Switching valve output in Marshall JMP1 preamp

[dobrota]

Hi everybody,
See part of schematic. HT is about 160V, power supply +-5.6V for 4053 switch so about +-6V absolute maximum analog signal level. I found other service information - signal level in various nodes for this preamp. There's 4V pp (+-2V, 3 diodes in series) on grid VB2a and 100V pp (say about +-50V, yes I know it isn't symmetrical) on C51/R92 and from 1.5V to 2.5V pp on IC12c pin 4 out depending on switch positions. I tried a PC simulation next part of circuit and there's +-26V on R91/R92 for switch in position as in the schematic , +-5V in second position (heavy load from fix tone stack circuit), and +-32V during switching time delay on IC pin 15 input.
I don't understand, how can IC12b section of 4053 handle continously so high voltage levels (four to five times more than maximum in datasheet) without damage. Some special unknown secret ?

[Enzo]

Probably the same way Leo Fender got away with running 6V6 tubes 100v over their "maximum rating" for all those years.

Looking at the internal circuits of the 405x cells, I see MOSFET current channels for the signal path not connected directly to the control circuits. SOmeone probably tried it and found they took signal voltages and didn't blow up.

Data sheets tell us what the IC maker will guarantee as performance... more or less. It does not tell us what voltages the chips will explode at.

[Steve Conner]

Like almost every other IC, the 4053 contains a pair of ESD protection diodes on every pin that will clamp the voltage on the pin to 0.7V outside the supply rails.

It's considered bad practice to rely on the clamping effect of these diodes in your circuit, but that doesn't stop designers from doing it.

[fuzznut]

Well that begs the question: Is that 4053 a contributor to the distortion "sound" of the jmp-1? I have one, and it would be interesting to run a sig-gen and look at the distortion waveform on each side of the switch.


edit: that won't work, I would have to lift pin 15 on 2B, and look at the waveform at R91,92.

[dobrota]

Thanks, but I think it's possible to mesaure signal level and see waveforms on pins 15 and 1,2 of 4053 with signal from generator in input JMP1 without disconnect pin 15 from R91/R92. There's about 100V pp, almost square waveform, on C51/R92 according to Marshall service info.

[Gregg]

There's about 100V pp, almost square waveform, on C51/R92 according to Marshall service info.

Well that begs the question: Is that 4053 a contributor to the distortion "sound" of the jmp-1?

It looks like it is because there's no way 4053 can stay "clean" when fed such amplitude signals and the signal after CF is almost a square wave before going to 4053.

[dobrota]

It looks like it is because there's no way 4053 can stay "clean" when fed such amplitude signals and the signal after CF is almost a square wave before going to 4053.

For low signal levels (clean or crunch sounds) 4053 switch is linear and for high signal levels (heavy distortion, almost square waveform on input 4053) another distortion creating inside 4053 (clamp diodes etc.) maybe hasn't almost any effect for sound and waveform. So it works in some strange way for both low and high signal levels. Or switch really also creates distortion using CMOS "tube like" sound (as various pedals or guitar amps using 4049 IC etc).

[Gregg]

For low signal levels (clean or crunch sounds) 4053 switch is linear and for high signal levels (heavy distortion, almost square waveform on input 4053) another distortion creating inside 4053 (clamp diodes etc.) maybe hasn't almost any effect for sound and waveform. So it works in some strange way for both low and high signal levels. Or switch really also creates distortion using CMOS "tube like" sound (as various pedals or guitar amps using 4049 IC etc).

The only way to find out is to try how it sounds with and without the 4053