Marshall JCM 800, Series 2000, Cross Channel Interaction

[g1]

Next, going back to what G1 had me do before, I plug into B and then put a jumper across R22. Yes, this fixes the issue.

The relays are not being used when using A or B inputs. So that was an error I made earlier.
What is going on with the A and B inputs does not make sense to what the schematic shows in terms of the switching logic.
So right now, the best thing we can do is verify the schematic. Post some clear pictures of the trace side of those jacks, showing the jacks, as well as where all the traces go.
Then also pictures from the component side so we can see the component names where those traces are going. A bright light from underneath so the traces show through would also help.

[TomCarlos]

The relays are not being used when using A or B inputs. So that was an error I made earlier.
What is going on with the A and B inputs does not make sense to what the schematic shows in terms of the switching logic.
So right now, the best thing we can do is verify the schematic. Post some clear pictures of the trace side of those jacks, showing the jacks, as well as where all the traces go.
Then also pictures from the component side so we can see the component names where those traces are going. A bright light from underneath so the traces show through would also help.

G1 and Team….

See my UPDATES below...

This has to be the worst schematic to amp documentation I have seen!! I am sharing the images below and notes. But it didn’t take me long to discover these findings once I began the tracing.

It turns out there is a Version 3 of this schematic, no major changes. But the errors on the pots and couple other errors are still there.

UPDATED: The Version 2 schematic only shows a single 68K resistor feeding into V1a and V5a (they should be R3 and R6). However, if you look at the PC board, there are two more resistors, R1 and R2. AND… if you look at the Version 3 schematic, there is a note next to V1a and V5a that says “On Valve Base.” So… we have three 68K resistors for each valve. And I guess we have multiple 68k resistors because of the channel switching. But why have a 68K hanging directly off the valves? You can see one in an image and the other is covered in orange heat shrink.

UPDATED: So here is how it measured out...

UPDATED: Plug into Channel B.
Channel A input is at ground.
Input B connects to R6 and this goes to V5a.
R3 is at ground.
R1 and R2 are floating.

UPDATED: Plug into Channel A.
Channel B input is at ground.
Input B connects to R3 and this goes to V1a.
R6 is at ground.
R1 and R2 are floating.

UPDATED: Plug into the Mix.
The Mix input goes to R1 and R2 (both 68K resistors), where R2 goes to V1a and R1 goes to V5a, through yet another 68K resistor that is connected right on the Grid pins.
R3 and R6 are floating


Next, the 1 Meg resistor connecting to R1 and R2… It appears these are drawn on the wrong side of the 68K resistors. I believed the 1 Meg resistor should be connected directly to the grid of V1a and V5a?

The Footswitch appears to be wired up correctly. One side of the jack has no connections beyond the pins and are closed to the other side. So when the plug is inserted, half the jack is not connected to anything. The other side of the Jack has a connection to ground and the other two pins connect to the footswitch board. See my diagram.

The Mix input is a mess (and I am talking about the schematic). I do not see any changes to wiring – except for fixing traces. So I am of the belief that the traces and wiring are correct. This MUST have worked at some point in time. And if you assume that my moded jack is working correctly, something else must be happening. I checked the diodes on the footswitch board – the seem ok. But it is the relays… testing those might be tricky. I still think something might be wrong with Relay 1.

It's hard for me to redraw the schematic. I think there should be two versions. One without the plug installed, another with the plug installed. I need to see what is happening under each condition. But as I stated, we have to assume this all worked at one point in time.

Oh, how I would love to abandon the mixing and have a single button that gives us Channel A or Channel B only. That is why I previously asked about taking R22 to ground or the wiper of Vr11 to ground, through a SPDT switch. We might have to rewire the inputs and bypass the extra 68K resistors that are not used. We would also need to move the 1 Meg and connect that directly to the Grids of V1a and V5a.

I know, confusing and frustrating.

I'll see if I can get a scope on this and look for the place where the signal is making it's way into A when I am plugged into B.

