JCM800 2204 Phase Inverter issue and other

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Please measure again where V2 is also removed (while leave 22n unsoldered and V1 removed).

(If 1V disappeared after V2, V1 and 22n are removed, then V2 might have some grid abnormal current leak). With 1V on top 470k, it's 323mV on grid of V2, so this 1V must come from top of 470k. in other word if 323mV come from grid of V2, you couldn't have measured 1V on top of 470k.

I know it seems impossible, but I measured in sequence the following:
- with only the 22n cap unsoldered
- in addition with V1 removed
- in addition with V2 removed
and the voltages are more or less the same. After 15 minutes of tests I've 800mV after the cap (point 1 in the picture attached) and 615mV at pin 2 of V2 (point 2 of the picture). The only strange thing is that this two values are slowly decreasing. At starting they was 920mV and 660mV and slowly decrease arriving at 800-615 in around ten minutes.
Any idea?
 

Attachments

  • 20180919_225531.jpg
    20180919_225531.jpg
    483.8 KB · Views: 160
Disabled Account
Joined 2013
Clean the circuit board where these components lay, using contact cleaner (maybe turpentine) and brush and drain dry with cotton bug. You can verify if any resistance between those terminal lug or tag. I see there are electrolyte capacitors nearby watch for leakage.
 
Last edited:
I try to clean as well as I can, with contact cleaner, brush and drying with cotton, but nothing change. So I decided to unsolder the components and try to wire them in-air (I hope is the right term). All the measures are perfect!

After I measured the voltages on the turrets without anything connected and the result is (see the attached picture, but consider that the components are removed):
- turret T1 --> 450mVdc
- turrets T2-T3 --> voltage continously varying from 2-12-25Vdc and again from 2Vdc.

Consider that turret T4 (quite near) is B+ for V1.
May be thet the board is conductive also if well cleaned? In this case why T5 and T6 haven't problems? What can I do in this case?
Thanks for suggestions.
 

Attachments

  • 20180919_225531.jpg
    20180919_225531.jpg
    484.4 KB · Views: 171
Disabled Account
Joined 2013
If the circuit material is FR4, probably it could catch fire. I once replaced a faulty electrolytic cap which has leakage electrolyte on PCB I forgot to clean, after one week the PCB burnt out about 1" long black track on PCB. Lucky no track is affected. So I just drill out the burnt area and clean, it's fine again.

Consider to lift the entire board, have a look and clean the other side. You can drill small hole to test but take precaution not to drill through the chassis, protect it by putting a sheet of wood or metal between.

BTW is your layout same as this?
 

Attachments

  • Marshall_JCM800_layout.pdf
    505.3 KB · Views: 68
Last edited:
Board already lifted and cleaned also on the back side. I also tested if there are voltages on the surface of the board (near the B+) but nothing. Do you think that it's conductive under the surface? When you say to drill a hole is to test voltage under the surface?

The layout is quite similar. I've potentiometers and tubes on the same side (attached a picture of the whole).

I read on the web about using a heat gun to heat the board. In your opinion can help? Or do you suggest to replace the entire board? :(
 

Attachments

  • 20180922_213312.jpg
    20180922_213312.jpg
    481.7 KB · Views: 153
Disabled Account
Joined 2013
Yes, it will test if the entire board thickness is leak or not, drill holes on the circuit board, around T4 as attached, preferable one at time.

Actually I have soaked the same board in soap water for hours, it does not leak, appear to be water proof. Maybe some defects or porous holes formed too deep, heat gun may only reduced leak on the surface. You probably could feed some heats already as it conducts some electric current, but the heat is not hot enough to burn off the conductive "liquid". If drill holes can reduce the leakage to acceptable level then you can still use it and check later.
 

Attachments

  • JCM800-b.jpg
    JCM800-b.jpg
    178.9 KB · Views: 159
Last edited:
20180923_124927.jpg
I'm quite concerned. Before proceeding how you suggested I try to unsolder the B+ (T3) and mount it in-air to verify if this solve the problem. Nothing changed, I still have 500mVdc at T0. So I try to mount in-air also T3 and this reduce T0 to 250mVdc. Unsoldering and mounting in-air also T2 give a value of 20mVdc on T0 (a good value, I suppose).

But the conclusion seems that every HV turret is involved and the conductive board bring this voltages everywhere. To confirm this I measured the free turrets and now I have 2-3Vdc in this zone (before 15-16VdC) but moving to the right (t0oward the power zone) the voltage raise up to 15-16Vdc in a free turret near the B+ of the PI (now I suspect that the probleam I've on the PI - the difference in plate voltage - may be related to this). I suppose also that the turret connected to ground mitigate the problem. What do you thing about? Any more suggestion?
 
