VHT Special 6 and the Parasite SOLVED!

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by RussB, Apr 25, 2016.

  1. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT
    [​IMG]






    I got myself a Special 6. Cool little amp and made to be tweaked. The issue I've had from the get-go is a nasty parasitic buzz on the notes as you turn up the volume. It "sounds" to me like a parasitic coupling, but I could be way off base.

    So far, I have shortened all grid wires including using a shielded wire to V1A's grid with the shield grounded and re-routed the heater wiring. I've installed a choke where R24 was. This makes no difference regarding the buzz.

    Just about every cap and resistor has been swapped, mostly just tweaking. Filter caps were replaced with F&T cap cans, a 50-50/500v and a 16-16/450v. The cathode bias cap was swapped for a 100uf/100v, and the bias resistor is a 250 ohm. I am running a 6L6GC power tube at around 85% dissipation.

    I have increased the power tube's grid resistor (R19) incrementally to 100k, thinking that blocking distortion was at play. I have installed a 470pf cap from the power tube grid to ground which helped a lot, but increasing it doesn't help any more.

    I have both of V1's cathode bypass caps on switches, and when switched off, eliminate the parasitic buzz. Basically what happens is when gain is increased, the parasitic buzz rears its ugly head.

    My problem is an inability to actually locate the source of the buzz. Any tips or advice appreciated.



    [​IMG]

    [​IMG]

    [​IMG]

    Note this pic does not show the 470pf cap from grid (pin 5) to ground, but it is there now

    [​IMG]
     
    Last edited: Apr 25, 2016
  2. UsableThought

    UsableThought Supporting Member

    Messages:
    1,622
    Joined:
    Jul 18, 2015
    Just a question - how do you do your shielded wires? I make mine up rather painstakingly with some shielded 24AWG I had left over from a guitar harness project; the shield braid is uninsulated so I have to add heat shrink. I haven't found a good source yet for 22 or 24 w/ outer insulation already made. Yours look quite nice from the pics?

    Also, for buzz, post a sound clip?
     
  3. pdf64

    pdf64 Member

    Messages:
    5,499
    Joined:
    Aug 9, 2008
    Location:
    Staffordshire, UK.
    Does the tone control have any effect on the buzz?
    As the build is compact, there's a fair amount of gain, and lot's of opportunity for unintended coupling via stray capacitance, this sort of issue can be expected.
    To verify whether this is just a gain/bandwidth stability issue, try adding a low pass filter or two, eg 100 to 220pF from each triode plate to 0V, to see if that helps to reduce or get rid of it.
     
  4. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT

    I got it from either Tubesandmore.com or Hoffmans.com.

    I'm gonna have to figure out how make & post a clip one of these days, although I've been saying that for years! The buzz gives the illusion that a string is buzzing on a fret, or a speaker has a bad voice coil
     
  5. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT
    No


    Will do. Should I put these directly on the socket?

    The 100k power tube plate load resistor only knocked some high end off the signal, and did nothing towards eliminating the buzz. The amp became dull sounding
     
  6. pdf64

    pdf64 Member

    Messages:
    5,499
    Joined:
    Aug 9, 2008
    Location:
    Staffordshire, UK.
    It shouldn't make any difference where they go; it's a different thing to grid stoppers.

    You meant grid stopper, rather than plate load, right?

    I think that it may also help if R16, the grid stopper for the second stage, was moved to the socket terminal, and increased in value to 10k - 22k; perhaps try that first, before the caps to ground from the triodes plates.
     
  7. TimmyP

    TimmyP Member

    Messages:
    2,473
    Joined:
    Nov 1, 2007
    Location:
    Indianapolis
    If the tone controls have no effect, the issue is in the V1B and/or B2 circuits I would think.

    Check for ripple on the power supply nodes. Poor filtering can cause issues between sections.

    Measure the DC voltage across R20/C15 and figure out the bias current to make sure that it is within the acceptable range for the B+.

    If R19 is not on pin 5 of the socket, move it off the board to the socket. And go back to the stock value, as it should be plenty.

    Borrow an oscilloscope and see if the amp is oscillating - I'll bet it is. If it is the V1 circuit, a small cap across R4 might solve it. If it is the V2 circuit, a small cap from the plate to ground might solve it. (Of course if it is oscillating only when there's signal, it could be a little harder to find - a scope with FFT is very handy in such a case.)
     
