Snubbers? Do I really need them?

[Greg_L]

I'm building a pretty standard Super Lead clone based off this schematic....


Those two .22uf caps bridging the PT secondary....do they have to be there?

Also I'm a little confused as to how they're implemented. Are they two non-polarized caps in series just bridging across the secondary wires before the standby switch? Are they supposed to reduce pops or something?

[Greg_L]

Also if it matters....I'm using beefy 1N5408 diodes for the rectifier. I know it's overkill but I have them already.

[pdf64]

I guess they were added as best practice to reduce diode switching noise.
Neither the 90s or the 2002 reissues bother with those snubbers.
https://www.drtube.com/marshall-reissues/

[Greg_L]

Neither the 90s or the 2002 reissues bother with those snubbers.
https://www.drtube.com/marshall-reissues/

Yup, I know. I've been inside many of the reissues and have never seen those snubbers on anything.

I have a 73 Super Lead and it has one snubber on the board but it had been disconnected at some point in the last 50 years. It's still there for originality I guess? It's not doing anything though.

It's just a curiosity for me.

[Helmholtz]

Not sure.#

They might be used to
- reduce standby switching pops.
- prevent arcing of the standby switch contacts for longer switch life,
- mitigate mains HF noise entering the amp,
- lower diode switching noise.

My '68 Super Lead doesn't have them but uses a different standby switch interrupting DC.

[Greg_L]

I'm using modern reissue power and standby switches. I guess I'll skip the snubbers and see how it goes.

I'm still curious about how they're connected though. The schematic shows a no-connection at the PT center tap. So where are they actually going? Two caps in series from one HV secondary to the other? I don't understand. Is that an error? Or is it because the center tap doesn't actually connect to 0v ground? WTF is happening here?????

[Chuck H]

I'm late getting to a vintage plexi build project but I've done a ton of research, and... Most of the guys on the Marshall specific forums have chosen to omit those snubbers in their builds (for their own various reasons) and I never read about it being a problem in my perusing.

[Greg_L]

I'm late getting to a vintage plexi build project but I've done a ton of research, and... Most of the guys on the Marshall specific forums have chosen to omit those snubbers in their builds (for their own various reasons) and I never read about it being a problem in my perusing.

If they are connected to the HT center tap, then it kind of seems like a bad idea maybe? If one of those caps shorts then you melt a PT? I admit I'm an ignorant mongo but I'm failing to see how the benefit outweighs the potential for disaster.

[Helmholtz]

I'm still curious about how they're connected though. The schematic shows a no-connection at the PT center tap. So where are they actually going? Two caps in series from one HV secondary to the other? I don't understand. Is that an error? Or is it because the center tap doesn't actually connect to 0v ground? WTF is happening here?????

I think it's a schematic error missing a connection dot.

[Greg_L]

I think it's a schematic error missing a connection dot.

Okay because the 50w schematics show the connection dot but the CT is going to actual 0v ground. I thought maybe there was some difference for some reason that I don't yet understand.

[trobbins]

As shown in that schematic, the CT is sitting at 50% B+, and the diode bridge acts as two sets of full-wave rectifiers with one set developing -50% and the other set developing +50%. That has the distinct advantage of allowing each 100uF cap to be stiffly regulated, and not need balancing resistors per se. But put bleed resistors in if doing a clone imho. That is not an uncommon rectifier configuration.

The B+ load is going to be reasonably high, even at idle, due to the quad - that means the charging current pulses have a highish peak as the peak depends on the half-winding resistance (which is lower than for a full winding) and the load current and the use of a reasonably large 100uF value. I'd suggest the 0.22uF caps are there to avoid diode turn-off and recovery induced noise from becoming noticeable. The reverse recovery effect can be mitigated by using modern UF5408. The turn-off effect can be mitigated by using a 'tuned snubber', as per the quasi-modo test setup (over on diyaudio). The noise is generated by the leakage inductance in the power transformer, which if left untamed can often easily leak out past the diodes and splatter into sensitive parts of the circuitry - perhaps not noticeable, but likely to be seen on scope plots as it can also easily couple over into the heater and bias windings to 'get out'.

