clipping LEDs to floating AC-ground?...makes a differance?

[Johan]

 a number of years ago I had one of those inexpensive Marshall solidstate rack preamps, model 9004. I remember it sounding pretty good and have been thinking about looking for another one for a while.
I remember what I liked about it was that it had some very tubelike dynamics, even at higher gain.
looking at the schematic  http://www.drtube.com/schematics/marshall/9004.gif I noticed that the clipping LEDs go to a floating point between two caps, with no dc-path to ground. it is obviously an AC-ground so clipping will occur, but could this be affecting the dynamics in the clipping ( the charge/discharge)or is it just a curiosity of this circuit, that has no real effect?...I mean, if it makes no difference, it would just be wasting two caps when they could just have used the ground, right?..
what do you guys think?


johan

[R.G.]

I'm not sure why they did that. There's no obvious reason to have done it that way, except that they wanted to return current to the power supplies instead of ground, which doesn't tell you much.

It it possible it was to keep the grounding quiet because of a nonoptimum layout or something.

Maybe they were just long on 2.2uF caps that month, or maybe they could tell when someone copied them by looking for extra caps. It's just speculation.

There should be no technical reason it would sound different if the power supplies and grounds were clean and low impedance.

[Gus]

C27 and C28?

I do not see a DC discharge path if JS1 ground is opened with the plug inserted.  The tone stack and C2 are AC coupled so the caps C27 and C28 might be charged different with an asymmetrical guitar waveform and the diode cap node will "move around" with time.  Now what would that do 1.5 to 1.7 or so voltage drop with a red LED how might the node move to the - 12 or +12 supply rail?
  What is IC 1a doing? 

[oskar]

IC1a looks like a special instance of a voltage follower. It's not needed though to drive the LED clipping stage so...
perhaps it's wellcome to the op-amp stacking thread?   

[Sir H C]

At first I thought, hmm maybe highpassing the clipping, but 13 Hz is the 3 dB point, so doubt that, I agree, it is odd, unless the DC at that point is not well controlled, no clue.

[oskar]

What is IC 1a doing? 

It grounds the boost channel when in normal mode. The schematic doesn't specify IC1 ( or IC3... ) but it should be the M5201 switching IC. nor does it show pin 1 which is the control pin.

[cpm]

i dont know the calculations, may those caps play like a freq. dependant clipping?

[Gus]

http://pdf1.alldatasheet.com/datasheet-pdf/view/898/MITSUBISHI/M5201FP.html

found the 5201.  IC 1,2 and 4 look to be 5201s

[Sir H C]

What is IC 1a doing? 

It grounds the boost channel when in normal mode. The schematic doesn't specify IC1 ( or IC3... ) but it should be the M5201 switching IC. nor does it show pin 1 which is the control pin.

The schematic does show the pinout only for that opamp so makes sense that that is what is used there.

[oskar]

But IC1 shows ordinary OP-amp polarity on the inputs. And doesn't indicate a common output except for the fact that they're tied together?   

[DougH]

I do not see a DC discharge path if JS1 ground is opened with the plug inserted. 

I don't see that either but there is a switch "arrow" symbol where the "LED section" connects to the top of the jack ground on the input. So that implies to me that there is some sort of switching action. Otherwise the boost channel input would be permanently grounded.

[Gus]

Oskar

   Maybe the pin number are correct for IC1 but the draft person mixed up the drawing with the - and + input.  Pin 6 as - input makes sense for an inverting opamp stage and the shared output pin 5 grounding the signal when the sides of IC1 are switched like you posted before.

[Johan]

I think Oskar is right about ic-1a is grounding the lead channel when the clean chanel is engaged. Marshall has ben using those opamp's eversince these amps were intrudused, through the jcm900 and jcm2000 range...simply a dual opamp with an internal switch to select wich one the output is tapped from
and the input jack is grounding it to to keep things quiet when nothing is plugged in.
what pussels me is the total lack of DC reference on the LEDs..like, how do you know if you fly or fall if you can't see the ground...so what's going on at the junction of c27,c28? is it stable somewhere in the middle?. what about when no signal is applied..will it drift?...doing it that way to keep the ground clean doesn't make sence to me. any noise introduced by the clipping will only be there when clipping and probably so much below the signal level I doubt it would be detectable?..they just seem to have gone so far in making sure the LEDs are isolated from any dc reference...including C2
..or perhaps they did it this way to confuse people on a stompboxforum 19 years later? 

I guess I'll have to breadboard this sometime and drag out the old o-scope to have a look...

johan

[Sir H C]

I think Oskar is right about ic-1a is grounding the lead channel when the clean chanel is engaged. Marshall has ben using those opamp's eversince these amps were intrudused, through the jcm900 and jcm2000 range...simply a dual opamp with an internal switch to select wich one the output is tapped from
and the input jack is grounding it to to keep things quiet when nothing is plugged in.
what pussels me is the total lack of DC reference on the LEDs..like, how do you know if you fly or fall if you can't see the ground...so what's going on at the junction of c27,c28? is it stable somewhere in the middle?. what about when no signal is applied..will it drift?...doing it that way to keep the ground clean doesn't make sence to me. any noise introduced by the clipping will only be there when clipping and probably so much below the signal level I doubt it would be detectable?..they just seem to have gone so far in making sure the LEDs are isolated from any dc reference...including C2
..or perhaps they did it this way to confuse people on a stompboxforum 19 years later? 

I guess I'll have to breadboard this sometime and drag out the old o-scope to have a look...

johan

And add a DC path to the circuit

[oskar]

Actually there is a DC path. The leak current from the caps. But that will make for bad aging habbits won't it? The cap's will age unevenly...
( The GGG graphic equalizer's first op-stage has an electrolyte cap blocking the DC-path, so it relys on capacitor leak for bias. )