[g1]

So the key seems to be you mentioning 3 versions of the schematic. So far we have only seen rev.2, and rev.3 is not much different. What is not shown on those is rev.1, which is what you have there.
Rev.2 and 3 corrected the crosstalk issue.
The grid stop resistors at V1a and V5a were located wrong, so the inputs did not get completely grounded when nothing plugged in.
For the update, R1 and R2 were removed and jumpers installed. R3 and R6 were moved to the tube pins and their spots on the board were also jumpered. Looks like that was done on one of your sockets (check that it is 68K), but the rest of the update was not completed. I expect that's why you only get the bleed when one input is used and not the other.
Once these updates are completed you can carry on with the mix jack issues.

Here is the rev.1 schematic that matches your unit. You can refer to the rev.2 schematic to help make the update to rev.2

[TomCarlos]

So the key seems to be you mentioning 3 versions of the schematic. So far we have only seen rev.2, and rev.3 is not much different. What is not shown on those is rev.1, which is what you have there.
Rev.2 and 3 corrected the crosstalk issue.
The grid stop resistors at V1a and V5a were located wrong, so the inputs did not get completely grounded when nothing plugged in.
For the update, R1 and R2 were removed and jumpers installed. R3 and R6 were moved to the tube pins and their spots on the board were also jumpered. Looks like that was done on one of your sockets (check that it is 68K), but the rest of the update was not completed. I expect that's why you only get the bleed when one input is used and not the other.
Once these updates are completed you can carry on with the mix jack issues.

Here is the rev.1 schematic that matches your unit. You can refer to the rev.2 schematic to help make the update to rev.2

G1.... We're making progress !!

As I was going through Rev2 vs Rev3, I noticed some changes on the Power Schematic. But those changes are NOT causing my issue. And YES, it looks like someone made some "some" changes in this amp (but not all) and left us with a hodgepodge collection of new and old.

I too did some digging around and found the Rev 1 circuit - where I spotted the multiple 68k resistors on the inputs. So unless you get to Rev 3, you do not see the note regarding putting the resistors on the valve. And Rev 2 and Rev 3 make NO mention of shorting out the 68k resistors on the PC board.

Ok... here is what I have done

-Put temporary jumpers across R1, R2, R3, and R6.
-Leave the 68Ks where they are (connected at Pin 2 of V1a and V5a).
-The input jack goes to a 1Meg and the Coax that goes to the 68k and tube,
-The other side of the 1 Meg is grounded.
(Let's avoid the Mixing channel for now...)

And voila !! I plug into Channel B and if I turn up the gain control on Channel A, the volume on B remains the same. BUT.... even though the input for Channel A is grounded, I get white noise (from that channel) that comes through the amp. And this is true if I plug into Channel A. If I turn up the gain on B, it does not change the volume level on A but we do get white noise that is audible (from that channel).

Of course, if I remove V1 and V2, Channel B is quiet. So I will need to take a deeper dive with a scope and see if I can locate the source of the noise. Maybe noisy tubes, leaky caps, aging plate resistors, we'll see if I can find the noise source.

I have seen many videos on the old Marshall amps where the problem of noise and grounding problems are discussed. I don't think this is a grounding issue. I think this is plain ole noise.

I'll spend some time on this. Later, I will try and tackle the Mix section. But I do believe I have a faulty Channel A (channel one) relay - save that for another day.

Thanks again G1.

[TomCarlos]

My Monday afternoon update…. and I think I am getting a better grasp on this.

I am plugging my guitar into Channel A (aka Channel 1), volume at 0, and checking tube voltages.

-V1a and V1b seem ok.
-On V2a, Plate = 334, Grid = 1.8mv, and Cathode = 2.99 (.90ma)
-On V2b, Plate = 430, Grid = 36v, and Cathode = 171. (1.69ma)

So I have 1.69ma flowing through v2b less the effect from R18, the 1 Meg. So in this case, the Grid is at 36v because of the connection through R18 to R19/R23? Is that correct?

Now then, I am back at plugging directly into Channel B. The B circuitry seems ok, minimal noise. HOWEVER, I still have white noise coming from Channel A if I turn the Gain control up. If I put my scope on R22, the 15K just before the Channel 1 Send/Return, I see a wave form when I play through B. Grant it, it appears to be (maybe) 15mv peek to peek, but it is there. If I put the scope at the Grid and Cathode of V2b, I hardly see anything.

So is this saying that like it or not, by the fact that Channel A and B meet at the junction of R203 and R204, even if Channel A input is grounded, I could get some noise that makes its way through the circuitry, through R21 (in Channel A), through R204, and now sits at Mix Return/Send and Preamp output? (And as you would expect, if I remove V2, everything goes quiet.

