Plexi Keeps Eating Up Screen Resistors

Discussion in 'The Workbench' started by SirArcsAlot, Dec 31, 2016.

  1. SirArcsAlot

    SirArcsAlot Active Member

    Joined:
    May 22, 2014
    Messages:
    399
    Likes Received:
    121
    Happy New Year, everyone!

    I've got a small problem with my 1959 plexi build. I have gone through multiple pairs of screen resistors in this amp, including up to 7w 1k ones and still they burn up. I've tried the obvious and put in new tubes. I even tried switching from El34s to KT88s yet still to no avail.

    The bias circuit appears to be working correctly, as I measure 35ma across each tube with 490 B+, and the rest of my voltages are correct.

    EDIT: I should also note that I am using 1.5k grid resistors on the power tubes.

    The only thing I can think of that isn't quite normal is my wall voltage. I measured it at 125v. My transformer's primary is rated for 120v. I don't recall having any troubles using my variac set to 120v, but from a practical stance I'd like to be able to run the amp without it (especially at venues).


    Also, here's the datasheet for the resistors that burned up. They were both wirewound and flame resistant.

    http://www.digikey.com/product-detail/en/vishay-sfernice/RWM06221001JA15E1/RWMB-1.0KCT-ND/1587958
     
  2. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    14,931
    Likes Received:
    4,878
    Location:
    US of A
    I know most 1959 schemes are typical but it may help if you post or link the exact schematic or layout that was used.
     
  3. matttornado

    matttornado Active Member

    Joined:
    Mar 14, 2012
    Messages:
    478
    Likes Received:
    95
    I have the same issue with my 1974 Superlead. Same B+ voltage too. Everything seems to be working fine though.
     
  4. SirArcsAlot

    SirArcsAlot Active Member

    Joined:
    May 22, 2014
    Messages:
    399
    Likes Received:
    121
  5. Jazz P Bass

    Jazz P Bass Active Member

    Joined:
    May 19, 2015
    Messages:
    228
    Likes Received:
    27
    Location:
    Wernersville, PA
    5 watt should be sufficient.

    By chance are you running through some kind of 'power soak'?
     
  6. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    14,931
    Likes Received:
    4,878
    Location:
    US of A
    I agree, 5 watts is enough.

    The layout diagram seems fine.

    I believe you have been having this amplifier for a while too.

    How about some good pictures in the chassis?
     
  7. matttornado

    matttornado Active Member

    Joined:
    Mar 14, 2012
    Messages:
    478
    Likes Received:
    95
    What would a power soak have to do with it? I use a power break.
     
  8. ampmadscientist

    ampmadscientist Well-Known Member

    Joined:
    Jun 26, 2014
    Messages:
    3,896
    Likes Received:
    1,276
    Location:
    Discharge is futile.
    The Output transformer may have secondary taps mixed up.
    Possible wrong impedance as load connected to amplifier.
    If the load does not match, that explains why screen resistors burn up- the screen is drawing too much power.

    Should be a 5 watt resistor, not 7, the resistor is a FUSE

    If bias is being measured incorrectly - that also explains it.

    HOW
    are you measuring bias...exactly?

    PICTURES- lets see the whole insides of the amp...

    No, not 125 VAC causing this...some kind of mis-match is causing this.
     
  9. m1989jmp

    m1989jmp Active Member

    Joined:
    Apr 24, 2014
    Messages:
    530
    Likes Received:
    74
    When they do blow up, are they blowing up in pairs?

    Always the same pair?
     
  10. ampmadscientist

    ampmadscientist Well-Known Member

    Joined:
    Jun 26, 2014
    Messages:
    3,896
    Likes Received:
    1,276
    Location:
    Discharge is futile.
    Power Soak is infamous for creating mis-matched loads...which burns up amplifier.
    You do not want to use a Power Soak....ever. It's junk.

    Power Brake is the professional method.

    A. See where pin 1+8 is connected to the chassis?
    Put a 1 ohm 5% or 1% resistor between 1+8 and the chassis.
    (connect 1+8 together, then connect that to the chassis thru the 1 ohm resistor)
    Connect the amp to the correct MATCHING load.
    Now measure in mv, across the 1 ohm resistor.
    32 mv across the resistor = 32 ma of cathode current.

    B. Connect a speaker cable to the speaker cabinet.
    Measure with ohm meter at the end of the speaker cable.
    WHAT is the reading (in ohms) from the speaker load?
    This is very important - don't assume that the load is correct until you have actually measured it.

    C. Meters: always use a 2nd meter to VERIFY your measurement. Meters do not always read correctly. Don't assume that the reading is correct until you verify it w/ a second meter.

    Stop assuming - start verifying.
    Then, you will discover where the problem actually IS.
     
    chuckharmonjr likes this.
  11. neikeel

    neikeel Well-Known Member

    Joined:
    Feb 12, 2014
    Messages:
    1,389
    Likes Received:
    434
    For your screens to draw that much current you must be ploughing through output tubes!
    As said before 5w should be more than enough (I use 3w on my 50watters).
    What is your measured screen voltages?
    I agree some pics of your build would be good.
     
