Power transformer troubles

[BiBi]

Hello,

I recently purchased a Fender Twin Silverface with the Ultra Linear taps. The condition wasnīt so good and the price was right so I was looking forward to doing some repairs and getting it back in order.

While sorting out some of the hum issues I was tightening and straightening out the filament wires. Unfortunately, I turned on the amp before checking if any of the pins on the 6L6 were touching. When I turned on the amp, I heard a loud sizzle then hurriedly turned the power off. It appears that pin 2 and pin 3 on one of the 6L6 were touching due to the length of wire coming out of pin 2. My straightening the wires had pulled the pins together to make enough connection.

The result is that the amp makes only wild oscillations-- I would thereby assume the OT is ok as it still makes a sound. The HT voltage has stayed roughly the same (470VDC) while the filament voltage is sagging. When unloaded, I am getting 6.3VAC while with the full complement of tubes in the voltage droops down to 5.6VAC. I believe I have ruined the power transformer in some way but I wonder if it is still salvageable by removing the filament wiring and installing a new 6A (4x6L6 + 6x12AX7etc = 5.4A) filament transformer. Am I courting serious disaster by doing this? Or should I simply try to replace the power transformer first (although there is no easy aftermarket alternative.)

Thank you for any help with this.

Bill

[R.G.]

While sorting out some of the hum issues I was tightening and straightening out the filament wires. Unfortunately, I turned on the amp before checking if any of the pins on the 6L6 were touching. When I turned on the amp, I heard a loud sizzle then hurriedly turned the power off. It appears that pin 2 and pin 3 on one of the 6L6 were touching due to the length of wire coming out of pin 2. My straightening the wires had pulled the pins together to make enough connection.

First, my condolences on the mishap. This kind of thing is always traumatic.

Second, it may not be the PT that has a problem. A previously-good PT might withstand a filament short for a short time without damage, and the oscillations may be something unrelated.

Third (and a bit guiltily) thanks for providing a real-world example of a situation that's one of my hot buttons. More about that later.

The result is that the amp makes only wild oscillations-- I would thereby assume the OT is ok as it still makes a sound.

This is a confusing symptom. I would not expect a heater short to have any effect whatsoever on the stability of the amp. It is conceivable that one or more tubes don't run well at the 5.6V you measure for heater voltage, and that's the cause of the oscillation, but that's even a bit unusual. Did the amp run without oscillations before the heater rewire job?

It is also possible that there was a second side effect of the heater rewire that caused an unrelated oscillation. One simple one would be reversing the two OT primary leads.

The HT voltage has stayed roughly the same (470VDC) while the filament voltage is sagging. When unloaded, I am getting 6.3VAC while with the full complement of tubes in the voltage droops down to 5.6VAC.

Did you happen to measure the filament voltages before the incident? 5.6V is low, but not disastrously so.

Transformers fail in two main ways: open windings and shorted windings. Open windings are easy, just look for output voltages or use an ohmmeter to check for opens and shorts between windings and between windings and core. Internal shorts are harder to find, because a shorted turn doesn't necessarily have much effect on the winding resistance.

It is plausible that an external heater winding short could overheat one turn enough that the insulation fails and leaves one turn shorted internally. However, the result of an internal shorted turn is more serious than just a lower voltage. It actually is a massive overload on the whole transformer, and would likely clamp the high voltage to be seriously lower than previously. The lower voltage is possible, but an unlikely result of an internal shorted turn. Maybe there's some other problem, like de- or un-soldering the lead connection to the filament output wires causing a high resistance ( this is relative; I make it about 0.128 ohm from the measured values ) to appear in series with the heater output.

I believe I have ruined the power transformer in some way but I wonder if it is still salvageable by removing the filament wiring and installing a new 6A (4x6L6 + 6x12AX7etc = 5.4A) filament transformer. Am I courting serious disaster by doing this? Or should I simply try to replace the power transformer first (although there is no easy aftermarket alternative.)

It is possible that this one is fried; however, it's also possible it's fine and there is another problem.

I would do this:
1. Find out if it's internally shorted; I would do this either by (a) removing all secondary loads, including tube heaters and the HV rectifiers, and just seeing if it overheats or pops the primary fuse over the course of maybe 8 hours sitting powered but unloaded; or (b) using an ohmmeter to look for open/shorted-together windings and the neon-bulb pulse test to see if there's an internal short. Option a also tells you if it's safe to use a separate filament transformer.
2. If it passes 1, hook it all back up and see if you can track down why it oscillates. In a previously-working amp, this is nearly always a wiring fault, bad resistor or cap, or a going-bad tube. If you touched the OT wiring at all, I'd check for reversing the OT primary leads, which will give oscillation in any amp with feedback from the output.

and finally
3. When you get it back to working, put a 6A slow blow fuse in series with each side of the heater winding. This would have prevented your current predicament, as one of the fuses would have popped. This is a little-known item, but it's pretty solid protection for heater wiring issues, including an internally-shorted tube. The math says that a shorted heater will sit there and overheat the windings enough to burn out and not pop the primary fuse. This is exactly what you report. There are people who refuse to believe this, even when shown the math; even people who should know better. It's cheap protection for the problem.

So do a bit more snooping before going off to power transformer replacement. Find out if it overheats when it's completely unloaded, or if it fails the pulsed-neon test. If it passes both of these and the voltages are OK, the PT is still good. The only reason I am being this cautious is that yo report a lower heater voltage under load.

[Steve Conner]

Pin 2 is heater, pin 3 is plate. So you didn't short out the heater winding, you applied plate voltage to it.

You will probably have burnt out any hum balance pots or resistors, and possibly fried a few tubes by exceeding their heater-to-cathode voltage rating. The preamp cathode resistors and their bypass caps may also be damaged.

Fix all that and the oscillations should be gone

[BiBi]

Thank you both for your assistance it is much appreciated!

I have removed all the secondary wires from the power transformer and it is running now unloaded.
I made notes this time of the voltages and resistance between the secondary wires:
HT-- both sides are around 4 ohms to center tap and roughly 175VAC so that seems ok.
Filament-- .3 ohms resistance and 6.4VAC. I seem to remember this was much higher unloaded before, perhaps 7 but I didn't make more than a mental note.
Bias-- 3 ohms and 51VAC so this is also in spec.

I measured resistance between all the secondary wires and there was nothing out of the ordinary. I also replaced the worn out original 6L6 sockets (after the troubles though) and put 10 ohm resistors from cathode to ground as a precautionary measure. When I applied power one of the resistors burnt up so i suspect that indeed one or more of the 6L6s is defective. I still have the original tubes so I will continue to measure on these after running the transformer for some hours.

I like the idea of putting fuses in the filament supply as I've also seen a power transformer that was burnt when the connections on the indicator light shorted.

Thanks for the help and I will report later today on how it goes.


Bill

[EFK]

Filament-- .3 ohms resistance and 6.4VAC. I seem to remember this was much higher unloaded before, perhaps 7 but I didn't make more than a mental note.

Did you check your wall voltage then vs. wall voltage now? Reason I ask is that my wall usually cranks out 124-125 VAC on average through the fall, winter and spring, but it drops every summer. Here in PA it has been pretty hot and I guess everyone is running their air conditioners, so every day I watch it start out around 117-118 VAC and by noon it it down to 114-115 VAC. This lowers voltages pretty dramatically throughout any amp.

I actually had to rewire my "vintage voltage adapter" (thanks RG) to ADD 6.3 volts rather than drop 6.3 volts!

[Chuck H]

The OP mentions that the amp was in poor shape. Which is not very specific. But "tightening and straightening out the filament wires" seemed necessary. Were they not original? If the filament voltage wasn't measured before the incident, and those leads arent original, I might wonder if the wire is too small. With twisted pairs and so many tubes there is a lot of lead length. This occurs to me because the unloaded voltage is in spec. If the low voltage was a result of shorted turns the drop wouldn't be relative to loading. Though there IS always some drop, and it's mentioned that the before incident voltage MAY have been around 7V. Still, have the filament leads been replaced? And, if yes, might they be too thin?

[R.G.]

Did you check your wall voltage then vs. wall voltage now? Reason I ask is that my wall usually cranks out 124-125 VAC on average through the fall, winter and spring, but it drops every summer. Here in PA it has been pretty hot and I guess everyone is running their air conditioners, so every day I watch it start out around 117-118 VAC and by noon it it down to 114-115 VAC. This lowers voltages pretty dramatically throughout any amp.

I actually had to rewire my "vintage voltage adapter" (thanks RG) to ADD 6.3 volts rather than drop 6.3 volts!

Good point. And WOW - your AC actually drops down to 115!

One thing we have to watch for here is that the AC voltage is set to about 125 nominally, drops to maybe 123 under full airconditioning load. On days with hot middles that drop to cooler nights, people (or likely, their thermostats) turn their air conditioners off in the early evening and the voltage shoots up. I've seen 135 once.

I agree - the OP needs to measure the incoming AC to figure out what the transformer ratios need to be. The value for 6.3V with no load is suspicious, as the open circuit voltage on these things is usually more like 6.8 to 7 because of the drop under load.

It's also hard to measure low resistances without special low-range ohmmeters. If the measured 0.3 ohms is correct, the 5.4A load on the tubes will cause the output voltage to sag by 5.4*0.3 = 1.62V. That can't be right, as an OC voltage of 6.3 minus 1.62V is 4.68V, less than is actually measured. I'd bet the voltage measurements are correct instead of the resistance reading. In fact, the better way to measure the resistance in this case is to measure the unloaded and loaded voltages and the AC current then get the resistance by calculation.

@Bibi - measure the incoming AC voltage. Carefully~!

[BiBi]

Thanks for all the ideas here...

As is, I have a unique situation in regards to AC power. For one, I live in Germany, the other is that my workspace is in an industrial building that has it's own generator (it also sells electricity back to the power company.) The AC voltage in the building stays very steady at around 225VAC all the time there isn't any industry left in the building to change it.

As is I have been using the 240V tap on the transformer as my HT was going over 520 volts on the 220V tap fully loaded and I wanted to go a bit easier on the tubes.