Is there anything else I should be looking at?

[Helmholtz]

[SIZE=16px][FONT=Calibri]

So I have 1.69ma flowing through v2b less the effect from R18, the 1 Meg. So in this case, the Grid is at 36v because of the connection through R18 to R19/R23? Is that correct?

The grid voltage is low, because your meter resistance pulls down the voltage.
Actual grid voltage is the voltage at the junction of R23 and R19 (should be close to 170V).
As there's no (grid) current, the voltages at both ends of R18 must be the same.
Grid bias is the voltage difference between grid and cathode, typically is between -0.8V and -2V with an ECC83.

Remember, the correct grid voltage with a CF or a PI can only be measured using a VTVM having an input resistance around 100M.

[TomCarlos]

Remember, the correct grid voltage with a CF or a PI can only be measured using a VTVM having an input resistance around 100M.

Thank you Helmholtz, that makes sense. And I am making a note of this in my "Lessons Learned" binder.

The readings I posted before are a "skosh" higher than normal because I did not have V5 and V6 installed. But it looks like V1 and V2 are working properly. I'll check V4 and V5 tomorrow.

[TomCarlos]

MEF Team...

I am making progress. All of the voltages for the 12ax7 tubes check out. But as I look around the power amp section, once again, I can see a hodge podge of old, new, and mixture of components that just doesn't make sense. The amp produces sound, no doubt. But what I see is a combination of V1, V2, V3 schematics but parts were not replaced correctly. So I will deal with this as a last step before I bring my friend over to test a fixed amp. And I will ask you all to weigh in. In short, I want to revert back to the Revision 2 Power Amp design. Just for fun, you can view the mess that Revision 3 seems to create. Look after the Choke, the 470 ohm resistor (and I have one of these installed in my amp), and see where that goes. On the top row of tubes, it appears to feed the Screen resistors but I see 1.5K ohm resistors. Now follow it around and see that it goes to Ground. I just see this schematic as all wrong. If it's "ME" who has it wrong, let me know. So again, the Revision 2 schematic for the Power Amp seems to be what you would normally find.

AND BE ASSURED - when I am done with this amp, I might open a new thread, dedicated to posting the schematics only, no trouble shooting, to discuss the three Revisions, Changes, Problems, etc. I am hoping someone out there can find the Review 1 diagram of the Power Amp. I would like to see what that looks like.

I found the ever elusive 348-526 miniature relay from a guy in the U.K. Yes, I am paying dearly for this part. BUT... I am absolutely convinced Relay 1 has a problem. The minute you try to engage the contacts, you can hear them clattering on and off, uncontrollably. Relay 2 seems to be ok. The one thing I am trying to find is a diagram for this six pin, 5vdc relay. If I knew how it works, I might be able to pull it out from the PC board and then use some jumpers to simulate turning it "On" and "Off."

I'm making a list of parts to swap out once I resolve the Mix A&B circuitry. There are some worn pots, smaller electrolytic caps, etc.

It's not just an amp... it's and adventure." I think I need to add that to my signature!!

[g1]

The rev.3 power amp schematic obviously has a few errors. Just a couple lines connected where they shouldn't and a couple not connected where they should.
I will caution about never trying to make an amp match a schematic untill you have verified 100% that the schematic is correct. Schematics are often wrong, or the wrong revision. No one proof-reads them except us.
Here is the corrections to the screen and plate supply lines in the rev.3

[TomCarlos]

MEF Team…

I think I figured this out…

It turns out, the Mix Input jack really wants to be Tip N/O, Ring and Sleeve both N/C.

This was a VERY ODD configuration for the Mix Input jack- I must say. I tested this by making up a temporary stereo jack where I pulled the Tip connectors apart. Once installed on the PC board, I can then plug into A and A works (with the Channel A LED turned on). There is no interference from Channel B. I can then plug into B and B works (with the B LED turned on) and no interference from Channel A. So far, so good. That resolves what type of Jack I need to make for the Mix Input. Good catch from G-1 on figuring out that this amp needs an odd-ball jack.

