Amp squeling when i rotate the gain knob

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by em07189, Dec 7, 2016.

  1. em07189

    em07189 Member

    Messages:
    15
    Joined:
    Sep 2, 2016
    Hi i need some help here to find what is the problem:

    Well, after i have done some mods to my Marshall dsl 15h the amp now starts squealing when i rotate the gain knob to add gain. If i rotate the gain knob up to 1 the amp doesnt squeal, but after this position if i rotate the knob to add more gain the squeal changes pitch and gets louder.

    What could be the problem here? if the squeal changes pitch could the problem be a capacitor?

    The amp has this problem even with anything plugged into the amp. i also have checked the preamp tubes ans eveything is allright.
     
    Last edited: Dec 7, 2016
  2. Blue Strat

    Blue Strat Member

    Messages:
    30,070
    Joined:
    Aug 2, 2002
    Location:
    Sterling, VA (not far from Washington DC)
    What mod did you do? Schematics? Photos?
     
  3. em07189

    em07189 Member

    Messages:
    15
    Joined:
    Sep 2, 2016
    Ok the squeal started after i have done the following mods:

    In the preamp:
    R34 changed to 220K
    C58 changed to a 500pf 500V silver mica cap

    in the poweramp:
    R91 to 10K
    C69 to 4,7 nf , 500 volts cap

    Before the mod described above, the amp had allready the following mods:

    in the preamp:
    C16 changed to 1 uf
    R15 changed to 2,7K
    C75 removed
    R80 removed
    C76 removed
    R41 removed
    R40 changed to 2,7K
    C34 changed to 0,68uf
    R51 removed

    The Marshall dsl15h schematic is in the link bellow:
    https://www.dropbox.com/s/0ekws3lamtrj1oy/DSL15-60-02-v04.pdf?dl=0
     
    Last edited: Dec 8, 2016
  4. em07189

    em07189 Member

    Messages:
    15
    Joined:
    Sep 2, 2016
    This problem only happens in the overdrive channel, any suggestions what could be the problema?

    I cannot find what it is?

    Changing R34 from 18K to 220K , could cause this problem?
     
    Last edited: Dec 8, 2016
  5. Blue Strat

    Blue Strat Member

    Messages:
    30,070
    Joined:
    Aug 2, 2002
    Location:
    Sterling, VA (not far from Washington DC)
    I haven't been able to find R34. Can you describe where it is?

    Any change of 10:1 or more should be suspicious. Where did you get this mod from?

    Change it back to 18K, or some value in between, and see what happens.

    Are you absolutely sure that the preamp tubes are not microphonic? High gain mods bring out the worst in preamp tubes.
     
    zzmoore likes this.
  6. Jeff Gehring

    Jeff Gehring Silver Supporting Member

    Messages:
    4,579
    Joined:
    Jan 17, 2008
    Location:
    Kansas City area
    Your R34 change increases the signal at V3A grid by eight times. Yes, that could cause instability problems.
     
    doctorx and TEPR like this.
  7. em07189

    em07189 Member

    Messages:
    15
    Joined:
    Sep 2, 2016
    Hi Jeff,

    Changing R34 to 220K would make the squeal changing pitch when i rotate the gain knob?
    Well after i have changed R34 to 220K, i have noticed more deep bass and more gain in the bass strings which i liked.

    I'm going to put a lower value resistor in R34 and see if the squeal stops.
     
  8. UsableThought

    UsableThought Supporting Member

    Messages:
    1,620
    Joined:
    Jul 18, 2015
    I'm not sure, but it looks to me like C25 in conjunction with the voltage divider (R32/R34) forms a low-pass attenuating filter. That means it attenuates everything above a chosen frequency, with the attenuation increasing as the frequency increases. This might explain why you notice "more deep bass" even though you also have an annoying squeal.

    Someone may need to correct my math, but here goes a try: With R34 at 18KΩ, the filter would have been attenuating everything above 9.3KHz, which is a pretty darn high note (here's an irritating 10K pure sine wave to give you the idea). When you changed R34 to 220KΩ, that brought this pivot point down to about 1.2KHz, which is quite a bit lower (here's a 1K sine wave, pretty close).