  12. Guitar-Rocker

    Guitar-Rocker Well-Known Member

    Joined:
    Mar 5, 2012
    Messages:
    2,433
    Likes Received:
    1,925
    Location:
    McCordsville, Indiana
    Your screen voltage should ride somewhere between 375V to about a maximum of 400V (assuming you are using EL34 valves). If you are riding at the 400V line or above, when idling and when you push the amp, that instant inrush of current is killing your valves and screens.

    I think a JJ EL34 lists the max screen at 450V, with plates much higher (maybe up to 800V?) and a 6CA7 screen is max at 425V? If your plate is at 490 as you stated, what is your screens at?

    This is assuming all other factors are correct as previously mentioned by other members here in the forum.
     
    Last edited: Jan 1, 2017
  13. SirArcsAlot

    SirArcsAlot Active Member

    Joined:
    May 22, 2014
    Messages:
    399
    Likes Received:
    121
    I've already got 1 ohm resistors off the cathodes and have always biased the amp to 70% dissipation on the tube that draws the highest current. My screens are usually 1 volt below B+, as I have never seen screens in any amp (either one that I've built or own) be several volts below B+ unless we are talking ultralinear. Oddly enough I haven't taken out a single tube, as I had the winged C Svetlana EL34s in there which can take some punishment. Also as I said, I was using KT88s to try to produce the same results. Measured my load, and I even verified the OT taps were on the correct part of the selector.

    Also I am not using a power soak, just straight into a speaker load.

    to answer m1989jmp's question yes, it's always one pair of them that get crispy.

    I'll see about getting some pictues
     
    Last edited: Jan 1, 2017
  14. SirArcsAlot

    SirArcsAlot Active Member

    Joined:
    May 22, 2014
    Messages:
    399
    Likes Received:
    121

    So if that's the case (which is what I suspect), would putting in higher wattage screen resistors fix the problem, or am I going to have to figure something else out?
     
  15. mickeydg5

    mickeydg5 Well-Known Member VIP Member

    Joined:
    Sep 30, 2011
    Messages:
    14,931
    Likes Received:
    4,878
    Location:
    US of A
    Every tube amplifier I have has screen voltages above 450.
     
  16. SirArcsAlot

    SirArcsAlot Active Member

    Joined:
    May 22, 2014
    Messages:
    399
    Likes Received:
    121
    So what makes even less sense here is the fact that even with KT88s the resistors are overheating. According to EL34 datahsheets they draw 50ma of screen current. KT88 datasheets specify 18ma of screen current. All of my voltages check out to be in spec.
     
  17. RickyLee

    RickyLee Well-Known Member VIP Member

    Joined:
    May 22, 2010
    Messages:
    7,537
    Likes Received:
    1,543
    Location:
    Riverside CA U.S.A.
    So when you say the screens are frying in pairs, is that on one side of the push pull? It is either the V4 V5 or the V6 V7 pairing?

    I am wondering if it is something to do with the signal getting through too hot on one side of your push pull. Have you tried measuring the audio signal coming into each side of the push pull on your pins 5 control grids while the amp is being played? And then also measure the bias on each leg of the output transformer while under load.

    This only happens when you get some good volume on the amp correct?

    How about the phase inverter circuit? If you do measure any odd out of balance audio signal coming into the EL34's, check out your phase inverter circuit thoroughly.
     
  18. m1989jmp

    m1989jmp Active Member

    Joined:
    Apr 24, 2014
    Messages:
    530
    Likes Received:
    74
    If everything else checks out, measure DC resistances of both OTs primary windings

    Other than that, I'd increase screen resistors to 1.5k 5W, grid resistors to 5.6k and bring the wall voltage to 120VAC; these extra 5VAC could raise your plate voltage by 15-20VDC

    I see your bias is on the safe side, but try lowering it a bit more just for testing purposes.
     
    Last edited: Jan 2, 2017
  19. ampmadscientist

    ampmadscientist Well-Known Member

    Joined:
    Jun 26, 2014
    Messages:
    3,896
    Likes Received:
    1,276
    Location:
    Discharge is futile.
    It makes perfect sense, when you consider the amp is not connected to the correct matching load....or there is a connection problem to the load...

    Always amazed when people have these type problems...they still just ignore the basics.
    They won't measure the speaker cabinet (just like this guy).
    They won't check to see that the correct tap is connected to the speaker...

    No wonder the screen resistors burn up. The load is wrong.
    No, you do not need 10W screen resistors.
    What you need to do is VERIFY your load and connections.

    So stop guessing, and start measuring.
    This is a "classic" mismatch problem.

    When you connect a 8 ohm tap to a 4 ohm load, the current DOUBLES.
    Then, you are biasing the amp, based on the wrong load...no wonder it's all goofy.
     
  20. ampmadscientist

    ampmadscientist Well-Known Member

    Joined:
    Jun 26, 2014
    Messages:
    3,896
    Likes Received:
    1,276
    Location:
    Discharge is futile.
    OK this is how you do it:

    Start at the output and work your way backwards.

    1. What is the reading on the ohm meter when you measure the speaker?

    2. What color tap wire of the output transformer is hooked up to the speaker?
    Can you post the transformer spec sheet? Or a link?
     

Share This Page