Marshall 1987x Plexi HF leakage via High Treble Vol Pot when plugged into Normal Ch

[nevetslab]

I have a Marshall 1987x Mk II Plexi 50W head on the bench, doing preventative maintenance. After swapping out the power tubes and re-biasing, next checking to see if everything works, I find going thru either input 1 jacks, using the Normal Ch volume (Loudness 1) pot, turning the volume pot up for the High Treble Volume (Loudness 2) adds loads of top end with nothing plugged into the #2 jacks. HF leakage, obviously. I just don't see the path for it. All of the input jacks are properly grounded, the two volume pots are grounded to the common buss bar that is soldered to the back side of all the front panel pots. I get solid ground to the chassis from that buss, as well as from the jacks. The normals on Ch 2 input jacks are all grounding the inputs to the second half of the input tube, as each channel has it's own half of that input tube. Swapped out the input tube, same thing.

I don't recall this problem before, though perhaps because I hadn't looked for it. Not that this is necessarily a bad thing, having an additional 'Brilliant Control' to add to the top end of the Normal Channel, but, somethings seems amiss. The High Treble channel works........top end that will make dog's whine 3 blocks away.

Marshall-JMP-Lead-50W-1987-Schematic.pdf

[alexradium]

turning up the bright volume you isolate the combo 470k/500pF from ground,i guess that is the main cause.

[Helmholtz]

Did you check the shorting contacts of the input jacks?
If the passive input is open it will receive high frequency content via capacitive coupling.

[frus]

turning up the bright volume you isolate the combo 470k/500pF from ground,i guess that is the main cause.

That's right, if the treble channel volume is at zero, it bleeds highs to the ground and normal channel becomes very dull. So one should always lift treble channel volume a bit when playing normal channel, so it becomes usable

[nevetslab]

Did you check the shorting contacts of the input jacks?
If the passive input is open it will receive high frequency content via capacitive coupling.

I did check the shorting contacts on all four inputs, and they all measure ground, until interrupted with a plug.

"if the passive input is open it will receive high freq content via capacitive coupling"......I'm not following you on this.

[nevetslab]

That's right, if the treble channel volume is at zero, it bleeds highs to the ground and normal channel becomes very dull. So one should always lift treble channel volume a bit when playing normal channel, so it becomes usable

You're right........with the Treble Ch volume at zero, it IS dull sounding. So, it sounds like this is normal.

[Helmholtz]

"if the passive input is open it will receive high freq content via capacitive coupling"......I'm not following you on this.

There is always some parasitic capacitance (in the low pF range) between adjacent wires and components. If the non-used grid circuit is not grounded or terminated by a rather low resistance, high frequencies passing through the parasitic capacitance from the other/active grid circuit will produce an unintended input signal.
Parasitic capacitance and input resistance constitute a high pass filter, i.e. a frequency dependent voltage divider. Higher input resistance allows for more audible spill-over.

[nevetslab]

There is always some parasitic capacitance (in the low pF range) between adjacent wires and components. If the non-used grid circuit is not grounded or terminated by a rather low resistance, high frequencies passing through the parasitic capacitance from the other/active grid circuit will produce an unintended input signal.
Parasitic capacitance and input resistance constitute a high pass filter, i.e. a frequency dependent voltage divider. Higher input resistance allows for more audible spill-over.

OK, now I'm following ya. The Grid resistance path is typical of most amps....1M + the 68k resistors, one used to form the 6dB attenuator on the lower input jack. The 'normals' on the input jack short out the 1M, leaving you with 34k to ground (68k's in parallel), with the 1M shorted out. So, you're left with that 470k/500pF summing network to the 2nd stage grid from the High Treble Channel Volume pot, which also has the 5nF bypass cap across the top of the pot and wiper, for even more HF boost, when there is signal passing thru. In this case, you have two parallel RC networks in series to ground off the 2nd grid, being fed from the Normal Ch's wiper. So, depending on how much you raise that High Treble volume pot, you're getting HF boost, even though it seems like you should have HF attenuation instead. You do, when the pot is all the way down. As you start to raise it, it's no longer HF attenuation with that 5nF in parallel with the 1M pot. Strange.

[J M Fahey]

Just thinking aloud: these are very simple, unsophisticated amplifiers, which happen to sound very good, so why bother?

Designed to be played live, LOUD ; doubt any user complains about a little HF feedthrough, even IF they notice it in a noisy stage, next to a LOUD drummer and surrounded by screaming headbanging fans.

[Helmholtz]

I don't think Marshall/Bran cared much about the normal channel in their lead amplifiers after ca. 1966. Rather they tried to squeeze out the most treble and gain of existing designs to satisfy their big name customers. A 5nF bright cap across a 1M volume pot seems ridiculously high. But those amps were typically dimed anyway, so the bright cap doesn't matter. Also it was common to bridge/cross-link both channels.

[J M Fahey]

Actually the established way to use them was:

1) bridge input channels

2) rise High Treble channel as much as needed to get the distortion/sustain you want.
IF that means setting it up to 10 (or 11 ), so be it, but remember anything above 2 is already distorting , 3 to 10 just controls sustain but does not actually change *volume*

3) sound is killer but thin, so you rise "Normal" volume only to get some body back, usually 2 or 3 tops; higher brings mud into the mix.

4) enjoy