How to cure JTM45/100 ghosting without altering tone?

[vintage_charlie]

Thanks for all the tips guys and sorry for not getting back to the topic sooner. Due to my living situation i moved the JTM 45/100 to my flat and i run it into a load and then into my DAW for cabinet/room simulation. My previous test that showed most of the ghosting was conducted under exactly these circumstances.

To my big surprize, the ghosting is basically gone after moving the head to my flat (it might still be there in minimal amounts, but at a point where it is hard to pick out). So it might have to do something with the wall voltage at my house? / orientation / location of the amp? or some sort of interference that is not present at my flat / a cable moved during transport?

I am not imagining things - the amp did have a text-book kind of ghosting going on previously.

So i am good for now - really happy with the tone now - no problems with neck pickup on les paul, even when cranked - nice thck solo sound. If it creeps up again, i will have some good starting points thanks to your suggestions here. So, once again, thanks guys and have a good one!

VintageCharlie

[Roe]

I will experiment a little with my underfiltered dual rectifier amp (1967 100w) to see if I can minimize the ghosting.

The following worked but changed tone and response too much:
- upping filtering
- 5h or 8h choke (modest effect)
- CLC filter (sic), 3H, 500mA
The latter works fairly well, but the amp sags more and sounds smoother and cleaner (too smooth for me)

The following things worked well so far:
- matched power tubes
- coldish bias
- improved grounding scheme
- formed RIFA caps (BC also good)
- full wave bridge rectifier bias supply (sic)
- decreasing low bass by changing signal caps slightly (e.g. 3n3 on v1b, 1nf PI input, 56nf PI caps)

I play to try the following:
- 6.7k presence pot for increased NFB
- 5nf or 10nf phase inverter input on V3B (instead of 100nf)
- balancing PI plate resistors, following calculator but adjusting for tube imperfections
- 390k input resistor on v3a, ala park 1210

[V8]

Delayed thanks Roe, just fired er up & not noticing any issues so I'll keep the 3H in there.


VintageCharlie, glad to hear your amp fixed itself

David

[Roe]

I play to try the following:
- 6.7k presence pot for increased NFB
- 5nf or 10nf phase inverter input on V3B (instead of 100nf)
- balancing PI plate resistors, following calculator but adjusting for tube imperfections
- 390k input resistor on v3a, ala park 1210

The two first worked fine. Two latter had little effect on ghosting. Will do more tests tomorrow

[Roe]

More testing gave a clear result. Balancing PI plate resistors worked the best. I keept the two 1m resistors on v3 but used 82k/95.5k for the plate resistors.

The ghosting is noticable but only at extreme settings when you really try to produce ghost tones. This amp doesn't ghost more than many 100w plexis with 50uf on the screens and mains.

[VintageCharlie]

Roe, thank you very much for following up the topic and stating your experiment results! This sounds very promissing. It actually sounds like what CHris might have described as the non-bnad-aid-fix.

So after all the things you tried and listed here, this gave the most audible effect?

Thanks!

Charlie

[Roe]

Thanks Charlie! I ended with the following specs:

6.5k presence pot for increased NFB
- 10nf phase inverter input on V3B (instead of 100nf) and 750pf on V3A (instead of 22nf)
- balancing PI plate resistors, following calculator but adjusting for tube imperfections
- 5nf and 3n3 coupling caps at v1 and 56nf on v3 (PI)