ohm resistor which would give about 2.4A (resulting in about 60VA...fairly heavy load, but it might work out).
Somebody run through that and double-check me, but it sounds as though it's just barely doable.

Transformer Testing

While we are discussing transformers, does anyone have a good procedure to test an unknown transformer for the correct VA?

I have a large selection of transformers I have collected from microwaves, TV's, kitchen appliances, old hifi gear and more. I have some very big transformers, one in particular that came from an old refrigerator, a 1500VA which is a 230 to 115 by 13 amps, this I know because it is marked on the housing, I am going use this to build a big power amp someday. But most of them have no indication on the transformer itself, and I can only guess from the specs on the original appliance.

I have tried resistor loads and etc, to get rough readings, but if there is a more precise method I would like to know about it.

Surf, Sun & Sound


2001-12-20 10:14 am
here we go

GRollins, thanks for all the info. I'm working off the original Zen plans, so it should be drawing 2A. One more question--each channel draws 2 amps, right? Is this transformer a little too small for both channels?

Just got some big heatsinks from Allelectronics, and am about to get elbow-deep in this....

'Bout the only method I've ever heard of is to load it down until the rail sags--hardly scientific, I know, but...
At least with the big one, you'll have a hard time doing that. You'll also need to arrange a center tap or some kind of floating ground if you decide to go for a bipolar power supply.
(...unless you run it backwards to get a 350V rail for a nice tube amp...shhh! don't tell anyone...Oh, Paul, uh, howya doin'? What? Me? No, sir! I ain't pushin' no tube ideas over here! Not me, sir!)
You could also use it to make a really obnoxiously huge Zen and be the envy of the diy clan.

Ouch! Two channels?
Yeah, that sounds like it's gonna get really hot unless you cut back on the bias. To do so you'd use larger values for the sense resistor in the current source. But I'm not sure (assuming that all that figgurin' I did above is anywhere near correct) that you'd be able to bias the circuit(s) high enough to be worthwhile. I hadn't thought the thing through as far as two channels--I was worried about getting enough current for one. Let's wait and see if someone catches me in a math error that will save the day.

I believe the gent did say Zen, no tubes here nosirree.
I would calculate it by assuming you can't exceed the VA rating of the transformer, and you can't exceed the current rating of any individual winding. Keep your eye on that second criteria.
The 115V winding would then be limited to 138VA / 115V, or 1.2A. The 375V windings would be limited to 110 mA, and you could use one or both in series. Feeding 115V in, you're limited to 115V * 110mA, or 12.65VA. Assuming 90% efficiency, you can only get 11.4VA out the other side.
The turns ratio would be 115/375 = 0.31 or 0.155 depending on whether you use one or both 375V windings. So my calcs say:
18VAC or 36VAC at 11.4VA; which using the standard 1.4 / 2 rule would give:
25VDC, 6 watts, 250 mA; or 50VDC, 6 watts, 125 mA.
The problem is that 110mA rating on the 375V winding. You could push it a bit, but not too far.
Hmmm. How about a nice tube amp? They're all the rage these days. Course, for the cost of the tubes, you could buy a transformer...
Cheap Transformers

you might try

they have a 120 v primary 48VCT 7A + 28VCT 3A secondary transformer for $9.95. They also have a 240v primary version that you could run on 120v if you need less voltage. I am about to order a few for my 5 W SOZ. BG Micro has 8200uF 50V caps for $1 I am also college student with no money so I know how it feels, but I think I will end up with a $60 5W x2 SOZ because of all my surplus parts.

They have to be better than my stack of modified computer power supplies powering my SOZ. (Modified so they could operate in series.)

Darrell Harmon