Smoke from the secondarys

[RockinRocket]

Ok, so my biggest and really only serious amp issue I've ever had. Smoke and gunk came out the secondary's. Transformer is a RS Deluxe clone and can only be hooked up for one ohm (16 ohm in this case) at a time due to the multi interweaving windings. Anyways, powered on it was quiet, maybe too quiet but I plugged in and strummed guitar and got a weird grumbiling sound and smoke poured out the secondary end bell instantly. Powered off.

So in a JTM 45 with negative feedback did I simply wire the primary's backwards?

If I did then the wires were reversed at the factory. Probally the last time I use a Marstran. Had a wire color issue with a power I had to figure out

Thanks

[xtian]

Is this a brand new build, and your first power up?

[RockinRocket]

Yes new build. Voltages and everything else checked out before plugging in. Triple check everything before power up.

I did take off the end bells and drill new mounting holes. Maybe it shorted somehow on the reinstall but haven't gotten that far in diagnostics. Have ordered a new transformer for the customer as I'm just going to scrap anything that had smoke come out it!

[Reeltarded]

They are Heyboers. Brian just specs it. He has all the wiring on his site.. if there is a problem with the transformer it would be the first I had known of.

[R.G.]

In transformers with multi-interleaved sections that you have to reconnect external to the transformer, it is VERY easy to get one or more sections misconnected. If you get two sections fighting one another, it can easily generate smoke. I don't know if that's what happened, but I have had issues with this when first connecting some of my handmade transformers before. It's an easy mistake to make.

Likewise, for handmade transformers, it's easy for the manufacturer to get the wrong color wire on one or more sections. I know this from my days of hand making transformers, then putting colored leads on them, as above.

[RockinRocket]

Yes the multi-interleaved sections that needed to be connected for 16 ohm are correct.
I guess I should have hooked up or rigged up a test with alligator clips and measure resistance between setup for 4 8 and 16 ohms.
But that was too much work. Lesson learned on that.

Still the amp did produce sound, quiet non the less, so I'm thinking it was the reverse primary's at the factory.

[R.G.]

As you wish.

[RockinRocket]

I do wish because I'm not going to dismantle the output until I get the replacement.
I took photos of the wires after I soldered them and the joints were solid before I shrink wrapped them.

So RG it sounds like you are convinced the weavings are the culprit and not a reverse primarys?

[R.G.]

I'm not convinced of anything other than the amp has one of two issues. Those are either a shorted turn (or turns) where the smoke came from, or the amp oscillating strongly enough at a high enough frequency to both cause the smoke and remain inaudibly high. It could be either of these two, but on a new build, it's much more likely to be misconnection of the described complex many-section secondaries than it is to be an internal shorted turn. On a new build, it's less likely (although possible) to have an internal shorted turn or inaudibly high frequency oscillation.

The low signal level is a plausible result of getting one of the secondary sections wired up so it parallel-opposes one or more of the other sections. As I described, it is easy to make this mistake, with or without color coded lead wires. I learned this from both winding transformers (where all the leads are the same color - varnished copper) and from personally mucking up my understanding of both correct and incorrect color coding.

Every winding on a transformer has a polarity. Imagine attaching a 9V battery in series with a 12V car battery; if you put the 9V battery one way and measure the voltage across the two in series, you get 21V. The other way, you get 3V. In the first case, the 9V is hooked in series aiding, so the voltages add. Reverse the 9V and you get the 9V opposing the 12V, so the voltages subtract. Same with transformer windings; one way they add, one way they subtract.

This is CRITICAL for hooking windings (and batteries) up in parallel. Imagine the 9V and 12V car batteries again. What happens if you hook them up in parallel? In either [+ to + and - to -] or {+to -} the 12V can force a lot more current through the 9V than the 9V can oppose, so the 9V battery either smokes - or explodes. Now assume you're hooking up two 12V car batteries in parallel. When you hook them plus to plus, minus to minus, they have (nearly) the same voltage, so not much if any current flows between the batteries. If you hook them plus to minus, they fight it out to the death trying to force the other battery to change voltage.

Same for transformer windings. Hook two windings in parallel and they'd better be identical polarities AND identical voltages. If not, they force currents limited only by the copper wire resistances to flow, and smoke comes out as they duel to the death.

That's the whole point of color codes on windings - to let you pick which windings are isolated from each other, and, critcally, to let you get the phases right.

There are ways to use either an oscilloscope or a DMM to figure out proper phasing and voltage.But you have to disconnect all of the windings to do this.

[Stevem]

One way to test the OT is to input 10vac across the primary, if the OT is for a pair of 6L6 tubes then you should read .62 volts out the 16 ohm tap, if it's for a pair of 6V6 tubes then you should read .44 volts out the 16 ohm tap.

These readings should read true within .08 volts out the secondary pretty much.

If you truly want to test out the integrity of the OT and have a veriac then you can input 100 volts into it and hopefully read about 62 volts out of it.

[sluckey]

If you truly want to test out the integrity of the OT and have a veriac then you can input 100 volts into it and hopefully read about 62 volts out of it.

That should be 6.2v, not 62v.

[R.G.]

Just thinking about the voltage ratio (and good point, Stevem!),

A pair of 6L6s is typically used at 4400 ohms primary in guitar amps. The output load is generally 4, 8, or 16 ohms. The impedance transformation ratio is the square of the voltage ratio. So 4400 to 8 ohms is 4400/8 = 550; the square root is 23.45. For 4, it's 33.166 and for 16 it's 16.583. In a 4400:8 ohm output, 10Vac on the full primary(that is, plate to plate) gives 10Vac/23.45 = 0.42Vac. For 4 ohms windings, 10Vac on the primary gives 0.301Vac, and for 16 ohms secondaries, you get 0.603Vac.

You're really, really close at 0.62 vac for the 16 ohm winding!

[RockinRocket]

Ok finally got the replacement RS output. Smoke didn't come from the secondary's but from the shrink wrapped extra primary taps... Must have done a poor job at isolating them . This output has taps for 6k6,8k,9k and ultra linier taps. I used the green pair for 6k6.

I'm assuming they simply shorted and that was the issue?
Transformer probably no good?





Marstarn RS 16 ohm schematic-
http://marstran.com/Wiring%20Diagrams.html