You may not notice and hence not worry about this issue, but if you do then I suggest at least use UF5408, and if you're really keen (and have enough instrumentation) then set up a quasi-modo style test - but note that the snubber capacitors must be AC rated for full-time 50% secondary voltage (plus margin) and be at least X rated (to avoid any stress if they start failing).

That rectifier configuration doesn't make it easy to fuse at the windings, and I'd suggest that if the B+ feed fuse popped then the OPT and choke would likely get hit with a stressful high voltage spike.

[Greg_L]

As shown in that schematic, the CT is sitting at 50% B+, and the diode bridge acts as two sets of full-wave rectifiers with one set developing -50% and the other set developing +50%. That has the distinct advantage of allowing each 100uF cap to be stiffly regulated, and not need balancing resistors per se. But put bleed resistors in if doing a clone imho. That is not an uncommon rectifier configuration.

The B+ load is going to be reasonably high, even at idle, due to the quad - that means the charging current pulses have a highish peak as the peak depends on the half-winding resistance (which is lower than for a full winding) and the load current and the use of a reasonably large 100uF value. I'd suggest the 0.22uF caps are there to avoid diode turn-off and recovery induced noise from becoming noticeable. The reverse recovery effect can be mitigated by using modern UF5408. The turn-off effect can be mitigated by using a 'tuned snubber', as per the quasi-modo test setup (over on diyaudio). The noise is generated by the leakage inductance in the power transformer, which if left untamed can often easily leak out past the diodes and splatter into sensitive parts of the circuitry - perhaps not noticeable, but likely to be seen on scope plots as it can also easily couple over into the heater and bias windings to 'get out'.

You may not notice and hence not worry about this issue, but if you do then I suggest at least use UF5408, and if you're really keen (and have enough instrumentation) then set up a quasi-modo style test - but note that the snubber capacitors must be AC rated for full-time 50% secondary voltage (plus margin) and be at least X rated (to avoid any stress if they start failing).

That rectifier configuration doesn't make it easy to fuse at the windings, and I'd suggest that if the B+ feed fuse popped then the OPT and choke would likely get hit with a stressful high voltage spike.

Thank you for this nice explanation. It's above my pay grade, but after reading it and researching a little further I think I sort of get it.

[Chuck H]

If they are connected to the HT center tap, then it kind of seems like a bad idea maybe? If one of those caps shorts then you melt a PT? I admit I'm an ignorant mongo but I'm failing to see how the benefit outweighs the potential for disaster.

On this matter, FWIW and unless I'm mistaken...

The "failure mode" for components is considered in their selection for duty (and indicated in their type model by specs). And I think "snubber capacitors" are designed to fail "open", rather than fail "short".

[Greg_L]

On this matter, FWIW and unless I'm mistaken...

The "failure mode" for components is considered in their selection for duty (and indicated in their type model by specs). And I think "snubber capacitors" are designed to fail "open", rather than fail "short".

So then it would be like they're not there at all. Makes sense.


Is it possible these things were spec'd in those early 70s units because there's no filter reservoir before the standby switch?

[Chuck H]

So then it would be like they're not there at all. Makes sense.


Is it possible these things were spec'd in those early 70s units because there's no filter reservoir before the standby switch?

Possibly? This goes beyond MY pay grade. Maybe inrush charging the caps after the switch caused a problem with some switch in some test. It's possible a different switch was used in the actual amps and so never was a problem. Or that the standby switch location was ignored on other models with snubbers indicated in the schematics. I've often seen design artifacts in amps or their schematics that don't make immediate sense, but don't usually cause any harm. And in this case, as in others, it's not uncommon to see that the actual amps don't always match the factory schematic.