Help me read my o-scope traces

Discussion in 'Amp Tech Center' started by timfred, May 11, 2019.

  1. timfred

    timfred TDPRI Member

    Posts:
    47
    Joined:
    Mar 14, 2016
    Location:
    Northern CA
    Do you see anything in these traces that would explain a harsh fizzy distortion?

    Amp is an Ampeg Reverberocket R12-RT (schematic is here.)

    Working well except starting around 60% volume the distortion has a harsh fizz. The pre-amp sections are clean all the way to 100% volume, and distortion starts in the 12ax7 floating paraphase PI.

    The PI distortion is strange in 2 ways. 1. When the PI starts distorting, the first triode clips asymmetrically then the second triode clips the other lobe, so the output tubes are getting different signals once the PI starts distorting. 2. There is some other non-clipping distortion in the waveform.

    Trace 1: 100% volume. The smooth trace is the input to the first PI triode (from a 0.2V ptp 500 Hz sine at jack) and the other trace is the input to the second power tube (from the second PI triode):
    [​IMG]

    Trace 2: Same signal but the distorted trace is at the speaker.
    [​IMG]

    Scales are different, obviously. The input to the PI is 1V/div (5V ptp amplitude). PI Cathode voltage is ~2V so I expect the PI to be clipping here. PI output is 10V/div.
     
  2. Greg_L

    Greg_L Tele-Holic

    Posts:
    781
    Joined:
    Jun 27, 2018
    Location:
    Texas
    I'm definitely no expert, but in my limited experience, I'd guess that's some sort of oscillation.

    I had a scope trace sort of similar to that, but to a lesser extent. I had little hiccups in my smooth sine wave after the first preamp gain stages. I had an oscillation problem due to poor lead dress. I chopsticked some wires, found the culprit (reverb driver plate wire on a Fender build), and fixed the issue.

    You could possibly be having the same issue somewhere maybe? VERY CAREFULLY start moving some wires around with a chopstick or some other non-conductive probing device and see if the kinks change in the scope signal.
     
    timfred likes this.
  3. kbold

    kbold Tele-Meister Silver Supporter

    Posts:
    237
    Joined:
    Jul 21, 2015
    Location:
    Australia
    Try changing the phase inverter. That would at least confirm or rule it out as the cause.
    I'm not familiar with that family of power tubes (the 7868 power tube), but I'm not a fan of 12AX7's as inverters. Maybe try a 12AT7. While 12AT7's have a lower gain, they have higher output current capacity than 12AX7's. The slight mismatch between 12AX7's and 12AT7's is not usually a problem.
     
    brokenbones and timfred like this.
  4. timfred

    timfred TDPRI Member

    Posts:
    47
    Joined:
    Mar 14, 2016
    Location:
    Northern CA
    Very possible. I poked around a little but I probably need to spend more time at this.

    I did try a 12AT7 as well as a 12AU7 and any difference was not really noticeable.

    Now that I look at the speaker trace, I'm thinking that maybe the unbalanced PI is causing crossover distortion.
     
  5. dogmeat

    dogmeat Tele-Holic

    Age:
    68
    Posts:
    908
    Joined:
    Oct 12, 2017
    Location:
    Alaska
    negative feedback loop ok?
     
  6. brokenbones

    brokenbones Tele-Meister

    Posts:
    431
    Joined:
    Mar 15, 2012
    Location:
    Indy
    Like @kbold said. Check against another PI tube. If it it's still present, double check the components all around V4. Dress and ground connections included. It may be a leaky cap.
     
    kbold likes this.
  7. kbold

    kbold Tele-Meister Silver Supporter

    Posts:
    237
    Joined:
    Jul 21, 2015
    Location:
    Australia
    Second that.
    I always start with a visual inspection, especially of capacitors, solder joints, and any signs of thermal damage.
    Cap's are probably #1 on the list of problem components, especially as gear ages.
    (More than half of the faults found before grabbing for a multimeter or CRO).
     
  8. printer2

    printer2 Poster Extraordinaire

    Posts:
    6,245
    Joined:
    May 24, 2010
    Location:
    Canada
    Put a bigger grid stopper on the output tubes, start with 10k.
     
  9. robrob

    robrob Poster Extraordinaire Ad Free Member

    Posts:
    7,468
    Joined:
    Dec 29, 2012
    Location:
    United States
    A leaky phase inverter input coupling cap can throw off the bias and cause severe asymmetric clipping.
     
    LudwigvonBirk and Wally like this.
  10. timfred

    timfred TDPRI Member

    Posts:
    47
    Joined:
    Mar 14, 2016
    Location:
    Northern CA
    Solved!

    Thanks for all the input, the solution was a little bit of everything.

    For reasons I can't explain, I thought I would try removing the bypass capacitor on the PI. When I did this, I saw a nasty ringing/oscillation on the post-PI signal, even for low volumes. I think the PI bypass cap was hiding it.

    Then I tried removing the NFB. This erased the ringing, but the harsh distortion was still there. So I decided to go step by step. I removed the power tubes and concentrated on getting the PI as balanced as possible.

    Re-biasing the PI by dropping the shared cathode resistor to 470R made the clipping the same on both triodes. but the magnitudes were still different.

    Sure enough, the 470K resistor in the voltage divider feeding the PI was more like 500K. Replacing it brought the two sides of the PI close enough for rock and roll.

    I had already reduced the 0.1uF coupling capacitors to 0.047uF, but increasing the grid stoppers to 10K eliminated the ringing/oscillation.

    After verifying that it was running cleanly (or cleanly distorted?) I reconnected the PI bypass cap and NFB. Honestly I can't tell much difference with/without the NFB, but the amp sounds great either way and much louder, as if these issues were robbing

    It sounds really good now, and much louder than before. Thanks all!
     
    robrob, dogmeat and Greg_L like this.
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.


  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.