Valve amp HV switches

Is it a problem with the make or break? any HV cap over the contacts? Parallel up a DPST switch, perhaps NOS would be better

Is it a problem with the make or break? any HV cap over the contacts? Parallel up a DPST switch, perhaps NOS would be better

This is a vintage(ish) amp that has snap-in rectangular Mains and standby switches. The Mains DPDT double width, the Standby SPST, single width.

I don't really want to be drilling holes for another switch that isn't even rated to 500v.

Gareth.

lack/dp/2068581

s t

I

This gets you to 400VDC: http://76.12.210.115/pdf/s-toggle-sw-rocker-400v- dc.pdf

and then, if you use a relay, you do not risk arc-welding the switch:

Hope it helps. Keep in mind that at higher voltages, arcing is the concern. But you knew that already or you would not be asking the question.

Peter Wieck Melrose Park, PA

This gets you to 400VDC: http://76.12.210.115/pdf/s-toggle-sw-rocker-400v-dc.pdf

and then, if you use a relay, you do not risk arc-welding the switch:

Hope it helps. Keep in mind that at higher voltages, arcing is the concern. But you knew that already or you would not be asking the question.

Peter Wieck Melrose Park, PA

Interesting that the switch you linked to uses a magnet to steer the arc, as used on some Speaker Relays.

I'm guessing that valve amp manufacturers have used underated HT switches for decades now, in much the same way that they have often put way higher voltages on pre-amp valves than they are designed to cope with. (Fender, Mesa reverb drive circuits for instance).

So I guess I shouldn't be so shocked at this revelation.

Gareth.

I'd be tempted to rewire the unit to activate a high-side PMOS power FET switch. The plastic switch could bias it from any positive or negative low voltage with a NPN level translator (emitter driven from low voltage DC on the replacement switch). Depending on the current you're switching, even an optoisolator might suffice.

** Is the failure mechanical or are the contacts burned up ?

lack/dp/2068581

** Where is the switch connected - in the AC feed ?

t ** And you won't.

I

** You have missed the fact that the breaking capacity of a switch depends on the current flowing at the time. That rocker switch will break 24V DC an d 250V AC all day long as long as the current does not exceed 16 amps.

In the Standby position of a valve amp, it maybe has to break 0.4 amps at m ost - so the voltage rating goes way up and it can be expected to do the jo b fine.

Plus any 250VAC mains switch is insulated to way more than the nominal volt age to allow for spikes etc - so that is OK too.

The best idea is to have the switch in the transformer wiring, breaking the AC voltage rather than the DC after the rectifier. Marshall did this in mo st models, using a double pole rocker.

But Fender got it *horribly* wrong in a great many models - like the Bassma n 100 - connecting the switch in the one spot it should never be.

Imagine the switch is operated when the amp is on and running at high volum e - maybe because of a sudden acoustic feedback.

Momentarily, as the switch opens, OT is fed HT from the 20uF electro but no w in SERIES with the filter choke as the main electros are disconnected. Au dio frequency current flows in the choke and generates a very large AC volt age pulse - making the switch contacts arc and the same voltage also appear s on the output valve plates.

I have measured this voltage at more than 3kV rms.

A similar thing happens when the HT fuse in a70s or 80s Marshall blows, sin ce it is wired in the same place.

Explains a lot of arc burnt valves and octal sockets.

.... Phil

** Is the failure mechanical or are the contacts burned up ?

** Where is the switch connected - in the AC feed ?

** And you won't.

** You have missed the fact that the breaking capacity of a switch depends on the current flowing at the time. That rocker switch will break 24V DC and 250V AC all day long as long as the current does not exceed 16 amps.

In the Standby position of a valve amp, it maybe has to break 0.4 amps at most - so the voltage rating goes way up and it can be expected to do the job fine.

Plus any 250VAC mains switch is insulated to way more than the nominal voltage to allow for spikes etc - so that is OK too.

The best idea is to have the switch in the transformer wiring, breaking the AC voltage rather than the DC after the rectifier. Marshall did this in most models, using a double pole rocker.

But Fender got it *horribly* wrong in a great many models - like the Bassman

100 - connecting the switch in the one spot it should never be.

Imagine the switch is operated when the amp is on and running at high volume - maybe because of a sudden acoustic feedback.