    DeaconBlues likes this.
  8. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT

    Are the pics showing? I had to delete my PC's cache so photobucket would work again.

    Bias is at about just over 22 watts (82mA @ 287pv)...slightly cold. That is with R20 @ 200ohm/10W

    R19 is on the socket

    I actually own an O'Scope, but I'm a tool & die maker who does this amp stuff as a hobby, I really am a neophyte with the scope. There is no oscillation without signal. The buzz only occurs as gain is increased and notes are played. The amp sounds great until you get past about 1/2 way on the volume knob.

    How small of a cap should I try across R4? I have 5pf to 47uf caps here...

    I have a feeling that it's coupling at the hi/lo power switch and its wiring
     
    Last edited: Apr 26, 2016
  9. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT

    Yes, I did. Thanks Pete!

    Right now I have a 33k R16 grid stopper. I'll try moving it directly on the socket
     
  10. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT
    I just swapped,

    R19 back to 5k6
    R16 to 10k and put it on the socket

    What I don't like is R21 mounted on the board, with a lot of wiring to the Standby/Triode/Pentode switch, rather that just being across pins 3 & 4 right on the power tubes


    [​IMG]

    [​IMG]
     
    Last edited: Apr 26, 2016
  11. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT
    ...and the buzz is still riding the notes
     
  12. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT
    I have been trying to get the 6L6GC to operate at 90% dissipation. Right now I have a 200 ohm cathode resistor, and come close to 90% at idle, but as I add signal via a signal generator (100mV @ 1kHz), the current draw drops by 15 or so mA, and the plate volts go up.

    Now my understanding of a class a power supply is that the tube dissipate a constant output that should not fluctuate with signal.

    Is this a sign that the power supply can not handle a 6L6GC?
     
  13. pdf64

    pdf64 Member

    Messages:
    5,499
    Joined:
    Aug 9, 2008
    Location:
    Staffordshire, UK.
    Oh dear, this is getting to be a tough nut to crack:(

    It may help if you could record and upload (eg youtube) the buzz.

    That's a good point; the screen grid resistor is a grid stopper too, and so is best mounted at the socket terminal. So maybe try that.
    It may change a bit, eg 10%; see some tube info sheets that have suggested operating conditions for class A, eg p2 2nd table of http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6L6GC.pdf
    If it changes a lot there may be several causes, perhaps acting in concert, eg it isn't centre biased, the HT is sagging (especially at the screen grid), bias shift, squish at the cathode.

    With such a low HT (for a guitar amp at least), there's no point in idling the power tube so hot (likely to be hotter than centre bias); see the suggested conditions on the info sheet.
     
    Last edited: Apr 27, 2016
  14. UsableThought

    UsableThought Supporting Member

    Messages:
    1,622
    Joined:
    Jul 18, 2015
    I'm not pdf64 and I don't play him on TV, and in fact I don't know much of anything! But this question strikes a chord with me (ha ha). Looks like you subbed the 6L6 in place of the 6V6GT shown on the schematic. My guess is that's pushing things pretty hard, though I have no idea if it could be relevant to the buzz.

    My tiny bit of relevant background: As a beginner, I am in the process of designing a little single-ended, class A amp, based on a small, high-voltage PT pulled from an old tube tape recorder - i.e. working up a load line, etc. I have the service manual for the tape recorder & it indicates that the PT isn't spec'd to support very much VA - only 50mA of current across ~ 380VDC or so. The original output tube was a miniature w/same specs as a 6V6. Early on I was hoping I could swap upwards to a 6L6 for more power, but there was just no way - too risky to subject a 55-year-old PT to current draw it wasn't built for.

    So I am wondering, do you know the VA for your PT? The VA is just a published number, an estimate by the manufacturer; but it might give a hint as to your question above. Generally the refrain I have heard on this forum (including when I asked about a 6L6 for my PT) is that an amp maker is extremely unlikely to over-spec something like a PT - they watch their costs closely & this amp is made in China where that's doubly so. So if the schematic shows a 6V6GT, it seems unlikely that the installed PT was over-spec'd enough to support a 6L6 instead - at least, support it comfortably. I may be reading the data sheet for a 6L6-GC wrong, but it seems to show a typical 72mA of plate current for Class A operation at 250VDC. That's versus 45mA for the ditto conditions for a 6V6GT. (If I've got those figures wrong, someone please correct me.) I also wonder what effect a straight swap has on the load line and biasing.

