I keep breaking standby switches....

[rccolgan]

ok... so maybe something I'm overlooking however I have this problem where I'm using quality Carling switches for standby on my Dumble builds (440vdc on the standby), some part numbers will short out internally and not recover! The switch is dead and always on from that point forward. What the heck am I missing? If it is relevant, I like to use .047uF 630v caps on the standby to eliminate the noise when turning the amp off, however I've had switches fail without the cap too.

Part numbers I've tried and they internally shorted out after first switching on:

• https://www.amplifiedparts.com/products ... t-tabs-15a

• https://www.amplifiedparts.com/products ... der-lugs-1

Part numbers I've tried successfully:

• https://www.amplifiedparts.com/products ... older-lugs

Per Rafael's suggestion, I'm going to try https://www.mouser.com/ProductDetail/691-2FA52-73 next because that's what he uses successfully. My goal is to have a solid "CA-CHUNK" feeling when switching. I find the https://www.amplifiedparts.com/products ... older-lugs feels kinda cheap and flimsy.

Can anyone clue me in on what is going on? Thank you in advance.

[xtian]

Do not use standby switches. Seriously.

[rccolgan]

Do not use standby switches. Seriously.

Absolutely agreed here the more and more I read. My next chassis designs will eliminate the standby switch. Unfortunately, I need to make this work per original this time around.

[sluckey]

Have you tried a DPST switch to break the HT AC wires rather than the B+ wire?

[maxkracht]

The cap should reduce arcing inside the switch, maybe adding a resistor across it as well would help? I think somewhere around 100k 2-3w is standard. Switching the AC line from your transformer could also be an option.

[Lauri]

Most switches aren't able to handle much DC and aren't even rated for it. If you absolutely must have a standby switch use the switch to short power tube control grids together or some other simple way to mute the amp.

[roberto]

Most switches aren't able to handle much DC and aren't even rated for it. If you absolutely must have a standby switch use the switch to short power tube control grids together or some other simple way to mute the amp.

I would say that, if needed, grounding the MV would be a simpler, safer and better solution.

[Stevem]

I use where I have room a SPDT 4 to 6 amp switch using both sides of the switch.

[rccolgan]

I use where I have room a SPDT 4 to 6 amp switch using both sides of the switch.

Never even thought of that! The switch that just broke was a SPDT. I ordered the one that Rafael suggested so crossing my fingers.

Thank you all for your input! More and more reason to ditch the standby switch next go-round.

[SoulFetish]

Do not use standby switches. Seriously.

I agree with xtian. I don't use them in my amps. Most people use them as a "mute" switch anyways, so there is no reason to have a standby switch open the HV supply anyways. If still wanted to implement a standby/mute switch there are many ways to do it. But the way Silvertone did it is easy, effective, and cheap.
They essentially short the grid connections going to each of the output tubes in a push pull amp. like this simplified schematic:

Silvertone Style Standby.jpg

because the switch is post PI after the coupling caps, it takes advantage of the common mode noise rejection, so it is very quiet.
If it does not act as a mute, you know you have an issue of one side of the push-pull circuit in failure. So it can act as in indicator of a problem in you output stage as well.

[R.G.]

I agree with the mute switch theory - if you're using it as a mute switch, make it be a mute switch, not a dramatic power on/off surge. Or just don't put a standby switch in there at all.

There is a side path. Don't turn high voltage DC on and off with a switch, use a device that was designed for it - the MOSFET. A MOSFET and an LED-photovoltaic isolator can turn your high voltage on and off all day long with no ill effect on the MOSFET. Switch current in the LED section of the isolator, the MOSFET switches. Devices rated for amperes and 1000+ volts are a couple of bucks, the isolators about the same. They don't arc unless you do a poor job of soldering up the board. You can then use a traditional bat-toggle switch on the front panel to satisfy purists who demand a standby switch.

There's another item here - the surge when the standby switch is flipped from mute to normal. If your standby switch interrupts the high voltage, then the reservoir caps are fully charged up when the standby switch is switched back to conduction. This makes for a big surge of current at high voltage, and it is hard on the capacitors, the standby switch, and the tubes that are connected, filaments hot. Our friend the MOSFET can help. There is a simple circuit that makes a high voltage MOSFET into a current limiter. You can literally dial the peak current out of the filter cap down to just a little more than the amp uses at full power, and the limiter will clamp surges down to the limit level, not interfering with max amplifier current use. There's a circuit for this illustrated at geofex.com.

[Mac Dillard]

I use the mouser 633-S1A https://www.mouser.com/datasheet/2/295/ ... 663420.pdf and have never had a problem.

[Colossal]

I use the mouser 633-S1A https://www.mouser.com/datasheet/2/295/ ... 663420.pdf and have never had a problem.

I would also recommend NKK switches. They are MUCH more robust than the Carling switches and give a satisfying, heavy snap, when thrown.

[SoulFetish]

I use the mouser 633-S1A https://www.mouser.com/datasheet/2/295/ ... 663420.pdf and have never had a problem.

2nd the NKK switches. Use 'em. Love 'em.

[Roe]

I've never broken a stdby switch but I use resistors across the switch. Still, it is easy to excede the voltage rating of the resistors. on a 45/100, there was 385v across a 330k resistor, but the resistor only seems rated for 350v