Disabled Account
Joined 2013
Could you also measure the resistance between those turrets? I didn't get any resistance with the eyelets much closer than your board, and I have about 500V HT, there is no leakage.

So far I have experienced with board used in Fender Bassman amp, it is scratchy and got cracking noise if the board is bent. I got voltage everywhere from front end to output stage. There is much leakage resistance between eyelets. I rebuild the circuit board with 1.7mm Bakelite & Fiber Phenolic Resin Flat Plate, so far good.

Leakage in old Fender board will go away after heating up but when cooled it absorb moisture, so the problem never go away. The other problem is the voltage withstanding properties, the higher voltage it got, the worse it get. High voltage can be relieved or reduced by removed some the material and replaced with empty air space esp. if it's in marginal area. If too many material has to be replaced by air space in every HT components, then it's better to get a higher voltage board rather than fixing it. I read that others have encountered the same problem because they admit used or supplied with incorrect type board. So what is the type of board you're getting?

If there are resistance between free turrets then the board quality is in question so that put others problem aside for now. If you measured higher voltage near PI, it maybe due the shorter path (more current flow hence higher voltage). Then you would check actual voltage on the PI and the difference. The current flows from HT to circuit ground but is your PI also signal ground?
 
Last edited:
All my issues were effectively due to a conductive board. In the attached pictures you can see the board modified with my homemade insulated turrets applied to HV nodes. This solve the problem of the voltage after coupling cap of V1a (now around 20mV against the previous 1V) and ALSO the issue of the not well balanced PI (on pin 1&6 now I've 252 and 257V).:)

I've to thank you all for the helps, and specially Koonw for the patience and all the suggestions (and to introduce me in the simulation software world): the next time you come in Italy I'll be pleased to offer you the best pizza&beer in my area.
 

Attachments

  • 20180923_195300.jpg
    20180923_195300.jpg
    294.6 KB · Views: 127
  • 20180923_195327.jpg
    20180923_195327.jpg
    350 KB · Views: 131
Unfortunately it seems that this conductive board generate other problems. For example I set the bias using 1 Ohm resistors and measuring the mV at two ends (33mV --> 33 mA of bias current in my case). But in this situation the negative voltage on the power tubes is around -26Vdc only, probably because the board conduct a positive voltage.

In addition the amp still not reach the full power, and I suppose that also in this case the conductive board has a role. So I'm thinking to substitute the board. A shop near me sells a board with the JCM800 turrets already in place, for 19.99 Euro. The material is bakelite. Your opinion or other suggestion?
 
Hello turk. Discover that my board is conductive was a long journey, and I arrive to the conclusion mounting in-air the component on which I've issues. In this way the issues disappear, but I was still working on the amp idle. Now, souning it, other issues appear.

I don't know what kind of troubles you suppose but there are no leakages, all the resistors were checked, on idle all the voltages are correct (with the exclusion of negative grid on power tubes (as I posted), power tubes are new.

But if you have more precise suggestion I'm here to learn.
Many thanks
 
Disabled Account
Joined 2013
Put the amp in "STANDBY" mode, this will disable the HT, but the bias is still on, measure bias voltage again. Do you do air-lift test on the HT components e.g DC end of diodes in power section? AC for bias could also be affected (say due leak to ground, AC for bias can be reduced on the resistor RB2 120k end), so maybe to air-lift the AC from transformer windings.
 
Last edited:
Hello Koonw, we are here again. Yes, I can try to do what you suggest, but if the problem in effect is the board I've to insulate near anything that can have a Voltage greater than few volts to have a good result. In this case is not better to unsolder all and transfer to a new board? Probably is faster and the result could be better and more sure. What about bakelite I mentioned in my previous post?
 
Put the amp in "STANDBY" mode, this will disable the HT, but the bias is still on, measure bias voltage again. Do you do air-lift test on the HT components e.g DC end of diodes in power section? AC for bias could also be affected (say due leak to ground, AC for bias can be reduced on the resistor RB2 120k end), so maybe to air-lift the AC from transformer windings.
In any case I take the measures with the amp in STANDBY mode:
- 32mV after the Bias trim (before RB1)
- 175mV after RB1(between RG resistors)
- 216mV after RG of V4 / 219mV on pin 5 of V4
- 670mV after RG of V5 / 685mV on pin 5 of V5

With the amp fully on I take the following measures:
- with a Bias current of 12mA on V4 and 13.2mA on V5 I measured -41V at the node between the RG resitors
- with a Bias current of 29mA I've -36.5V at the same node, but when apply a signal (400Hz, 150mA, sinusoidal) to the High input V4 start to red plate
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.