    Did you get this buzz out of the box with the stock 6V6? You say it was there "from the get-go."
     
    Last edited: Apr 27, 2016
  15. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT
    The amp is touted as being able to run a 6V6/5881/6L6GC/EL-34, and VHT sells an adapter to run an EL-84

    There are many folks doing so with seemingly great results. The 6L6GC adds head room and brings the power output to about 10 watts. I plug an 8 ohm load (speaker) into the 16 ohm tap to better match the tube to the OT

    Yes, and with substitute 6V6GT's as well
     
  16. UsableThought

    UsableThought Supporting Member

    Messages:
    1,622
    Joined:
    Jul 18, 2015
    Odd that neither the product pages nor the user manual mention the power tube swapping - you'd think they would want to highlight that. http://www.vhtamp.com/images/Manuals/vht-special 6 manual.pdf

    But maybe they don't need to. I find plenty of forum threads on modding & @robrob loves the Ultra version - I wonder if he would have thoughts on the buzz? https://robrobinette.com/Special6UltraMod.htm

    If you really get stumped, a signal tracer is very good for finding where a persistent noise enters the signal chain and so much easier than guessing. A bit easier to use than a scope but you have to put together a probe & audio amp etc.
     
    Last edited: Apr 27, 2016
  17. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT

    A tough nut indeed!


    My goal behind getting the bias up near 90% was to see/hear if crossover distortion was an issue. Playing the amp this morning, I find that moving the R16 (10k now) to the noval socket helped reduce some of the buzz. I have a head cold/allergies raging right now so my hearing is junk.

    When the pull-boost is engaged, the amp gets sloppy sounding after 3 o'clock on the volume. I measures 45vAc going into the power tube's grid with the pull-boost mode wide open! That seems pretty high. It's about 27vAc in the "clean" mode. I can lower that by swapping R22 for a smaller value

    Last night I was playing guitar into the amp while hooked to a dummy load, and watched the sine wave on the scope. There are certain notes where the buzz/parasite is pronounced. While plucking those notes, I could see a second wave form appear and then disappear as the note faded. I have seen the devil, now I must exorcise him!
     
  18. pdf64

    pdf64 Member

    Messages:
    5,499
    Joined:
    Aug 9, 2008
    Location:
    Staffordshire, UK.
    It's single ended, so crossover won't be an issue;)

    Grr, I can't find R22?
     
  19. RussB

    RussB low rent hobbyist Silver Supporting Member

    Messages:
    11,069
    Joined:
    Jul 3, 2004
    Location:
    CT
    R22 (68k) is on the push/pull pot


    I pulled all the triode/pentode/standby switch wiring out today, eliminating R21, and wired it pretty much like a Champ with a choke. No change.

    Rectifier > C42 > choke > C44, C45, R25 to B+1 > 1k R26 > C46 to B+2 > 10k R5 > C7 > V1

    Then I put R20 back to 470 ohm, and stuck the 6V6GT back in it. The amp has less power & headroom, but it did a lot to eliminate (but not fully get rid of) the buzz.

    Me thinks the OT is the source of the buzzing. VHT can't-won't sell me another to try
     
    Last edited: Apr 27, 2016
  20. UsableThought

    UsableThought Supporting Member

    Messages:
    1,622
    Joined:
    Jul 18, 2015
    If you still don't know if it's preamp vs. output stage, including OT, and if you have a power amp w/ a line in, you can temporarily tack a line out into your preamp after the coupling cap C14. I tend to be very careful doing that sort of thing to check & make sure I really am getting just the signal voltage, but that's me. I haven't done this with anything but a naked preamp, so I'm not sure if you'd leave the dummy load in place or just take out the output tube. If taking out the output tube safeguards the speaker I'd do that.

    Or probably easier - if you can see the waveform on the scope as you say, just tack a short wire temporarily into place right after that same coupling cap, so that you can clip a scope probe onto the wire (also taping it or otherwise securing it) - thus leaving your hands free to again play the guitar and look for your ghost note - obviously adjusting the vertical div for signal voltage.

    Or if a tone generator produces the buzz, then just do that and run the scope through the path. But someone other than me will need to tell you where/how it's safe to scope the output tube since I've only scoped preamps to date.
     
    Last edited: Apr 27, 2016
    pdf64 likes 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