Standby Switch Problems

[Rodango]

I'm having difficulty with standby switches surviving in my 300W KT88-based bass head that I DIY'ed. And now that I have taken the KT88s to 600V on the plates, even my nice HIWATT standby switch was welded closed.

When I close the standby switch, I get a large "pop" and a visible arc in the switch. This happens every time the switch is closed and eventually whatever switch I put in welds together.

The Hi-Voltage PS is essentially PT -> SS Bridge rectifier -> 50uf -> 10H -> 900uf -> standby switch -> Power Amp

Should I try the switch at a different point in the circuit? Is the choke too large? I need to calm this thing down a bit....

Thanks!

Dave

[richie00boy]

Supress the switch by wiring a 4.7nF Y rated capacitor across it.

[Corax]

How about using a high-voltage power MOSFET doing the 'hard' switching nice and softly with a few added discrete components and replacing the manual switch at the same position?

[Ty_Bower]

Use separate power transformers for filaments, bias supply, and B+. Put the standby switch on the AC primary of the B+ transformer. Leave the filament and bias supplies hot when you switch the B+ supply to standby.

[Tom Bavis]

Are there any capacitors on the load side of the switch? If so, there's a hell of a lot of energy being transferred at the instant of switch closure. The switch probably has a lower DC voltage rating than its AC rating... AC is much easier to switch since it drops to zero 120 times a seconds, so arcs don't last. I'm nut sure you can get ANYTHING except for a vacuum relay that's rated for 600V hot switching (actually more since it's unloaded).

[Rod Coleman]

I would say the problem is the high value of the 900uF cap. The standby switch is making all that energy available to uncharged valves (risking arcover if you change to cheapo KT88s. The instantaneous turn-ON current is probably enormous! Have you any EMP-damaged computers nearby?

Try a cheap inrush-suppression NTC. Use one rated the same as your HT fuse (which I guess is about 1A-1,6A for a 300W-er). Install in series with the switch. Putting another one after the bridge wouldn't hurt.

Snubbing the switch (10nF 1500V LCR Capacitors type PC/HV/S and series 10 Ohm) will also help switch life.

Stick with it though - Marshalls and Laneys and HiWatts all used to run 550V - 600V through a standby switch, so it can be made to work.

[SpreadSpectrum]

I put a separate filament transformer in and now I switch the primary of the HV power transformer. Much easier on switches. Power transformer runs cooler now too.

AES sells a 16A 6.3V Hammond transformer for under $35. It might be worth considering if you still have a little space.

[audiowize]

Wow! Switching B+... that's gutsy! I hope it was a plastic switch, this just seems very, very risky. The separate filament transformer is a good idea!

[Rodango]

Got it fixed ...

I removed the 225uf of capacitors beyond the switch and suppressed the switch with a .0039uf 1600V Orange Drop. No more fireworks. I'm sure it was the 225uf which was doing it.

I found a couple rectangular pieces of plastic under the chassis. I was trying to figure out what they were and finally realized ther were the sides of the switch! The 600V inrush current arc blew the sides off of the switch!

Thanks for the help!

[MrTransistorm]

Originally posted by Ty_Bower
Use separate power transformers for filaments, bias supply, and B+. Put the standby switch on the AC primary of the B+ transformer. Leave the filament and bias supplies hot when you switch the B+ supply to standby.

That's how it's done in the Ampeg SVT. The switching is nearly silent.

If you can't use a separate heater/bias transformer, try putting the standby switch directly after the secondary before the rectifiers. That way you'll be switching AC. Correct me if I'm wrong, but I believe that switches can handle much higher AC switching currents than DC because the arc extinguishes as soon as the voltage alternates. With DC the arc can remain indefinitely, destroying the contacts in the process.