GZ34 Anode Series Resistance

[harry]

I recently read that modern GZ34s are very critical of the series resistance feeding each anode: it should not be less than 175R.

Supposedly this should be made up of the DC resistance of the transformer secondary (from centre tap to anode) added to a fixed resistance.

Discussed here (Radford Sta15 Problem - UK Vintage Radio Repair and Restoration Discussion Forum) is a method of using a 1N4007 with a 1kV cap across it on each anode instead of adding series resistors to bring it up to 175R.

Will adding a series resistance create extra sag in the amplifier? What rating resistors are we talking here? I'm guessing wirewound, over 5W.

Any other comments are welcomed... I'd like to hear if this is common knowledge and I've just missed out on it until now.

[Steve Conner]

Yes, adding series resistors will increase sag.

The 1N4007 trick amounts to putting a diode rectifier in series with the tube one. It'll stop the tube from arcing back and blowing your fuses, but it won't stop it from being worn out prematurely by high peak currents, only the series resistors will help that.

Of course it could be that the arcing in new production GZ34s is just a transient problem at startup, and once the heaters are fully warmed they can handle the high peak currents fine. In this case, the diode trick might be all you need, but remember that once you've added the diodes, the rectifier tube is basically just there for looks.

The stuff in that Vintage Radio thread about softening the current pulses is kind of true, but a resistor would do that equally well, and you could leave the tube rectifier out altogether.

[Bruce / Mission Amps]

...
Of course it could be that the arcing in new production GZ34s is just a transient problem at startup, and once the heaters are fully warmed they can handle the high peak currents fine. In this case, the diode trick might be all you need, but remember that once you've added the diodes, the rectifier tube is basically just there for looks. ...

I think, if the Si diodes are before the rectifier and before the first filter cap, the rectifier's impedance will still drop the pulsing high voltage DC, especially while the PT and rectifier tube are under load.
The 'sag" effect of the tube will should still be in effect.
So in that application, I'm thinking the rectifier will be there more then just for looks...

[Steve Conner]

Yes, the rectifier will introduce some sag, but if you replaced it with diodes, you could put the sag back in with a series resistor. That's what's inside a Weber Copper Cap or the like, two diodes and a resistor. One resistor in the common cathode point, or two in the anode circuits, it doesn't matter, except that if you use two, they dissipate half the power each.

You can estimate the resistance needed by looking at the V-I curve in the rectifier datasheet, drawing a best-fit straight line passing through the origin, and calculating R=V/I for any point on the line. The rectifier is non-linear, but I doubt anyone would notice the difference, and that's why I argue that it's just there for looks.

In response to another previous question, I guess 5 watt cement block resistors would be fine, maybe 10 if you want to be bulletproof and have the space to mount them.

[harry]

Thanks for the suggestions everyone.

In the end, I used two 68 ohm resistors, one on each side. The rectifier still arced (much to my annoyance as it was a new JJ GZ34). In the end, the resevoir capacitor seemed to be the problem, lowering it from 48uF to 33uF corrected the issue.

Interestingly, both rectifiers still work, even though they had arced at some point. My understanding was that once they've arced, they've had it... didn't seem to be the case.