Testing an output transformer

[lowell]

I have a crate palomino here that has a bad speaker output. the line out is good, all pins/voltages checkout... pre and power tubes have been replaced - bias is good. The scope is showing good signal on the output but the amp is low volume and distorted. I did not understand why the scope shows it good but it sounds awful, then I saw 2 signals at certain frequencies and at lower settings on the clean channel volume. I'm wondering if there is some phasing issue here. If I don't have the volume knob turned up over half way there seems to be 2 out of phase signals on the output. This goes away on the scope when the volume is turned way up. Now I suspect the OT since it seems everything else checks out.

I measured primary resistances and they seem ok 64ohms/58ohms.
I measured secondary resistance at .7ohms.
Now, I'm not sure that resistance ratio is the same as impedance ratio. Actually in thinking about it, impedance ratio is turns related and resistance is not so much. For sh's and giggles I calculated it anyway.

It seems with 4 el84 in p-p the ratio of primary to secondary should be 8k/8 for an 8ohm load. My resistance math doesn't work, but I think I'm wrong in assuming resistance is going to work for this.

I guess I need to inject some signal to see.

Ok so I set my signal gen to .5vac peak-peak at 1khz. I disconnected all OT primaries and secondaries. I connected the .5vac to the secondary and double checked that it was .5vac. It was much lower when connected to the secondary so I readjusted it. My reading on the secondary is 10vac peak-peak.

The ratio of primary to secondary is 10/.5 or 20/1. Shouldn't 4 el84 in push-pull be an 8k primary? One side being 4k and the other side being 4k? Or is only 1 side supposed to be considered in the calculation?

Finally the ratio is 20/1=20. 20x20x8=3.2k which seems reasonable. I don't think the OT is bad. Anyone have any advice on what could be wrong w/ this amp?

[defaced]

Resistance doesn't mean anything, you need to inject a signal into the secondary, read the voltage across the loaded secondary and open primary, and calculate turns ratio.

RG posted an OT diagnostic tool the other day. Here's the thread:
http://music-electronics-forum.com/t17678/

[lowell]

Ok yeah that's the conclusion I came to, and I measured it and it seems good.

Any ideas what to look for in this amp? Here's a schem too.

[Jazz P Bass]

OPT test

I just went through this with a customers Swart Space tone (Champ knockoff)
Resistance checks where good.
Did the applied AC thing. Looked good.
The problem was that I was getting a 0.6 volt output signal. Max!
I had a real healthy signal on the grid.
Subbed in an old Princeton transformer (did not connect the middle wire) & up came the output voltage. 6 Vac.
So what did I learn.
When everything looks good, except the output signal, it pays to have a spare OPT laying around.

[MWJB]

"Ok so I set my signal gen to .5vac peak-peak at 1khz. I disconnected all OT primaries and secondaries. I connected the .5vac to the secondary and double checked that it was .5vac. It was much lower when connected to the secondary so I readjusted it. My reading on the secondary is 10vac peak-peak.

The ratio of primary to secondary is 10/.5 or 20/1. Shouldn't 4 el84 in push-pull be an 8k primary? One side being 4k and the other side being 4k? Or is only 1 side supposed to be considered in the calculation?

Finally the ratio is 20/1=20. 20x20x8=3.2k which seems reasonable. I don't think the OT is bad. Anyone have any advice on what could be wrong w/ this amp?"

400ohms times 8ohm speaker output = 3200...sounds feasible.

I'd also check that with 0.5VAC on the secondary, that you are getting 5VAC each side on the primary, when measured to the CT...a big imbalance might be the result of some shorted turns?

The DCR test is of some use, a little mismatch is not a problem here, one side of the primary winding has more wire in it than the other.

How are the dc voltages for power tube plates, screens & cathodes, likewise for preamp plates & cathodes?

[lowell]

How are the dc voltages for power tube plates, screens & cathodes, likewise for preamp plates & cathodes?

all those had measured fine.

I tried a sub OT and voila. Let this post be a lesson that even after testing an OT, one should try a substitute to be sure.

[Jazz P Bass]

OPT Test

Rock On!

[J M Fahey]

Yes, what we are testing is the output transformer "job" here, not optimizing the power stage.
Even a humble 15W OPT can be used (if nothing else available) to test, say, up to a 100W amp, even mismatching impedance.
We don't aim for the full 100W ; 10 to 15 clean watts with it compared to 1 to 4 W with the original one, show that the original one has shorted turns and that output tubes, bias, PI, are at least not grossly wrong.

[Enzo]

Amen. I have an old Fender 50w Bassman OT on the corner of the bench. I clip it in as a substitute OT in ANYTHING.

As JM says, it isn't about sounding good, it is just about finding out if the original part is bad.

[R.G.]

I'll defer to Enzo's bench experience any day.

Just as some comments on testing OTs and about OTs:
- To completely test a transformer that once worked:
1. Use your ohmmeter to see if the windings are open where they should be connected.
2. Don't relay on measured resistance to tell you good or not. Resistance of windings tells you little or nothing about either correct operation or matching other than open or not. The variation in resistance for non-interleaved / sectionalized windings from layer to layer is bigger than the variation caused by either an adjacent turn short or an adjacent layer short. The only failure where this would help is a short from a lead coming out to an outer layer.
3. Use your ohmmeter to see if the windings are shorted one to the other, or to the core. Core shorts do happen, and also happen in chokes. The transformer can work fine if it's out of the chassis if it has a core short.
4. Use the neon-bulb kickback trick to see if there is a shorted turn inside the trannie.
5. Tests one through four get the entire possible set of transformer flaws except the following: (a) intermittent shorts/opens which only happen when the transformer is hot enough and (b) intermittent shorts which only happen under very high voltage. It is possible to make a tester for (b) where you look for breakover voltages of bigger than the breakover of a single neon bulb, but you need to be careful that this can't itself cause a punchthrough voltage on adjacent wires.

The failures from item 5 are very rare.