>

Ghosting in Superlead

Discussion in 'The Workbench' started by Inisio, Jun 18, 2021.

  1. Inisio

    Inisio New Member

    Joined:
    Jun 7, 2021
    Messages:
    7
    Likes Received:
    0
    So i got a Super lead MKII '74 on bench right now with quite a bit of ghosting going on.
    According the person who owns it, the ghostnotes was just there all of a sudden.
    After only visual inspecting i saw that the filter caps was bulging it, so time for recap and some maintenance.
    But once done the ghosting was still present.

    What has been done so far is:
    1. Filter caps replaced
    2. Screen resistors replaced
    3. almost every dropping resistor in the B+ since they were way way off.
    4. Balance resistors on the screen filter caps. Upped them to 220k.
    5. Checked all ground connections.
    6. Brand new tubes all the way, checked with other know to be good tubes aswell.
    7. checked coupling caps.
    8. checked diodes.
    9. Measured choke, seems fine at 110ohm. I even tried with a Resistor in its place just be sure.
    10. Measured OT.
    11. Checked all the voltages, everything is within "spec".
    12. Tried with upping the main filtering to 200uf.

    I tried to scope where the ghost notes originates from and i cant quite pin it down.
    I know that with every tube except v1 pulled it is still present after the bright cap, even with the cable from the wiper pulled.
    So i tried to provoke on the normal channel with a bright cap on the volume and it was there as well.

    I'll attach a picture of what the sine wave looks like on the wiper with volume set to 2-3.
    Ghost.jpg

    I presume that that harmonic is my little ghost showing.

    So some hints on what to look for next would be very much appreciated. any ideas?
     
  2. 351

    351 Member

    Joined:
    Jun 8, 2021
    Messages:
    86
    Likes Received:
    48
    Look at the guts of a similar amp, compare the wiring layout of the amp you have there.
    Take note of grid wire positioning for all tubes.
    Is the layout of the amp the same as when it left the factory, many amps have been messed with, perhaps a photo or two of the amp itself would help.
     
  3. Matthews Guitars

    Matthews Guitars Well-Known Member

    Joined:
    Aug 17, 2019
    Messages:
    4,259
    Likes Received:
    6,405
    That kink near the zero part of the sine wave is crossover distortion.

    First, check the bias. You may be biased so cold you're heading past Class B and into Class C.

    Then, go stage by stage to find the crossover distortion.
     
    thetragichero likes this.
  4. 351

    351 Member

    Joined:
    Jun 8, 2021
    Messages:
    86
    Likes Received:
    48
    If you checked ground connections, did you take out all the pots and clean around there where the pots go in? ( clean the front of the pots too)
    This is an area that can cause big problems.
     
  5. Inisio

    Inisio New Member

    Joined:
    Jun 7, 2021
    Messages:
    7
    Likes Received:
    0
    Everything looks as orignal as it gets when it comes to the wiring, the only things that been messed with is that someone snipped the 330uf bypass cap on the normal channel and the bright cap was gone.
    But yeah i am gonna doublecheck just in case.

    That picture is with every tube pulled except for v1, is it really crossover distortion? if i plug in all the tubes again, and look at the sine wave on the output instead i can move the notch angle around with the treble pot.

    That is something i haven't done, good thinking! will do this asap. I just cleaned the pots on the spot.
     
  6. Pete Farrington

    Pete Farrington Well-Known Member

    Joined:
    Jan 18, 2021
    Messages:
    367
    Likes Received:
    359
    The kink / non linearity of the scope trace is probably just distortion from your sig gen, which the bright cap is accentuating. Or perhaps the input signal level is pushing the grid a little too hard, although the symmetrical nature of the kink makes that less likely. Whatever, it’s not significant or pertinent, a red herring.
     
    Travis398 likes this.
  7. myersbw

    myersbw Well-Known Member

    Joined:
    Oct 22, 2015
    Messages:
    1,253
    Likes Received:
    716
    Location:
    SW Ohio
    Hang that probe off the speaker and let's see what you have...all tubes in. If you hear it, it should appear like a phase shifted duplicate (and likely distorted) version of the original signal. (Not small and insignificant like what appears to be a bit of crossover distortion. (And it's location is a bit premature for crossover.)

    Also, make sure you've properly adjusted the probe compensation so a square wave has no over or undershooting spikes.

    One more thing...did you replace the bias circuit electrolytics? (I'm assuming you're replacing quite a few because they're likely original?)
     
    Last edited: Jun 19, 2021
    NewReligion and Pete Farrington like this.
  8. neikeel

    neikeel Well-Known Member

    Joined:
    Feb 12, 2014
    Messages:
    6,017
    Likes Received:
    4,741
    You've done all the things that I would normally do (and have done in the past and typically one of the things solves the problem - particularly odd as it is a new problem in that amp, so presumably a component failure.
    I have same on my Princeton build sounds good but still ghosts a bit (I even added a choke) despite all of the above (incl extra filter node in preamp too) not got my scope out yet -that is next!
     
    Joshabr1 likes this.
  9. Pete Farrington

    Pete Farrington Well-Known Member

    Joined:
    Jan 18, 2021
    Messages:
    367
    Likes Received:
    359
    Yes, scope the output at high power, at thd onset of clipping.
    Measure the VDC and VAC at the OT and screen grid HT nodes, at idle, and at the inset of clipping.
    Ghosting is often used to describe the effect of a vast excess of HT ripple, due to poor filtering. Lots of intermodulation of the audio signal by the ripple, creating ‘sum and difference’ frequency products, and amplitude modulation of the audio at the ripple freq.
     
