Ghost Notes Solved

glip22

glip22

Well-known member
I can do alot of things on an old Marshall. I can replace everything there is from transformers to pots to caps. and do extensive mod work as well. One thing I can never track is the source of ghost notes.
Speaker cry, edge yowl, etc always ruled out. I do not have a signal generator or an O scope. I have experience. For example, recently I had some minor snap crackle pop. I went right to the 1st stage plate load resistor and replaced it. Problem solved. I knew it was that based on what I was hearing. I Where do I look? Chris Merren told me years ago the secret is in the PI. I know about PI tilt as a cause but how do I solve it? Tubes obviously ruled out first and all filters replaced including bias caps . My assumption is without equipment it's a rabbit hole because it could be anything from a far drifted resistor to a leaky coupling cap to a bad ground. I will not replace everything on an old Marshall as the exact parts don't come easy. Post up if you are a ghost buster
 
I thought so. Even the really good techs have trouble with it. It's considered to be part of the sound by many in an old Marshall. Some have it very faint at volume and others more so. It's not usually an OT but I have had a PT once act up at volume where it was causing a voltage drop; not ghosting but an odd warble of the notes when doing bends.
Some transformers in the old amps have been run with incorrect impedance loads or no load at all multiple times over the long life. I assume if done too many times it can some issues. Maybe I will start replacing Piher resistors and mustards from the PI down and see what happens. :unsure:
 
I’ve had PI issues or I probably should say lessons learnt on some other recent home brews. My Brown Deluxe clone was fistfucking one power tube and pinky fingering the other. The PI is very Marshall except it has a 6k8 tail instead of 10k which even further imbalances the PI. So I increased the 6k8 to 10k and instant improvement. Now the tubes were driven more evenly. So then I measured the PI plate resistors and since they were Carbon Comp they had drifted. The 82k plate tends to dominate the 100k plate anyway but they had drifted such that it dominated even more. So I swapped them for more accurate values and again instant improvement.
 
My 82k is close but I suspect the 100k drifted up.I’ll have to take it out and test it. PI tilting too far on one triode I think can cause ghosting but I’m not sure
 
All I know is my 69 plexi ghosts when cranked above 7 or so. Cab is irrelevant. Caps all in good order. Just a nature of the amp on the old ones, I guess
 
It's ripple on the power supply intermodulating with the power amp. Two things exacerbate it: power amp imbalance and small B+ filter caps.

To solve the first thing make sure your phase inverter tube is in good shape and the plate resistors haven't drifted. Also be sure to use matched power tubes.

Increasing the B+ filter caps will reduce it as well but will affect the feel of the amp as it will change the time constant.

IME 100W Plexis are worse because the supply sags more so it gets more ripple on it. A beefier power transformer is another solution but that's an expensive one.
 
If I remember correctly from Metroamp esepecially with Marshall Plexi' and Superleads low filtering on the Phase Inverter, screens and Preamp tended to produce more ghosting. The early 68 plexi ran 64uf on the PI and that was increased to 100uf in late 68 into 69 as well as the 69 filtering was upped to 50uf on the screens and 100uf on the phase inverter.

I seem to remember people having more ghosting issues running anything less than 100uf on the PI and running 16/16uf on the preamp.

My 68 runs 2-100uf inseries for 50uf mains, 2-32uf in series for 16uf screens, dual can 32uf/32uf preamp, and dual 50/50uf can in parallel for 100uf Phase Inverter. With this of the stock 69 filtering I have not had any ghosting issues even variacing the amp down to 70-80 ACV. I usually run the amp on 7 bright channel volume. My 69 SL with new ARS 50/50uf caps had no ghosting either. If they did it was so minimal I do not really hear it.
 
Last edited:
cliffc8488 said:
It's ripple on the power supply intermodulating with the power amp. Two things exacerbate it: power amp imbalance and small B+ filter caps.

To solve the first thing make sure your phase inverter tube is in good shape and the plate resistors haven't drifted. Also be sure to use matched power tubes.

Increasing the B+ filter caps will reduce it as well but will affect the feel of the amp as it will change the time constant.

IME 100W Plexis are worse because the supply sags more so it gets more ripple on it. A beefier power transformer is another solution but that's an expensive one.
Click to expand...
+1 to all of this
 
glpg80 said:
+1 to all of this
Click to expand...
Thanks. 75 Superlead. Its fully recapped and has 50uf/50uf caps. I suspect the 100K plate resistor has drifted up. I am reading 164K but this is in circuit so it can be hard to tell. Teh 82 k reads 80k in circuit. PI is very closely balanced as well as power tubes are matched. Can the 100k plate resistor cause this?
 
glip22 said:
Thanks. 75 Superlead. Its fully recapped and has 50uf/50uf caps. I suspect the 100K plate resistor has drifted up. I am reading 164K but this is in circuit so it can be hard to tell. Teh 82 k reads 80k in circuit. PI is very closely balanced as well as power tubes are matched. Can the 100k plate resistor cause this?
Click to expand...
No not at all. Only the phase inverter plate resistors being out of tolerance can cause any issues with ghosting. If it’s the original 50/50 cans then you either need to replace the caps or consider upping the filtering or both.
 
glpg80 said:
No not at all. Only the phase inverter plate resistors being out of tolerance can cause any issues with ghosting. If it’s the original 50/50 cans then you either need to replace the caps or consider upping the filtering or both.
Click to expand...
Thats is what I measured in circuit. The 100k plate load resistor. I fully recapped it with ARS. I have recapped probably 30 Marshalls
 
glip22 said:
Thats is what I measured in circuit. The 100k plate load resistor. I fully recapped it with ARS. I have recapped probably 30 Marshalls
Click to expand...
Did the amp ghost the same with the old filter caps? Have you checked the new filter caps to make sure they are within spec?

Just throwing some ideas out there. I hope you get it figured out. If you suspect that 100K PI resistor then pull it out and test it and or replace it.

Here's a video with Steve Fryette working on Dave Friedmans 68 Marshall and he goes in depth on the effects on the Phase inverter, ripple and crossover distortion. You may find it interesting.
 
harddriver said:
Did the amp ghost the same with the old filter caps? Have you checked the new filter caps to make sure they are within spec?

Just throwing some ideas out there. I hope you get it figured out. If you suspect that 100K PI resistor then pull it out and test it and or replace it.

Here's a video with Steve Fryette working on Dave Friedmans 68 Marshall and he goes in depth on the effects on the Phase inverter, ripple and crossover distortion. You may find it interesting.
Click to expand...

It did the same thing with the old caps. I have recapped maybe 30 Marshalls with ARS and never had a problem so I doubt it is the caps. I will check out the video. Appreciate it
 
You must log in or register to reply here.
Back
Top