Quad II - PT autopsy begins...

[dougietamson]

Late model Quad II, serial 50053.

Bought-as-is, blowing fuses with no valves.
I desoldered all wires to the turrets and removed the PT to the bench.
The 240 mains pri reads at 12? dcr and 1.96H, I connected it to a light bulb limiter with no load and light up very bright.

The 320-0-320 reads 68?-CT-72? with total 640v at 140? and 3.73H-CT-3.74H and 14.7H total.

5v and 6.3v read 0? and 0H on my DMM, I'll check with a sig gen and the scope later.

It's potted in bitumen, time to get it out. I didn't go too hot in the oven, 120c,
for 1hr and it slid out with moderate encouragement.

Next step is to clean off most of the remaining bitumen once it's cooled enough to handle.

Some pics attached.

[dougietamson]

I've just checked it with a 10vpp 100hz sine input to the pri 240 tap.

The readings look normal, though low voltage.

On the 240 pri tap and input 10.8vpp I get 27.4vpp at the 640v tap
On the 220 pri tap and input 10.8vpp I get 30vpp at the 640v tap
On the 200 pri tap and input 10.8vpp I get 33vpp at the 640v tap

Maybe try it on the variac and bulb limiter next.

[lesmw0sec]

Going by your earlier tests, I'd reckon that that transformer may be what we in the trade refer to as knackered!

[G6Tanuki]

Shorted turns won't necessarily reveal themselves to a low voltage from a DMM, inductance meter, or a test with a signal generator and scope.

[Ed_Dinning]

Hi Dougie, probably shorted turns on the HT winding, so all should be rewound including Pri as it will have been stressed.

When reassembled it is worth fusing both sides of the HT winding as a valve A-A short will kill the transformer if the fusing is only in the CT lead

Ed

[dougietamson]

Originally Posted by Ed_Dinning

Hi Dougie, probably shorted turns on the HT winding, so all should be rewound including Pri as it will have been stressed.

When reassembled it is worth fusing both sides of the HT winding as a valve A-A short will kill the transformer if the fusing is only in the CT lead

Ed

Thanks Ed, autopsy to follow to find where the short is. Once cleaned up I'll warm it and ease the lams out.
I'll unwind each layer and counting the turns as I go. The low cost hand winder is great for keeping count.

Any suggestions for the spec of the new wire?

[kalee20]

If you have shorted turns - particularly just a few shorted on a low-current winding (that's wound with fine wire) - you may see the voltages looking 'reasonable' at the appropriate ratio, when you energise the primary by a low-voltage supply that's got a fair bit of grunt behind it.

However, at full voltage, it will develop hot-spots and suck excessive current.

All that is consistent with what you are seeing.

It's going to be an interesting strip-down... thanks for the photos so far!

[dougietamson]

Wee update before the strip down.

I thought I'd just take it up to 240 on the variac and a singe 60W bulb in series, no load, but with a previous test at 240 and no load the blub was full on bright.
All looked good, bulb stayed dim and the HT winding measured 660v (no load)

Maybe something during the un-potting brought it back to life.

I'm still going to do a full rewind as I'm not risking my pair of grey glass GEC KT66's

[dougietamson]

Also of note for the low voltage test, it didn't 'load down' the 10vpp input.

I have a 2nd Quad II with a PT issue, I'll run the same tests tomorrow.

[Richardgr]

One failure condition with quads is the choke sinking in its can due to getting hot and shorting to earth. Could that be the case here? I.e. PT OK? It looks like it could be possible via the rectifier heater winding.

[kalee20]

I don't think the KT66's would be jeopardised by a dodgy mains transformer, except perhaps if it spat flaming pitch at them!

Though at the same time, a proper rewind using good quality modern enamelled wire should result in a blameless transformer.

[Radio Wrangler]

You've already seen enough to know that you can't trust it any further. If there are shorted turns, the heating can move bitumen and oily solvents around within the windings and re-insulate two previously shorted strands.

Run the amp, get it warm and you shake the dice again. There is only one way to get it right and to know that it's right.

David

[trobbins]

Here's hoping you haven't 'jumped the gun' based on just the initial light bulb test. Can you elaborate on what is meant by "The 320-0-320 reads 68?-CT-72? with total 640v at 140? "

Are you able to measure the mains current while bringing the primary up on a variac. I'd anticipate the current would become excessive as you approach 240Vac - although the best comparison is with measurements against a good PT.

Do you have a megger/ insulation resistance meter, to assess any other fault mechanism at say 1kV.