  10. tschrama

    tschrama Well-Known Member

    Joined:
    May 15, 2010
    Messages:
    1,240
    Likes Received:
    523
    Location:
    Netherlands
    Ghosting is intermodulation distortion.

    If there is humm on the bias voltage, the humm would cancel out on low signal levels in the pushpull stage , but would become appeared as ghosting on higher signals.

    I would check the bias voltage for ripple. And I would see what effect a higher or lower bias point has on the ghosting.

    If the ghosting becomes worse with lower bias current, and less with higher bias current, then that would be an indication that the bias voltage has ripple.
     
    Pete Farrington likes this.
  11. tschrama

    tschrama Well-Known Member

    Joined:
    May 15, 2010
    Messages:
    1,240
    Likes Received:
    523
    Location:
    Netherlands
    Btw 1

    Turn off the treble and mid pot, so the harmonix of you sine-source dont get accentuated.

    Btw 2
    Ghosting, or intetmodulation in a guitar amp had nothing to do with the preamp stage, so I would focus entirely on the power stage.
     
    Pete Farrington likes this.
  12. Inisio

    Inisio New Member

    Joined:
    Jun 7, 2021
    Messages:
    7
    Likes Received:
    0
    This could be the case, looking at the sine wave with probes hooked up to the signal generator there seems to be a tiny tiny kink at the peak.

    Yepp, the electrolytics on the bias has been replaced aswell. The picture below is with probe on the Speaker jack with everything set to noon except for volume that is set to 2 which should give a clean signal.
    Scope output.jpg

    If lets say the bias has ripple, what would cause that? i mean the bias circuit has been recapped as well. I tried going hotter on the bias with 45ma, but it was still there.
     
  13. 351

    351 Member

    Joined:
    Jun 8, 2021
    Messages:
    86
    Likes Received:
    48
    Have you checked the pot to chassis grounds, bias circuit would ground there..
     
  14. Joshabr1

    Joshabr1 Well-Known Member

    Joined:
    Sep 2, 2013
    Messages:
    1,424
    Likes Received:
    678
    Location:
    Ky
    I like a small amount ghosting but am probably in the minority on that. I had a 74 100 watt that had some ghosting after a cap job. Switched caps from JJs to ARS and that cured it. But all my favorite old Marshall’s have some amount of ghosting going on.
     
    Last edited: Jun 19, 2021
    nickfox likes this.
  15. tschrama

    tschrama Well-Known Member

    Joined:
    May 15, 2010
    Messages:
    1,240
    Likes Received:
    523
    Location:
    Netherlands
    Hmmm. Well i’m puzzled then
     
  16. Inisio

    Inisio New Member

    Joined:
    Jun 7, 2021
    Messages:
    7
    Likes Received:
    0
    Yes, pulled all the pots and cleaned the fronts, the holes and reflowed the bus connections. No difference sadly.
     
  17. Pete Farrington

    Pete Farrington Well-Known Member

    Joined:
    Jan 18, 2021
    Messages:
    367
    Likes Received:
    359
    See
    and
     
  18. myersbw

    myersbw Well-Known Member

    Joined:
    Oct 22, 2015
    Messages:
    1,253
    Likes Received:
    716
    Location:
    SW Ohio
    Well that sine wave distortion is symmetrical, I'd think the power tube section is ok...pull the tube before the phase inverter and inject your signal there. Lets see what the signal generator wave looks like at the source and speaker output with only PI + power tubes. The process is to conquer & divide to start honing in on the culprit. Did I read you right earlier in that this issue was there only when V1 was in?
     
  19. Inisio

    Inisio New Member

    Joined:
    Jun 7, 2021
    Messages:
    7
    Likes Received:
    0
    Here's the on the output with Treble and mid at zero. I reckon that fuzzy peaks is just interference from my room.
    Output 2.png


    Here's the on the output with my signal tapped in just before the PI
    PI Output 2.jpg


    yeah, but like Pete Farrington said earlier, What i saw on the scope with only V1 could just be a red herring. Seems like i have some distortion coming from my Signal Generator.

    So i have measured every sinlge component in the circuit now, and all of them measures good with tolerance considered.
    It seems unlikely, but could a Coupling capacitor or bypass cap produce a ghost note? i mean, they measure their correct values and none of them seem to be leaking any DC.
    One thing that stands out somewhat is this, if i have my scope hooked up to the speaker jack the sinewave gets very fuzzy if i try to measure the voltage on the grid of the 22nf side of the PI. If this is relevant i do not know.
     
  20. Pete Farrington

    Pete Farrington Well-Known Member

    Joined:
    Jan 18, 2021
    Messages:
    367
    Likes Received:
    359
    Trying to measure the LTP grids produces a nonsense reading, as the meter loading pulls the grid voltage out of whack.
    Near as dammit, the grid voltage will be the same as the node at the top of the 10k tail resistor.
    To confirm that, masure the VDC across a 1M LTP grid leak; it should be 0V, +/- a few mV.
     
    Last edited: Jun 20, 2021

Share This Page