Now then, onto the Footswitch A & B jack. Using a footswitch cable that has open ended leads, I discovered the following:

-If the Tip wire and Ring wire are NOT connected to the Sleeve wire (ground), both channels are on.
-If the Tip is connected to Ground, Channel B is on.
-If the Ring is connected to Ground, Channel A is on.

So there is NO WAY that the footswitch can be a simple two button, on/off SPST switch - not when you Switch and Add channels at the same time. My friend provided me with a footswitch that he said worked at one time. No way!! Its not that we do not want to trust our friends... however... "Trust but verify..." I posted a photo of this footswitch earlier but I did not think it through. I guess tonight, the light shined.

I will try to figure out how to rig up a footswitch using two buttons. I think it might look like the attached file. If this works, I would love to add LEDs (powered by a battery) onto the footswitch!! So it would be an LED for A , an LED for B, or both A&B are lit up.

On the Switching board and for good measure, I replaced D203 (IN4007), ZD201 (15v Zener) and C203 (220uf Capacitor). I checked D201 (IN914) and D202 (IN914) with my Atlas tested. Both are good. I was thinking there might be something bad on this board. All along, it was the footswitch configuration.

So I will test this again tomorrow... but I think this might be the solution!!


Thank you!!

[TomCarlos]

Team...

I continue to make slow progress on this amp. I have fabricated a few different versions of the Mix Input jack. The issue is that modern jacks have different dimensions compared to what was in the amp before. So I have having to create the TIP's NO sleeve and making sure it will stand the test of time. I am thinking I might make a dongle for that jack - a six inch patch cable with a male at one end, female at the other. My friend would then connect into the female end and leave the dongle in the amp, only using the Mix Input going forward. The reasoning is that I just don't want the internal jack to fail if it gets constant use. These jacks have very narrow tolerances and I am finding that my home made jack has issues when inserting, withdrawing, and repeat the cycle. The other option I have is to separate the Tip connectors and then add a small toggle switch on the amp face- in one position it would be for the Mix Input where we short out the Tip. On the other position, the switch would open the Tip connectors and that would allow for using the Channel A or Channel B jacks. More to come on this.

In the mean time, I was concerned that one of the Electro-Mechanical relays was bad. So silly me, I bought one (and maybe the last one of its kind) from a guy in the UK. So a total of $30 means I now have a backup. This is a mini relay, very small, with an unusual pinout. So I thought I would post a photo here and pinout diagram. This same relay can be found in the Series 2001 Bass Amp Head.

Tom

[TomCarlos]

Gents...

I am doing some testing on this amp before calling my friend to come over and have a listen. So far, the fabricated Mix Input jack is working well. The cross channel interaction is gone!

Talking with my friend today, he mentioned to me that compared to his other Marshall Amp, the Volume control on Channel B seems to have a different sweet spot compared to his other Marshall amp (that has a Gain and Volume control). When I say "sweet spot", I am talking about how far up you turn the knob before it gets loud!! And I just noticed something on the Revision 3 circuit. For VR11, Marshall went from using a 1K Log pot to a 5K linear pot. So if I am looking at this correctly, we have a voltage divider with R48.

Stupid question perhaps, but other than touch and feel or a sweet spot, would a 5K pot allow for more signal to pass compared to a 1K pot? At some point, as you turn that 5K pot it will have the equivalence to 1K?

I am trying to find a simple way to explain this - to my friend and to myself !!

[Helmholtz]

... would a 5K pot allow for more signal to pass compared to a 1K pot? At some point, as you turn that 5K pot it will have the equivalence to 1K?

Yes, there's a voltage divider effect with the 22k series resistor.

[TomCarlos]

Yes, there's a voltage divider effect with the 22k series resistor.

Ok... so we have a voltage divider and my assumption is the 5K pot will allow for a greater signal to pass through - correct? So why not have that value is the previous revisions (unless an oversight)? Just curious.

I also assuming too much signal will overdrive the next stage. I just need to answer the question... do we leave the 1K in the amp or change it out (because there is something to gain by doing so).

Thanks.

[The Dude]

It's a matter of taste and what the user is looking for. The 5K will give you more signal driving the stage because less of it is going to ground. I imagine it might give you a more aggressive sound and possibly earlier breakup. IMO, it's down to what the user is looking for. It's probably not going to be a radical difference. If he's able to get what he wants by simply having the volume on a different number, I don't know why it would be an issue.