Channel bleed JCM 800 2205

[Wes]

Any ideas as to what causes this? It seems to be common and I've read about the occurrence
quiet a bit, but I can't seem to find what is causing the issue in the particular case I have in
front of me.

-With MV at zero I can turn the normal channel way up and it bleeds through.

-With MV at zero, normal V off, and boost V and G all the way up nothing.

Not really a big deal here; no problem.

-Normal channel selected (or without a footswitch plugged in) gets no bleed from boost channel.

-With footswitch plugged in and boost channel selected, the normal channel bleeds in with wherever it is set.

The version I have here has the TL3046 chip and doesn't have a couple of the caps that are in the schematic.
It also has what is labeled C40 attached on the underside of the board (an addition to original design I'm assuming)
but it does not connect to the wiper of the normal channel volume as in the schematic. It is connected to the wiper
of the normal channel treble pot. Disconnecting this cap does not affect the above mentioned operating scenarios
either way.

I'm trying to grasp how this chip switches. I take it that its just 5 transistors in a package together. I have taken voltage
readings for the pertinent pins and they are as follows:

BOOST OFF

3 Emitter 0
4 Base .7V
5 Collector 0

7 E 0
6 B .1V
8 C 0

13 E 0
12 B .7V
14 C 0

BOOST ON

3 E 0
4 B .04V
5 C 0

7 E 0
6 B .75V
8 C 0

13 E 0
12 B .04V
14 C 0


Is boost selected by having the transistors at pin 5 and 14 not conduct and the transistor at pin 8 conduct?
With the transistor at pin 8 conducting- is that supposed to "ground" what is C40?

Is it possible that this thing is just leaking elsewhere?
2205prem.pdf

[g1]

-With footswitch plugged in and boost channel selected, the normal channel bleeds in with wherever it is set.

Are you saying the bleed stops if you turn down the normal volume when in boost mode?

Is boost selected by having the transistors at pin 5 and 14 not conduct and the transistor at pin 8 conduct?
With the transistor at pin 8 conducting- is that supposed to "ground" what is C40?

Yes to both. Pin 8 should kill the clean signal whether it's connected to the volume wiper or treble wiper so I don't think that matters.
Set it up so it bleeds, boost on with normal volume up to a point you can hear the bleeding clearly. Use a jumper to short the treble wiper to ground. If the bleeding stops, the CA3046 is not doing it's job, or C40 is bad.

[Enzo]

Not all bleeding is on the schematic. Crosstalk is common and does not follow the signal path.

[g1]

Enzo, agreed on the crosstalk, I was working on the assumption he can kill the bleed by turning off the clean volume, which is essentially what the chip "should" do.

[Enzo]

Yes, I wasn't faulting your approach, it was just a reminder in general. Finding out just WHAT is wrong is the first and most important step. SO all possibilities need to be explored.

[km6xz]

Also, listening to the bleed signal will give clues as to how it is getting where the signal is not intended to go. If it sounds normal but just low in level it is probably the switching is not working correctly. If there is primarily high frequencies, it is likely capacitive coupling between traces or wires that do not connect and is a design problem unless there is some factor like a dried liquid spill on the board. If it muffled, mostly low frequency content it would probably be magnetic coupling or power supply filtering.
As with anything, knowing what is normal for the circuit is the first and most valuable clue. Many techs hear noise and write it off as just plain noise but important clues are imbedded in it as to its origin if one pays attention to its characteristics.

[Wes]

Wow! Thank you for all the quick responses!

I was really racking my brain thinking about how the switching works. I did see other posts dealing with issues with this amp
and the mentions about crosstalk, so I understand what you guys are talking about there.

Yes, I can actually fully eliminate the bleed simply by turning the normal volume down. Any amount of volume there is just
additive, full frequency range (actually sounds pretty good), to the boost channel.

I will test C40 and then short it out with boost on and see if that shows the chip as faulty and go from there.

Again, thanks for your time and help!

[Wes]

Grounding C40 eliminates 95% of the regular channel bleed into the boost.

Ordered a new CA3046.