[Cobaltblue]

For the ? read the ohms symbol. Doug clearly didn't realise that the forum software has stopped accepting special characters some time ago.


Cheers Mike T

[dougietamson]

Originally Posted by trobbins

Here's hoping you haven't 'jumped the gun' based on just the initial light bulb test. Can you elaborate on what is meant by "The 320-0-320 reads 68?-CT-72? with total 640v at 140? "

Are you able to measure the mains current while bringing the primary up on a variac. I'd anticipate the current would become excessive as you approach 240Vac - although the best comparison is with measurements against a good PT.

Do you have a megger/ insulation resistance meter, to assess any other fault mechanism at say 1kV.

The initial light bulb test was just a gauge of how much the PT was drawing with no load connected, a dim glow on the bulb could be in normal range for an old transformer but the bulb (60W) was full on.

Sorry about the omega ohms symbol, it's handy on my macbook just to hit the option key and z for the symbol, it reads ok when typed and using the preview post button before hitting submit reply button but seems it's not displayed in the final post.

I didn't measure the mains current before taking it out of the pot, my clamp meter was sitting a couple of feet away! I should add a meter to the bulb tester.

I have an AVO Megger BM403/2 that does 250, 500 and 1000v and test after Ive had my 2nd morning coffee.

[trobbins]

Yes it's interesting to diagnose a bad transformer that doesn't show itself simply as an open winding. I haven't used the 'Q' or transient ring type test for shorted turns - that could be relatable to say an impedance frequency plot of a winding - but I guess that would also need a comparison to a known ok PT.

Have you been able to compare your measured winding resistances to another ok PT sample, given some allowance for differences due to copper tempco ?

If there was a shorted turn, such as a side-by-side turn, then that could be quite difficult to discern in a tear down unless the very local region showed heat stress that was markedly different from any general heat stress of the winding.

[kalee20]

Originally Posted by trobbins

If there was a shorted turn, such as a side-by-side turn, then that could be quite difficult to discern in a tear down unless the very local region showed heat stress that was markedly different from any general heat stress of the winding.

Yes I'd agree - a single turn shorted in the HT winding wouldn't suck enough power to seriously disrupt operation - though it could cause a hot spot, and make the lamp limiter glow gently.

A shorted turn in a heater winding - or an interlayer short of maybe 50 turns - would be a different matter, possibly causing fuse blowing, or limiter lamp being much brighter.

One possible approach, might be to run the transformer from a Variac, via fuse or lamp limiter, for a significant time (supervised!) so that the faulty bit gets hot enough for long enough to give obvious heat discolouration. Down side is that as this transformer is pitch-impregnated, everything is going to be black and sticky anyway.

[dougietamson]

Originally Posted by trobbins

Yes it's interesting to diagnose a bad transformer that doesn't show itself simply as an open winding. I haven't used the 'Q' or transient ring type test for shorted turns - that could be relatable to say an impedance frequency plot of a winding - but I guess that would also need a comparison to a known ok PT.

I have another Quad II, it's transformer is not 100% though, not blowing fuses but gets quite bright on the bulb limiter on the variac at 120v.

The de-potted PT is measuring 320MOhms on the megger HT sec to Pri 240, 2GOhms 5v to 6.3v.

The 2nd PT (still in its pot) is measuring 225MOhms HT sec to Pri 240, 2GOhms 5v to 6.3v.

On both the CT for the HT, labeled 'E' is also a CT for the 6.3v.

In preparation for getting the 2 Quad II's up and running I bought a 320-ct-320 5v 6.3v PT listed as 150W from China, though it only has a 220 mains tap, I've stripped it down, seems well made, calculated the t/v and added turns for a 240v tap. I can use this as a good benchmark. I've tested it on the megger at 1000v and it's >10GOhm between all winds.

[trobbins]

Originally Posted by dougietamson

I have another Quad II, it's transformer is not 100% though, not blowing fuses but gets quite bright on the bulb limiter on the variac at 120v.

Would that be similar to your now unpotted PT ?

If so, then perhaps if you had time, you could measure the mains primary current with increasing voltage. To me, that could be the better characteristic to use for comparison with a 'good' original PT, and also for comparison with the unpotted PT. For an unloaded PT, the primary current waveform would likely be quite distorted, but an rms value could still be a good comparison metric.

The other valid metric would be 'joule heating' as previously mentioned.

[dougietamson]

I've just checked the current drawn by the 2 QuadII PT's.

The un-potted one draws 70mA at 240v set by the variac.

The other one (has bitumen oosing out of the 6.3v turret - see photo) draws a lot more so I just stopped when it hit 500mA (90v on the variac)