    Even if my math is off, if the divider plus C25 really does form a low pass filter, this means that what you are hearing as "more deep bass and more gain in the bass strings" is really "frequencies above X Hz getting progressively rolled off, so everything below that sounds louder by comparison when you turn up the gain."

    This might also explain why the pitch of the squeal seems to change as you turn up the volume. If the pitch seems to rise as you do this, that would fit: as the volume goes up, high frequencies that had been inaudible due to attenuation start to become loud enough to hear.

    Anyway, if you really want to keep this effect of "more deep bass", that ought to be easy. First put R34 back to 18K, as it used to be, so you get rid of the crazy gain & with it the squeal; then make C25 bigger. I think that 7n or 8n, instead of the present 1n, would do the trick, getting you pretty near 1.2KHz for the point where attenuation of higher frequencies begins; someone else should verify my math & correct it if necessary. The formula I'm using, borrowed from p. 95 of Merlin Blencowe's guitar preamp book, is

    f = 1 / (2 pi C ( R1 × R2 / R1 + R2) )

    where all values of course have to be full units - in this case farads not nanofarads. Note this kind of low-pass filter plus interstage voltage divider is also described on this page about preamp mods, by an unattributed author.
     
    Last edited: Dec 8, 2016
    pdf64 likes this.
  9. pdf64

    pdf64 Member

    Messages:
    5,490
    Joined:
    Aug 9, 2008
    Location:
    Staffordshire, UK.
    It's a miserable job trying to search that schematic for part references.
    I don't see how the amp could work properly without R41 (unless C27 was replaced by a short)?
     
  10. UsableThought

    UsableThought Supporting Member

    Messages:
    1,620
    Joined:
    Jul 18, 2015
    Download the PDF, and you should be able to do full text search inside whatever PDF reader you use. All the component names have been OCR'd.
     
    pdf64 likes this.
  11. em07189

    em07189 Member

    Messages:
    15
    Joined:
    Sep 2, 2016
    UsableThought i forgot to mention that i also removed C25, because it did make the amp more "edgy", The 1 nf removed almost all the mids.
    IF i put back R34 back to 18K, what other mods do i have to make to the result became similar to 220K resistor?, taking in cosideration that C25 was removed, when R34 had the value of 220K


    PDF64, maybe the problem could also be the remove of r41, because when i removed R41 i have put a 220K resistor in C27 position, but i didnt like the result because it added a lot of bass and the amp seemed that have loss dynamics, now i have putted back the 47nf in C27 with R41 removed.
     
    Last edited: Dec 8, 2016
  12. UsableThought

    UsableThought Supporting Member

    Messages:
    1,620
    Joined:
    Jul 18, 2015
    Well, it completely destroys my theory, anyway. You will have more high frequency above 10KHz without that cap; and overall gain will go up due to this. I have no idea why you would seem to hear more bass. I seem to have mislaid my parachute; does anyone know where it is?
     
  13. em07189

    em07189 Member

    Messages:
    15
    Joined:
    Sep 2, 2016
    Hi again UsableTought and thanks for the help , Well C27 with the value of 47nf from what i have notice takes some bass of the early preamp stages and makes the amp less muddier.
     
  14. makerdp

    makerdp Member

    Messages:
    348
    Joined:
    Nov 24, 2016
    A tube amp that squeals when you turn up the volume is almost always because the output transformer leads are reversed. Flip them and see if it goes away. If not, switch it back your mods did something. Your mods may have just pushed it over the edge or lowered the threshold to where it kicks in earlier now though.
     
  15. UsableThought

    UsableThought Supporting Member

    Messages:
    1,620
    Joined:
    Jul 18, 2015
    Thanks for clarifying what was happening with C25. At least that does confirm my guesstimate that C25 was acting to set up a low-pass attenuating filter. Filters are not my strong point, esp. when math is involved.

    It's hard for me to answer this for a couple of reasons:

    First of all, I am trying to imagine what this amp could possibly sound like with R34 as 220K. You said in your opening post that you can't rotate the gain knob up past 1 before you get squeal. Does that mean the improved bass gain (which I gather is the improvement w/ 220K resistor for R34 that you want to keep??) is something you're hearing pretty well provided the gain is low? I ask because I would have thought that the squeal would get in the way of really listening to how the amp sounds w/ that mod in place.

