Transformers and Goofy Windings

[The_Armadillo]

Hey everybody,
So, a while ago I came across this old GK-something amp that was all blown up. I tried to part out as much as I could, but nothing was really worth keeping except the big transformer. If I remember correctly, there are two secondary windings, one at 90VCT with ~4.5A capability, and another at 82VCT with ~2.5A. My question here is whether I could connect the two windings in parallel, in order to get higher current capabilities from the transformer. I'm trying to get at least 500W out of it.

I'm not sure if this is the right place to post it, but I figured that there's plenty of guys here that could steer me in the right direction.

Thanks dudes.

[defaced]

No. Parallel windings need to be voltage matched. If they're not, the higher voltage winding will force current through the lower voltage winding causing it to overheat and short.

[R.G.]

No.

What he said.

[The_Armadillo]

Thanks guys, I thought it was somthing along those lines, but I just wanted to be sure.

[R.G.]

The 90Vac winding probably has under 0.1 ohm resistance, the 82V winding the same. When you connect them in parallel, the 90V winding wants the 82V winding to be 8 volts higher. The only thing preventing the current from being infinity is the 0.2 ohms, so I = 8V/0.2R = 40A of current flow

*Note that the 0.1 ohm resistance is a wild-eyed guess. But whatever it is, the current is bad. Circulating currents are the bane of all paralleled transformer connections. This gets especially bad in 3-phase delta systems feeding different impedance loads.

[Perrow]

I'm guessing here, but if they each have their own rectifier, wouldn't that help. When loaded the 90v winding would drop to where the 82v winding would kick in. I'm guessing (again) that it would be "soggy", but perhaps worth it.

[R.G.]

I'm guessing here, but if they each have their own rectifier, wouldn't that help. When loaded the 90v winding would drop to where the 82v winding would kick in. I'm guessing (again) that it would be "soggy", but perhaps worth it.

You're right - if the 90V winding ever dropped that far. Back to Ohm's Law: the peak voltage of a 90V AC sine wave is 90*1.414 = 129.6V. The peak of an 82V sine is 82_1.414 = 115.948. The difference is 129.6 - 115.95 = 13.5V. The resistance this drops in is the 0.1 (or whatever it really is) in the winding, plus the resistance of the diode(s) minus the forward drop of the diode. Call the diode drop 2V at high currents. So you have 11.5V across 0.1 ohms, or 115A.

Obviously, this only happens when it's charging a filter cap, and the currents can be big then. But even then it's asking a lot for a 4.5A winding to be charging at 115A so it will sag enough for the 82V winding  to start conducting much at all.

Better, but not much help.

[The_Armadillo]

Thanks for all the replies. I'm trying to get this figured out.

So now that I have the actual tranny in from of me I can give the exact listed values.

Primary:
0--------120V

Secondaries:
First winding:
47.6V--------0--------47.6V (4.3A)

Second winding:
41.2V---------0-------41.2V (3.8A)

My eltimate goal is for a class D amp sort of rig. Specifically, this one on eBay: http://bit.ly/18FB5Dn
even though the description says AC50V max, I asked the seller, and it was a mistake. Supposed to mean +/-AC50V.

Would another route be to buy two of the modules , use an active crossover and run one amp for high and one for low, then bridge them at the output? Could also give the option of stereo.

EDIT: Wrong link.

[R.G.]

OK.

Let's do some math.  You're going to feed some CT transformer into a full wave bridge, that into two filter caps. The ebay ad says "AC 50V", which I take to be 50Vrms. 47.6Vrms on each side makes +/- 47.6*1.414 = +/-66Vdc about.

You say the rating is 4.3A. Is that secondary RMS? or DC out?

If it's RMS amps, then the DC you can get out is 4.3/1.8 = 2.39A, because of the peak currents and RMS current heating from full wave rectification. So the DC available is 2*66*2.39 = 315W. For an honest 500W DC out, you need 500/(2*66) = 3.78A. If the rating is 4.3A DC output from a rectifier, you're golden.

However, music isn't pure sine waves, and is not constant loudness. It pulses up and down. This is know in the biz as the crest factor, the amount the peak loudnesses exceed the average power level. Transformers average their internal heating over periods of tens of minutes to hours. So if your music hits 500W peak but only 300W average (and that's HIGHLY likely to be the case) you're probably fine with your 300W transformer. The only good way to tell is to run it with 500W peaks and see if it overheats (usually, outside temp >130F).

As I keep flogging and you're finding out, a power amp is essentially a power *supply* and the power amp circuit almost doesn't matter.

There's another almost imponderable. Not using the 82V winding means it's not heating the insides of the transformer, and so the 90V winding can do MORE than it was rated for without dying. How much more? Uh... what's the thermal conductivity of layers of copper wire, paper, varnish... AGH!!

Using a second module and running in bridge is not a certain advantage. The second amp adds more voltage across the load, but has the same current flowing in it. This means that the "low power" amplifier side running from 82.4VAC@3.8A will limit the amount of current the "high power" side can put out, so the current limit with the additional amp in bridge is smaller than before. But you have twice the voltage, so running twice the speaker impedance would get more power.

You want a simpler, more effective way to get more AUDIO power? Use two different speaker cabs (which you'll probably need anyway at 500W), use a second amplifier module, but make an entirely separate +/-DC power supply from the 82V windiing and just run the second speaker on that power output, like a stereo amp, one amplifier of which has a bit more reserve power. The problems with hacking power supplies in series/parallel, bridging amps, etc. all now vanish.

[R.G.]

I figured out one of the things that was bothering me about that amp module.

If you buy that, mount it where the heat sink fins are vertical - that is, the PCB on it side, with cool air being sucked in the bottom by the hot air exiting at the top. As illustrated, the fins are choked almost to uselessness by being horizontal, and blocked on the open side by the filter caps.

Oh. One more. Notice that they recommend a lower voltage, lower power transformer for it. The recommended transformer is 300W. It makes me a little suspicious when a power amp seller recommends a smaller power supply than is needed to drive it to max output.

I suppose that is another display of my general twitchiness about how the heat gets out of anything electronic, especially anything with the word "power" in the product literature. Class D is a lot better than Class AB for heat dissipation, but the waste heat is not zero.

[The_Armadillo]

I'm not sure whether the winding current is DC or RMS, it didn't say. The power amp on that winding was 300W AB sorta circuit, if that helps. I want to be an optimist and say its DC.

Yeah, the recommended tranny was low.. Made me wonder too.

I think I may end up buying two anyways. Stereo is more fun.

I've been looking at this one too. Almost the exact same ratings, only difference I can find is price, some of the hardware and parts, and the heatsink. http://bit.ly/15heKjx

Now, if I had two mounted in a box with the tranny, couldn't I use a fan or two to force the cooling?

[R.G.]

I'm not sure whether the winding current is DC or RMS, it didn't say. The power amp on that winding was 300W AB sorta circuit, if that helps. I want to be an optimist and say its DC.

The crest factor argument might let you get away with that OK.

I've been looking at this one too. Almost the exact same ratings, only difference I can find is price, some of the hardware and parts, and the heatsink. http://bit.ly/15heKjx

Notice that this one is mounted so you *can't* get natural convection air flow. The PCB blocks this one.

Now, if I had two mounted in a box with the tranny, couldn't I use a fan or two to force the cooling?

Yes, you could. It's a question of objectives. Most musicians object to fans. A lot. There's the issue of dust collection, filters, cleaning, etc. to deal with too. Natural convection is by far less maintenance and more quiet.

It always bothers me when the seller says "I've tested it and it's good when it leaves here. After that you're on your own."

But then, selling high power electronics to anyone at all, I guess that's not too unreasonable, especially at slim profit margins.