Marshall 2203 bias balance help sought

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by Neil, Mar 10, 2004.


  1. Neil

    Neil Member

    Messages:
    321
    Joined:
    Feb 11, 2004
    Location:
    San Francisco Bay area
    I was checking the bias on my 1988 2203 head tonight after fitting a set of new GT EL34Rs. I was using a biasright type tool. On the first tube I got a reading of 25 ma. Way low so I set it to 38ma. I then measured the third tube and got a reading of 45ma. A bit high from what I have read but I was more surprised about the difference between the 2. I went back and forth a few times and left it at 35ma and 42ma. I have biased 2 power tube amps before and never seen more than 1ma or so difference between the pair. This is my first 4 power tube amp.

    What could be causing this? The tubes are a proper matched GT quartet so they should be ok. Could something in the amp have drifted? What effect would a difference in bias like this have on the sound? I did not try the other 2 tubes - didnt seem much point if there is a problem with the amp.

    Can anyone help?
    Thanks
    Neil
     
  2. John Phillips

    John Phillips Member

    Messages:
    13,080
    Joined:
    Mar 17, 2002
    Location:
    Scotland
    Definitely try the other two tubes - that should show up whether the problem is with the tubes or the amp.

    It's entirely possible that one of the tubes is out of spec - even if they were matched correctly at GT. Knocks and vibration in shipping can mess up tubes. The tube characteristics are critically dependent on the precise physical positioning of the grids, which are made of very fine wire. Hit the tube hard, and they flex under their own mass. Hopefully, they return to where they were... but not always.

    If the problem is with the amp, you almost certainly have a very tiny DC leak in one of the coupling caps from the phase inverter. This will reduce the bias voltage on the side with the bad cap, and so increase the idle current. Measure the bias voltage on both sides with no power tubes in (and the amp fully on). If you've got a leak, change the caps - best to do both of them. It could be a failing bias feed resistor or grid-stopper resistor too, but less likely. If so, measuring their values (with the amp off) should show it up.
     
  3. Swarty

    Swarty Member

    Messages:
    1,143
    Joined:
    Dec 16, 2003
    Location:
    Michigan
    Try swapping positions of the 2 tubes and see if the current reading stays at the socket or moves w/ the tube. If it moves w/ the tube, it is a bum tube issue. If it stays at the socket, it is an amp issue. I'd bet it is a tube issue.
     
  4. Neil

    Neil Member

    Messages:
    321
    Joined:
    Feb 11, 2004
    Location:
    San Francisco Bay area
    Thanks for the input.

    It didnt occur to me that a matched set of tubes would give different bias results. I went back and switched tube positions around to try and determine if the tubes were the cause or the amp was the cause.

    The results. Between the 4 tubes in the 4 positions I noticed a maximum difference of 15ma between the highest and lowest bias current readings. The typical difference was 10ma. The highest reading was consistently achieved with one particular tube, the lowest with another particular tube. The other 2 tubes fell inbetween these readings. This confirms that the tubes are responsible for the problem.

    I did notice that moving a tube to another location and then back to its original position did not necessarily give the same reading. It could be ~1+ma different. Not sure what causes this (amp could be drifting with time and or temperature, measurement error).

    I thought matched tubes were matched for bias current draw along with transconductance etc etc. If so then these Groove Tubes are suspect. I understand that tubes are delicate and that a knock could change the bias but this seems extreme and such a susceptibility would render them useless for and amp that is to be moved around in a car or truck.

    I dont know if this is range of differences in a tube is acceptable. I suspect it isnt and I think I just bought my last Groove Tube product.
     
  5. John Phillips

    John Phillips Member

    Messages:
    13,080
    Joined:
    Mar 17, 2002
    Location:
    Scotland
    Out of interest, what type - manufacturer, or at least country of origin - are these tubes?

    It does sound like more than one is bad - I would have expected three to be very close, with one 'rogue tube' which could be down to shipping damage.

    The errors in returning a tube to the same position could be caused by time and temperature. You really need to let the tubes settle for several minutes before measuring current. If you did, that's more evidence that the tubes aren't really right.

    In any case +/- 1mA in 30-40 isn't important - that would be considered pretty well-matched, being under 5% - but 10-15mA is definitely not close enough.

    Have you tried contacting Groove Tubes?
     
  6. Neil

    Neil Member

    Messages:
    321
    Joined:
    Feb 11, 2004
    Location:
    San Francisco Bay area
    Hi John,

    The tubes are the Groove Tube EL34R (I think R stands for Russian - I dont know which Russian company though).

    I did let the tubes settle before taking the reading and waited until the current measurement appeared to have settled before taking the reading. I probably could have got more reproduceable results by leaving it longer.

    I havent spoken to GT yet thats a good idea.

    Cheers
    Neil
     
  7. Swarty

    Swarty Member

    Messages:
    1,143
    Joined:
    Dec 16, 2003
    Location:
    Michigan
    They should definitely be sent back or taken back to the store for exchange. You pay a huge premium for the GT name which is supposed to include tight matching (not sure what they guarantee, but it almost certainly less than 5mA). FWIW, I like really like the Slovokian E34L (GT/JJ/Tesla, insert brand) which is supposed to be a ruggedized version of the EL34.
     
  8. Neil

    Neil Member

    Messages:
    321
    Joined:
    Feb 11, 2004
    Location:
    San Francisco Bay area
    Here is the upshot.

    I took the tubes back to the store and it was clear from looking at the boxes they came in that they were not a proper quartet. One box (of 2) was from a quartet and the other box was a duet. Somehow they had become taped together (as GT does for its Quartets) and sold to me. Impossible to say if this was a store problem or a GT problem. If I had been awake when I bought them I would have realised the problem and saved all this hassle.

    The store replaced them with a (real!) quartet of GT E34LS's which are now fitted and sound amazing. I shouldn't really try and compare results from 2 (probably wrongly biased) duets with a quartet but the E34LS sounds much better than the EL34R.

    Thanks for the help.
    Neil
     
  9. Carl Zwengel

    Carl Zwengel Guest

    Although I'd like to see a tightly matched quad, your situation isn't all that bad really. This is just going to take a little finesse. What you need to do is pair up 1 & 2 then 3 & 4 such that each side has the same current draw and make sure that no single tube is exceeding its rated power dissipation.

    For example, lets say you've got one tube pulling 20mA, two pulling 30mA, and the 4th pulling 25mA. You'd pair up a 20 with a 30 on one side and the 25 with the other 20 on the other side. THis will net you 50mA on the left bank, and 45mA on the right. Then you just set your bias to the hot tube.

    In your case, just label each tube 1-4 and it's respective current draw. Then match up the pairs so you get the closest left to right current balance.

    Hopefully I didn't confuse you.
     

Share This Page