    Second, I am wondering what your goal has been with all of these mods. From your comment #3 it looks like you did two waves of mods, with the squeal starting after the second wave (which included the R34 swap). But you haven't said how you went about the mods. Did you make the component swaps all at once, in either case, or component by component? Where did you get the idea to change these specific components to these specific values? And what were your goals for both the first wave of mods, and the second wave of mods?

    It's pretty much impossible to know what you are after with changing R34 to 220K without knowing at least a little about the above questions. Maybe a short way to put the question is, was it a "happy accident" that 220K seemed to produce more bass, or was that an outcome you were hoping for?

    EDIT: Also I gather this amp actually has 2 channels ("classic" and "ultra"), and thus separate gain/volume controls for each - at least, that's what pictures of the front panel show, see below. From the schematic, it looks like both gain pots are upstream of the voltage divider w/ R34. So does this mean that you hear the squeal when you turn up the gain for either channel? Or just one of the channels - & if so, which one?

    [​IMG]
     
    Last edited: Dec 9, 2016
    SDMF38 likes this.
  16. Blue Strat

    Blue Strat Member

    Messages:
    30,070
    Joined:
    Aug 2, 2002
    Location:
    Sterling, VA (not far from Washington DC)
    Try a smaller value than 220K, start with about 50K and work up to about 100K. I don't know where you got the 220K "mod", but I can't imagine it working except to make that stage an oscillator (which you've already experienced).
     
  17. em07189

    em07189 Member

    Messages:
    15
    Joined:
    Sep 2, 2016
    Yes you are rigth, i remenber now that i might have changed the R34 to 220K first and played a little bit with the amp without the squealing and liked the result that i have described, then i modded the negative feedback and after this it started the squeling i also damage a little bit the tracking terminals on the R91 , i'm going to put a lower resistor in R34 and see what happens.
    Would a damaged terminal of the tracking in the R91 resistor cause the squealing? i have to check this also.

    Also what i'm not understanig is if the change of the R34 to 220K is a problem in the V2B stage, why is not a problem in the V2A stage?, because the values of the resistors are almost the same, the R49 is a more higher value than R34.

    My goal is to achieve a more sounded Marshall amp, with a more tight bass response the cathode capacitor and resistor changes to 0,68 and 2,7K, helped to tighting the bass respose but the amp missed a more deep bottom end. The changes in R34 was a advice in other forum , and was a outcome that i were hoping for.

    I'm only hearing the squeal in the ultra chanel.
     
  18. em07189

    em07189 Member

    Messages:
    15
    Joined:
    Sep 2, 2016
    I think i have found the problem, after checking the tubes for microphonic hammering them with a pencil i saw that v1 made noises, so i have changed the v1 with other tube, after hammering again the v1 tube , noises again so i have replaced again v1 with other tube and i get the same noises, so i realized that is impossible to have three microphonic tubes, so i have start to hamer the components near v1 and i realized that when hammering the capacitor C17 this produce noises, so the problem must be C17. right?
     
  19. UsableThought

    UsableThought Supporting Member

    Messages:
    1,620
    Joined:
    Jul 18, 2015
    Thanks for the detailed answers. I'm actually hoping that someone who knows this amp, or Marshalls in general, will step in & give you better advice than I can. With your answers, I think things are a bit clearer, so maybe that will encourage more comments.

    The only other thing I have to suggest for now is, solve one problem at a time! Too many mods at once become a moving target, impossible to debug. In particular, if I were you, I would definitely follow Blue Strat's advice & before anything else make R34 smaller until you get that squeal on the Ultra channel to go away. Good luck.
     
    Advisable Owl likes this.
  20. Advisable Owl

    Advisable Owl Member

    Messages:
    1,192
    Joined:
    May 1, 2013
    Location:
    The Ohio State University
    V1 is pretty much always sensitive to noise and micro phonics.

    Your methodology of going through some of these mods is really flawed and has left you with a lot of potential culprits.

    The thing with Marshall
    Amp is that there really isn't a ton of gain on the plexis, just a **** ton of volume, easily driven speakers, and a design that pushes some of the components passed their intended limits.

    As to where you want to go from here, know that an 820//.68uf has a cutoff freq of 285hz or so if I did my math right. Everything above that is getting full gain while the Lower freqs you are getting are getting attenuated quite quickly.
     
    pdf64 and UsableThought like this.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice