Marshall 1959T (year1969) problem Ghost Notes

[binito]

Hi,
I've a Marshall Super tremolo 100w of year 1969.
The problem is with treble notes( 12th fret in one string guitar for example) the amp sounds a double note or ghost note.
I read about this issue which is usually by the 120Hz filter.
I tried change the Filter caps in B+ to 47+47uF caps , and screen filter caps 100+100uF (even 220uF). I change the preamp filter cap to 50+50uF. (all caps F&T).
The PI filter cap is 50+50uF LCR( It is the only one not changed)
I changed all tubes also.
And the ghost note in treble sounds is equal.
I know that is a famous problem in these amps. Someone knows how to remove the ghost notes and the amp sounds correct.
track with the sound problem (second 0:14 ...) https://dl.dropboxusercontent.com/u/...%20%281%29.m4a
Photos: https://dl.dropboxusercontent.com/u/...remolo100w.zip
Thanks

[dai h.]

hi, the bias circuit filters appear to be unchanged from stock, so that might be a possibility for hum to get in and intermodulate the signal

[mhuss]

Did you try it with a different speaker cabinet?

[binito]

hi, the bias circuit filters appear to be unchanged from stock, so that might be a possibility for hum to get in and intermodulate the signal

Today i changed the bias circuit filters and the problem is equal.

[binito]

Did you try it with a different speaker cabinet?

I tried with a 3 cabinets (4ohm , 8ohm & 16ohm). The problem is same.

[binito]

I saw the central power tubes don't have 1k5 resistances. The other two tubes have a 1k5.
It could be the problem?

[pdf64]

It is very beneficial for tubes to be fitted with grid stoppers; it makes no sense to me for only 2 of the 4 power tubes to have them.
So yes, fit them to the inner sockets.
However, I'm not sure will solve the problem.

I hate to download zip files; could you put the files on to an image hosting site and provide links?

If you've definitely eliminated the tubes and power supply caps, I think that the HT 0V arrangement may be a good topic to investigate next.
Just regarding the power tubes, were the sets reasonably well matched, was the bias adjusted for a suitable plate current?
What is the amps HT voltage, eg at rectifier output?

[binito]

It is very beneficial for tubes to be fitted with grid stoppers; it makes no sense to me for only 2 of the 4 power tubes to have them.
So yes, fit them to the inner sockets.
However, I'm not sure will solve the problem.

I hate to download zip files; could you put the files on to an image hosting site and provide links?

If you've definitely eliminated the tubes and power supply caps, I think that the HT 0V arrangement may be a good topic to investigate next.
Just regarding the power tubes, were the sets reasonably well matched, was the bias adjusted for a suitable plate current?
What is the amps HT voltage, eg at rectifier output?

https://dl.dropboxusercontent.com/u/...20foto%201.JPG
https://dl.dropboxusercontent.com/u/...20foto%202.JPG
https://dl.dropboxusercontent.com/u/...20foto%203.JPG
https://dl.dropboxusercontent.com/u/...20foto%204.JPG

B+ = 490Vdc
V5 = 18mA
v6 = 30mA
v7 = 31mA
v8 = 20mA

https://dl.dropboxusercontent.com/u/...0/IMG_6703.JPG
https://dl.dropboxusercontent.com/u/...0/IMG_6704.JPG
https://dl.dropboxusercontent.com/u/...0/IMG_6705.JPG
https://dl.dropboxusercontent.com/u/...0/IMG_6706.JPG
https://dl.dropboxusercontent.com/u/...0/IMG_6707.JPG

[dai h.]

hi, maybe check out R.G. Keen's GEO site (there was a section on ghosting if memory serves). Also, this is an old ampage thread on ghosting (I remembered this (seemed pertinent and more comprehensive) and tried to find it searching the ampage archive search, but it didn't seem to come up) :

START NEW THREAD REPLY Top of Thread LIST
Read 87 times
From: Darren @
Date: 2/15/2002 3:52 PM
Subject: Help please: Ghosting problem, waveform included (long)


Hi! I've been working on my Bluesbreaker RI for quite awhile, since it has a ghosting problem after being modded by a tech for vintage tone.

By ghosting, I mean that there is a low frequency sub-harmonic "under" the fundamental of the note played. It varies with the note (in the same direction pitchwise) and is very unmusical. I've tried everything, different speakers, cabs/impedances, ALL different tubes, including power tube types, different power locations (houses), different OT.

I put the amp back to stock, except for the MM OT, preamp filter cap (16x16u as opposed to stock 50x50u), and all other parts are back to RI spec. I am, and have always been, using Chinese KT66s. I left the 16x16u filter can in because using 50x50u did not help the ghosting. The grounding schem and lead dress will never be stock again. I've played with both of these elements for quite a long time. Then I broke down and bought a scope, a used Tektronix 7603 and a signal generator.

Schem:
raw sewage


I've redone the grounding scheme a few times, and now its very much like Aiken describes here:

http://members.shaw.ca/house-of-jim/...unding_ra.html

The amp is now as quiet as it ever has been.

The last thing I did to it to help hum also helped reduce the ghosting to a large degree. I shielded the cable from V7 (rectifer) to R28 (220K) since it runs a long way, and lies right next to the B+ supply line (R29 power resistor to choke node). I used coax with 100% shielding and an 18 guage conductor.

But alas I still have ghosting. Here's a waveform...

http://www.guitaristforum.com/uploads/PI_out.jpg

Middle waveform is input signal 1x probe on the 100mV scale. I've varied the input signal from 50mV to 100mV to 200mV and get the same results. Larger wave form is a 10x probe at 500mV scale, and we're looking at the plate of the PI tube. There is a little spike on the leading edge of each peak-valley of the waveform. Its not easy to see.

I've traced the signal thru the entire amp, starting at the input to the output. The spike first appears at the PI plate, and is at the power tube grids, and output. IT appears to be of the same amplitude at each of these places!?!? The only ripple on the B+ is a fuzzy 120Hz 10mVp-p.

Any ideas please????

START NEW THREAD REPLY PREVIOUS LIST
Read 79 times
From: r dupont @
Date: 2/17/2002 1:30 AM
Subject: Re: Help please: Ghosting problem, waveform included (long)



I am not an expert but...Have you tried different preamp tubes?
That solved a sizzle type of ghost problem for me once. Perhaps ferrite beads along the lead dress may reduce occilations. Make sure the lead dress does not paralel amplification leads ...cross at right angles.
You say it is a LF subharmonic..does it sound like sizzle?
Ron

START NEW THREAD REPLY PREVIOUS LIST
Read 59 times
From: Darren @
Date: 2/21/2002 3:18 PM
Subject: Re: Help please: Ghosting problem, waveform included (long)



No it doesn't sound like sizzle. It just sounds like a note, an overtone. Its pitch varies in proportion with the note. Its worst at about 348Hz (fundamental).

Preamp tube swapping has not helped. I may have to shield some stuff in the PI circuit though.

Darren

START NEW THREAD REPLY PREVIOUS LIST
Read 46 times
From: Randy Jamz @
Date: 3/9/2002 9:06 PM
Subject: Re: Help please: Ghosting problem, waveform included (long)



If it turns out not to be lead dress, another good guess is the decoupling. If the resistors have been lowered in value, the decoupling between stages won't be as efficient. Also, watch out where each section gets its' B+ from. The preamp should be the lowest B+, followed by the cathode follower driving the tone stack, and then the phase splitter, then the screens, and then the plates. Seems logical, but sometimes it gets screwed up. When the ghosting becomes evident on an amplifier I've built from scratch, I redo it all; the lead layout, and I increase a decoupling resistor. It seems to work, but then I may have had a horseshoe up my bum and just got lucky. I don't think you hurt anything by trying, though. Make sure all of the decoupling resistors are 'correct', and the caps are the right value; they are chosen to have a time constant, and upsetting one might be your problem.

START NEW THREAD REPLY PREVIOUS LIST
Read 29 times
From: stephen conner @
Date: 3/14/2002 5:27 PM
Subject: Re: Help please: Ghosting problem, waveform included (long)



have you made sure that the bias supply to the output tubes is properly smoothed- any hum here will modulate the signal and show up on the scope just as you described. Try replacing the bias supply filter cap. Also make sure the ground point for this filter cap (ie its positive terminal since the bias is negative) is right on the output tube cathodes.

Steve C.

START NEW THREAD REPLY PREVIOUS LIST
Read 29 times
From: Darren @
Date: 3/14/2002 5:55 PM
Subject: Re: Help please: Ghosting problem, waveform included (long)



After more debugging, the ONLY places that ghosting appears are:

1)power tube screen grids
2)power tube plates
3)output.

All components in the B+ & bias supply, power resistor,choke,filter can have all been checked. All PI caps replaced (origionals measure good), NFB path checked.

I've tried many different grounding schems, none effect ghosting, just hum.

The problem is that many solder traces have been rendered partially burned and questionable by the mod'ers soldering work. Note-ably, solder traces in the PI circuit.

Then I find that the ghosting is VARIABLE. One day its worse than another.

I think the PCB is toast.

Thankfully, a new friend in CA is going to take a look at it for me.



p.s.

another thread:

http://music-electronics-forum.com/t15294/

[pdf64]

To me, the issue with binito's amp seems to be HT ripple corrupting the signal path; binito, please could you confirm whether the ghost note frequency varies with the note being played, or does its frequency stay the same whatever note is played?

[Leo_Gnardo]

One seldom considered source of hum/modulation on old Marshalls is the location of the bias supply ground connection. It's usually found where the factory put it: brought to the bus wire attached to backs of all to control pots. Not such a good place. Bring that ground to a chassis mounted lug at one of the filter caps.

Can't be certain this will help but it's worth a try and will only take a minute.

[binito]

One seldom considered source of hum/modulation on old Marshalls is the location of the bias supply ground connection. It's usually found where the factory put it: brought to the bus wire attached to backs of all to control pots. Not such a good place. Bring that ground to a chassis mounted lug at one of the filter caps.

Can't be certain this will help but it's worth a try and will only take a minute.

Yes , Thanks Leo_Gnardo . Now I conect the ground to chassis with cathode power tubes and negative of first filter caps. Unfortunately The issue is equal :-(
I will check ground connections . What would be the best method of connecting the grounds ?

[binito]

To me, the issue with binito's amp seems to be HT ripple corrupting the signal path; binito, please could you confirm whether the ghost note frequency varies with the note being played, or does its frequency stay the same whatever note is played?

thanks pdf64.
The Frecuency varies with the note played. Link with sound more notes:
https://dl.dropboxusercontent.com/u/...0notes%202.m4a

(00:00-01:00 ghost notes)
(01:00-01:25 low volume ,no ghost notes))
(01:25 .....more volume = ghost notes)

[binito]

I tried to install PPIMV with typical A500K stereo .
when pot of PPIMV up to 8 , sound ghost notes.
If the PPIMV below 8 does not happen ghost notes.

[binito]

and a choke problem??