Momentarily, as the switch opens, OT is fed HT from the 20uF electro but now in SERIES with the filter choke as the main electros are disconnected. Audio frequency current flows in the choke and generates a very large AC voltage pulse - making the switch contacts arc and the same voltage also appears on the output valve plates.

I have measured this voltage at more than 3kV rms.

A similar thing happens when the HT fuse in a70s or 80s Marshall blows, since it is wired in the same place.

Explains a lot of arc burnt valves and octal sockets.

.... Phil

Thanks, Phil, that makes a little more sense now. This switch is connected to the main filter caps, as in the Bassman schematic, though I don't think it has a choke, it's a pretty sparse 50W EL34 design (Roost, made in UK). One Mullard EL34 is toast, the other is OK. It also had a shorted cap in the tone stack totally crippling the whole amp.

I haven't yet managed to get the switch out of the chassis, it's a tight bastard, was going to see if I could repair the contacts, but to do so it needs to come out intact, or I need to source a replacement.

Cheers,

Gareth.

** The bad EL34 may be responsible for the burnt switch.

The owner may have tried to use the standby switch when the valve was passing a large DC current - like a couple of amps. At that level, a continuous arc would form and quickly destroy the contacts.

I see there is no DC choke in the schem but the fusing would allow that sort of current to flow.

Maybe when you replace the switch, rewire it in series with the HT fuse - that should allow arcs to extinguish at switch off.

.... Phil

Also, if you can get a 2-pole switch that fits ok, then wiring the two sets of contacts in series ought to improve the ability to extinguish the arc, (for a given contact separation etc.).

Chris Jones wrote: >

** If you look at the schem, the HT fuse is in the rectifier ground path where the voltage and current levels drop to zero at twice the supply frequency.

So for a couple of mS each 10mS, there is not enough of either to sustain an arc.

... Phil

Yes, I noticed that. It is much better than the original arrangement. If C23 and C24 are in good condition and the B supply rail doesn't get shorted, then the fuse (and new switch) should only be carrying current for a quite small fraction of the time, so arcs should extinguish well, perhaps even better than with a resistive load on AC.

Yes, I noticed that. It is much better than the original arrangement. If C23 and C24 are in good condition and the B supply rail doesn't get shorted, then the fuse (and new switch) should only be carrying current for a quite small fraction of the time, so arcs should extinguish well, perhaps even better than with a resistive load on AC.

Thanks Chris and Phil, that's very interesting.

Why wasn't this practice adopted decades ago? Were Fender (Bassman) really that clueless? Beggars belief.

Cheers,

Gareth.

[...]

Don't make the mistake which comes around time and again* of switching a charged capacitor directly onto a discharged one, the switch won't last long. Use a resistor to limit the surge current.

• The AR88 communications receiver and a Philips transistor car radio (40 years apart) spring immediately to mind.

--
~ Adrian Tuddenham ~ 
(Remove the ".invalid"s and add ".co.uk" to reply) 
www.poppyrecords.co.uk

h

in

** Nice example of how so often, the devil is in the detail .....

lly

** Fender originally placed the Standby switch after the main filter electr os in consideration of the valve rectifiers being used. Placing it earlier creates a large current surge every time the switch closes while the electr os charge. Such surges well exceed the maximum ratings for 5U4s or GZ34s an d will drastically shorten their life spans.

When Fender changed to using strings of silicon diodes, the problem vanishe d and they ought to have reconsidered the standby switch scheme as Marshall did. Ferder's "The Twin" model ( 1993) finally has a double pole switch in the A C like Marshall.

.... Phil

** You mean like the Mesa Boogie MkI model does ??

Note the 60F, 500V electro fitted AFTER the standby switch !!

The main electro bank is inadequate ( 2 x 60uF, 350V in series )so someone decided to add an extra cap. There is a loud *snap* whenever the standby is closed and it soon burns the hell out of it.

Later versions used 2 x 220uF in the main bank and a 0.1uF 600V film cap after the switch ( so from OT centre tap to ground ) which sometimes fails short.

For sheer stupidity it is hard to beat early 70s Marshalls that switched OFF the bias supply with the standby.

While the 400V HT come up instantly, output valve grid bias take 3 or 4 seconds to arrive !!

.... Phil