21H leslie problems

[Guest]

Its a 21H amp....ive included pics....i tested the resistance on the choke...it was at 64 ohms....And it does have the OLD caps in it

[Guest]

A few more pics....

[David Anderson]

Ok, quick visual evaluation:

This is a mid-1950s 21H amp that should have four 6V6 output tubes and two 5Y3 rectifier tubes in it.

Someone put General Electric 6L6GC output tubes in it for some reason. They look fairly high-mileage, and they draw more heater current than the power transformer is rated for. I'd recommend a matched quartet of JJ 6V6s. You didn't include a photo showing the rectifier tubes.

Two of the original wax-dipped paper capacitors have been replaced with brown epoxy-dipped Mylar capacitors between the 6SN7 and the output tubes, which was a good idea. Three of the original paper capacitors remain, the tan, tubular, waxy-looking things, two at the input and one on the motor sockets. The remaining paper capacitors should all be replaced with modern polypropylene types like Mojo Dijon or 716P orange drop. The Mojos are axial and fit better at the input. A shorted capacitor on the motor socket will make the motors run all the time, regardless of relay position.

There should be a 0.0047uF capacitor (0.005uF on the schematic) across the output transformer primary winding. It is missing from this amp, but it needs to be there for output stage stability and to limit voltage swings at frequencies above the output transformer's bandwidth. It needs to be rated at least 1,000V, preferably 1,500V or 1,600V. Leslie service manuals warn against removing this capacitor.

Two 220k resistors attached to the 6SN7 input socket were replaced with 270k resistors for unknown reasons, probably because someone was out of 220k resistors and only had 270k. This will reduce the input signal slightly. The 10k wirewound resistor dropping voltage to the motor relay is supposed to be 15k, 2 or 3W. (One early schematic says 22k, but I always see 15k in actual amps and on the schematic on the side of the amp.)

The basket of the original Jensen F15LL woofer is gone, but the field coil remains, which works as part of the amp. Presumably, the original woofer's voice coil or cone was damaged, so someone made the criminally insane decision to toss it and throw in an Eminence woofer, which is considerably inferior to the F15LL. Why people don't get the F15LLs reconed (easy to do) makes my head hurt, but some people think "new" is always "better." Maybe you could find an F15LL basket missing its field coil and have a speaker reconer put it all back together.

If the original filter can capacitor needs to be replaced, the CE Manfacturing 30/30/30/30 @ 475V is a decent choice. If you get the 30/30/30/10 version, the 10uF section can end up in the wrong place because the orientation of the capacitor socket is different from later Leslie amps that use the 30/30/30/10 can.

[Wes]

If you order the can capacitor from Rick at tonewheelgeneral.com, you can ask him to cut the tab correctly to match your amplifier. Or he can leave it uncut and let you do it.

David, why does the capacitor need to be rated so high? Surely the output tubes don't swing 500V each? Some kind of flyback effect?

Wes

[David Anderson]

David, why does the capacitor need to be rated so high? Surely the output tubes don't swing 500V each? Some kind of flyback effect?

Last month, I removed a 600V replacement capacitor from this position in a 32H amp. It was completely shorted.

You can get quite large AC signal swings across transformer windings, given the right conditions. There was just a discussion on a guitar amp tech forum about flyback diodes in high-powered guitar amps, the best types to use, and their value in protecting output transformers, especially in the event of a speaker or speaker cable failure creating a no-load situation.

Most of the classic Leslie tube amps have output transformers with a 5k primary impedance (though that varies with frequency). Calculate 40 Watts AC RMS across 5k, and then calculate peak-to-peak values from that. Make allowance for the fact that the impedance may be more then 5k at speaker impedance peaks and allow for transient peaks in excess of constant output. Then, give yourself a safety margin for the capacitor. 21H/44W amps are nominally 20 Watts, but I've clocked a rebuilt 44W at ~28 Watts.

I didn't understand this issue either, until I started investigating how magnetic coils (of which transformers are one type) work. The speaker magnetizer I rebuilt for a local company has to have a 100 Amp flyback diode to keep the coil from self-destructing when you switch the power off. I even found I needed a flyback diode on a stand-alone power supply I recently built for field-coil speakers to keep the field-coil from reversing the polarity on the power supply when switched off.

[Wes]

Yikes! Thanks, David -- that was a very interesting post. I had never really thought about flyback effects in output transformers before; normally I only think of them in the context of relay coils.