Putlec, Peerless S217, transformer rewinds and motorboating

[njm]

Got given an old EQP-1 to sort out, it was putting out a popping/crackling sound. Output transformer was on its way out. Sent it off for a rewind. Transformer came back. Wired it in. Powered it up. Smoke! A couple of wires were reversed on the primary and tertiary windings, took out the 1k on the centre tap of the tertiary winding.

So I sent the transformer off again. Wired it back into the Pultec, no smoke this time! But it was squealing around 16khz, whichever way the feedback winding was connected. Bypassed the negative feedback winding by connecting the 360ohms feedback resistors straight to the 1k on the centre tap, then that amp was stable but obviosuly not unity gain. I scoped the output of the feedback winding, checking freq response and phase, -3dB at 6.6Khz with 90° phase shift, turning a full 180° by 27khz.

So the transformer went back again. They realised the previous fault would have put DC current through the windings  and probably demagetised the laminations. So they pumped 50Hz through it to sort that out, primary inductance went from 18H up to 24H. Got the transformer back, put it in the Pultec. Now it's motorboating!

Now I'm pretty confident there is nothing wrong with the Pultec itself, the unit was previously operating fine appart from the crackling noise. The only part I've changed since then is the 1k resistor that burnt out. I've checked all the grounds, bypass caps, resistor values, valves.

Anyone know what the inductance of the windings of this transformer should be? I'm kicking myself for not taking measurements before sending it off for a rewind! Anyone got any ideas of where they'd go from here? I'm starting to question my own sanity!

 

[njm]

The transformer frequency response/phase and voltage ratios are all fine now. These are the readings I'm getting for the windings (measured end to end/series)
Lpri 26H
Lsec 2.5H
Ltert  57mH
Rpri 400r
Rsec 17r
Rtert 3r4

 

[radardoug]

Motorboating is often caused by the decoupling caps on the ht lines having lost capacitance, allowing lf signals to propagate through the power supply.

 

[CJ]

sorry to hear of your transformer hassles,

those feedback windings are tricky to get right,

what kind of meter are you using to check inductance?
probably a meter that uses 1000 Hz.

there are 1575 turns on each primary so 3150 total.

inductance for a  7/8" stack of 12 L lams is  u x N^2 x 1.86189^-9

so for 3150 turns you have L=1.84746^-2 x U

don't know the exact perm of that alloy, probably around 40 K

so primary inductance s/b 40,000 x 1.84746^-2 = 740  Henries

but that is for a core that is running at about 20 Hz, so as you go up in frequency, that number will decrease quit a bit.

 

[njm]

Thanks chaps!

CJ - yes the meter I have tests at 1k, and that's reading the values I put in that other post.

Radardoug - yes I've seen that before. I've replaced all the decoupling caps, hasn't made a difference.

Ah this thing is so close to working! If I take AC feedback from the output winding rather than the tertiary winding, the amplifier is stable. So in my mind, this would suggest this is a problem with the tertiary feedback winding

 

[Bo Deadly]

The 1K resistor that you replaced is rather important since it's the cathode resistor for both input triodes. Are you sure that was restored 100% correctly? Are you seeing the voltages listed on the schem?

 

[gyraf]

..make sure to phase the feedback winding correctly.

Get motorboating, first thing you do is reverse polarity of feedback winding.

/Jakob E.