I agree that the filament situation is strange! Per Chuck H, I straightened out the wires because they were really haphazardly placed and I was having some hum problems-- I kept the original wires though.

The power transformer has now been on for over 8 hours and is warm, certainly not hot to touch.
Now I get the following readings:

on 240VAC tap--

filaments 6.4VAC
HT 174-0-174VAC

on 220VAC tap--

Filaments 6.7VAC
HT 189-0-189VAC

Next step will be connecting the secondary connections again...

Bill

[Tom Phillips]

...Now I get the following readings:

on 240VAC tap--
filaments 6.4VAC
HT 174-0-174VAC

on 220VAC tap--
Filaments 6.7VAC
HT 189-0-189VAC
...

Well now we have another reading that doesn’t make sense. Your HT voltages are measuring way too low. Normal would be more like 330-0-330 ą~20V. In your first post you reported that your B+ was 470V. You would not get 470VDC if your transformer HT secondary was putting out the low VAC readings that you posted.

I suspect that your test meter may be giving incorrect readings. Do you have access to another meter to compare it to?

Regards,
Tom

EDIT: I just remembered that the Twin with Ultra linear taps uses a tapped bridge rectifier. Given that fact, your HT voltage readings appear to be within normal range. (The 189-0-189VAC format made be think of a FW rectifier). Sorry for the false alarm.

[R.G.]

As is, I have a unique situation in regards to AC power. For one, I live in Germany, the other is that my workspace is in an industrial building that has it's own generator (it also sells electricity back to the power company.) The AC voltage in the building stays very steady at around 225VAC all the time there isn't any industry left in the building to change it.

As is I have been using the 240V tap on the transformer as my HT was going over 520 volts on the 220V tap fully loaded and I wanted to go a bit easier on the tubes.

OK, this explains some things. The filament voltage is low because you're feeding from a too-high tap. It reasonably should use the 220V tap.

I agree that the filament situation is strange! Per Chuck H, I straightened out the wires because they were really haphazardly placed and I was having some hum problems-- I kept the original wires though.

I wonder if the hum problems came from the unusual AC power wiring. Is your AC power properly grounded? Many generator outputs are not. This could represent not only a hum problem, but also a safety issue. You need to get that checked out.

The power transformer has now been on for over 8 hours and is warm, certainly not hot to touch.

That means that there is not an internal short. It would overheat quickly if there were. If it gets merely warm after 8 hours and the voltages are close to correct, it is very likely that the PT is OK.

Now I get the following readings:
on 240VAC tap--
filaments 6.4VAC
HT 174-0-174VAC

on 220VAC tap--
Filaments 6.7VAC
HT 189-0-189VAC

The 220Vac tap is the one to use.

When you connect things back up, connect up the filaments first, and measure the AC volts on the heaters with all tubes in. It should be 6.3V +/- 10%.

[Chuck H]

And AFAIK most modern 6l6's can be operated safely at over 520Vp. I haven't done it. But I've read several reports here at MEF. Even if you expect to run the amp hard. Though with a UL Twin I don't know that even an industrial park is a non obtrusive place to do that.

[BiBi]

Thanks a lot for all the replies.

I will try to run the amp on the 220V tap then. I am still concerned as to why the filament voltage has dropped so much but there doesn't seem to be a clear answer does there? Maybe this will be moot when I put in the old 6L6s, at least I hope so!

I won't be able to get to the amp again until Monday but will give a full report then. As I will have to adjust the bias supply to accommodate the higher voltage I wonder if it is worthwhile to use full wave rectification (the original simply grounds one leg of the secondary)?


Bill

[BiBi]

Ok now things are getting strange.

I reconnected the secondaries of the power transformer and installed the old output tubes. I am still using the 240 tap since I havenīt adjusted the bias supply as of yet. I am getting around 470 HT which is normal, and the filament voltage rests at around 5.9VAC with the four output tubes and the 12AT7 PI. When I put a sine wave on the input grid of the PI the filament voltage rises to 6.3VAC! As well, there are still only very ugly oscillations from the output...

I am going to follow the tack suggested by Steve Conner and try replacing some of the cathode resistors in the PI, but is there any good explanation as to why the filament voltage would actually rise?

Thanks for any help on this perplexing problem!


Bill

[Steve Conner]

As I suggested, the first thing to do is check all of your tubes for heater-cathode shorts. The easy way to do this is to substitute them for known good ones.

The filament voltage would rise because current was dumping into the heater circuit from the B+ through a damaged tube. More likely RF current from the oscillations that just confuses the meter, rather than actually raising the voltage, which would take a huge amount of current.

Or one of the filaments is going open circuit intermittently, again because it was damaged by a heater-to-cathode arc.

Or the heater wiring got burnt and is now failing to make contact somewhere. Perhaps at the pin 2 where it arced?

[BiBi]

I used the original 6L6s for this test-- they were working fine beforehand. I can put some other ones in to see if there is a difference. The 12AT7 is new so it should be ok.
The burning of the heater wiring is interesting. I can replace it with some new wire but, after the fault, I did put in new sockets for the 6L6s thereby resoldering the connection to pin 2.

Thanks for the ideas!