Vox AC50 cathode follower dissipation

[pdf64]

Re Vox AC50, JMI SS rectifier, pre 1965 bright switch
I've worked on a few of these and have been concerned by the plate dissipation / current that the ecc83 cathode followers are operating at. Basically the have a marshall style direct coupled common cathode / cathode follower tone stack driver. However, unlike marshall the cathode resistor is 56k, and the supply voltage is high. This leads to about 250vdc across the cathode resistor, 4.5mA plate current, which is off all the charts I've seen, plus it's considerably over the heater cathode insulation ratings (anything from 100 to 200vdc, depending on manufacturer). The 1watt 56k cathode resistors have always oozed wax and drifted way up, both about 75k for the one I'm working on at the moment. I think that JMI just transfered the tone stage from the AC30 (which they borrowed from which Gibson amp?), without taking account that the B+ was 150vdc higher for the AC50. I've taken to reducing the voltages and currents to reasonable levels, by increasing the cathode resistor to 100k and increasing the B+ dropper resistors to get less than 200vdc across it.
Has anyone else considered this a problem and what are fixes for it?
Or should I just replace the stressed resistors for stock values and not worry about the tubes burning up?
Plus they always seems low on power output, never more than 40watts before clipping, even with newcaps and el34s?
Thanks for your input - Peter

[R.G.]

Any time you have parts melting down, it's an issue. Here's something that's a slam dunk fix. Go read the MOSFET Heresies at GEO. You can directly replace that cathode follower with a high voltage MOSFET follower with other changes to the circuit. Plus, this frees up a tube section for use in other stuff if you're so minded.

No, you can't hear that it's solid state. Read the article. Tube crazies have built this follower in and have reported that they can't tell it's SS by listening.

Another option would be to parallel another tube section with the triode follwer to lower the current pulled by one section, but then you have to cobble in another tube, worry about heater current, etc.

That does not solve the 56K resistor dissipation issue. You have some choices there. Either put in a high-power resistor of 56K value, change the resistor to 100K, or put a resistor in the plate side of the tube to drop voltage to that tube. The grid is sitting at 250V nominally, so an 18K or 20K resistor in the plate would drop off 80 and 90 V respectively from the tube dissipation. this would not help with the 56K resistor dissipation at all, though. But it's easier on the tube.

You could also put in a constant current source load for the follower, whether you use a triode or a MOSFET for the follower. That would force the follwer device to run at a lower current, but the same voltage, and also improve the follower performance at the same time. But it's more work.

A power MOSFET could also be used for the CC load on the existing tube. It would force the existing tube to run at 1 or 2 ma, without changing anything else. This would replace the 56K resistor. Most power MOSFETs of an appropriate 800V - 1kV voltage rating will dissipate 1W with no heatsink, and up to 20-50W with a sink bolted to them, so they can handle the heat.