Tube Amps / Music Electronics
|For current discussions, please visit Music Electronics Forum.||The sunn still shines online!|
|3/21/2006 4:51 PM|
||Are you techs experiencing the same problem w/ JCM2000s?|
It's not my first (or second) JCM2000 100W head (DSL or TSL) that eats power tubes/blow PTs.
Even (very) conservatively bias adjusted (less than factory spec) , after a heating period, they just don't sustain bias voltage; current starts to rise dangerously...this is with chassis on the bench, without cabinet, tubes upside. If you put it upside down (for more proper readings inside the chassis), things get worse more quickly. A (big) fan really helps to stabilize things, although not a solution in real world.
Tried everything: redid bias section, changed output coupling caps, nothing helps (besides every attempt destroy some pcboard trace on the crappiest doubleface board I ever saw !)
Apparently is heat related, no tubes or 2 tubes do fine; which is no help either...I'm thinking about subbing the EL34s for 6L6s, at least is 2.4A less on filament heat...
Anyone from an authorized Marshall repair center out there with a tech bulletin about it ? (it must be a design flaw - or I'm the baddest luck guy this side of hemisphere...)
Also any bright idea gladly accepted,
|And now, a word from our sponsors:
|3/21/2006 6:12 PM|
Find out WHY this is happening. AMps should be on and running for a good 20 minutes before adjusting bias, but if it runs away no matter what, then why? Does the bias voltage at each grid start to drop or does it remain the same? Both sides or just one. If bias voltage drops, then of course current will rise. If bias voltage drops at all tubes, does the bias raw supply drop or does it remian stable? C42 in the DSL100. If supply is stable but bias V drops at tubes, we either have a couple bad bias trim pots of the caps from the PI plates are leaking. Did you replace the trimmers?
Does screen current start to rise or stay the same? Does B+ stabilize or does it change? DO both phases of the output stage climb or is it predominantly just one side that climbs? Scope the output, is there any high freq oscillation building? Does it matter what impedance the load is? You are using a load during test I hope.
I am unaware of any service bulletins on something like this, and have not seen the problem. Any TSL/DSL amps I have had in for service had other issues than this. The thing we need to do is find the point of failure, then we can determine why the failure occurs.
The one problem I see way more than I would have thought is that stupid little 22pf disc cap over by the end power tube - between screen and plate. They short, making the end tube a triode and upsetting things. Check that socket for a pin 3-4 short. Almost invisible when you are in a hurry, it is C46, down next to the lower right power tube on the DSL100 drawing.
|3/21/2006 8:16 PM|
||Re: Are you techs experiencing the same problem w/ JCM20|
Thank you for your reply Enzo;
Actually amp is on for several hours now. After warming up, I biased it pretty cold - about 58mV per side (not really 29mA per tube, i guess those 1ohm cathode resistors are not very good), after get it real hot reachs close to 90mV per side. (despite the values, it's a big flutuation, and obviously situation would be worse with chassis inside the cabinet)
I thought the bias tap could being affected with overall tranny heat, but cooling the PT directly with a fan makes no effect, but pointing the fan to the tubes lowers the current in a few minutes.
Cooled, grid voltage is -42V , bias tap 44VAC, ; raw bias voltage is -55.8V.
Heat on, grid voltage drops to -39V (it does happen on both sides), bias tap still on 44VAC; raw bias - 54V.
Vplate and screen seems unaffected (of course, a few volts difference with lower/higher currents) and no difference between output taps.
Scoped output looks ok, no oscillation or uneven clipping.
I did not replaced the bias pots.
I could not manage to find C46 on the board, but V8 has a 10V difference between plate and screen, so i guess cap is ok.
Actually, with the tubes upside down I could shunt the tranny, the 88mV per side reading translated into 25mA per tube, which is okay (even being on the cold side); to be honest, after 5 or 6 hours on, heating/current seemed to stabilized/reach max point; what is pissing me off is that the user needs to deal with a real crappy sound (very overbiased) for about an hour before amp gets hot enough to bring the current up to the desired level, which cannot be adjusted at first, without catastrophic failure past few hours...
|3/21/2006 9:01 PM|
I have noticed a good number of JCM2000 series amps that have the bias adjustment pots hooked up backwards. Aparently the connector is reversed on the PC board which makes the Pot adjust the set of tubes opposite its placement on the chassis. This can make adjusting the bias a bit tricky as you will see some adjustment as you turn the pot but the other pair of tubes will be going wild. Another common thing in these amps is poor soldering around the power tubes in particular and in the amps with a DIN footswitch connector - there is no support for the connector which puts all the strain on the connector and the pc board mounted components. I have used a large fender washer and a 1/2" or 3/4" punch (I can't remember - it fit the radius of the connector perfectly) to make a bracket to hold this DIN connector securely to chassis. Actually one washer can make two brackets as I cut it in half. I used a couple 4-40 machine screws & Keps nuts to secure it in place. Marshall has added a bracket to the newer amps but there are a lot of the older ones out there.
|3/22/2006 2:04 AM|
I don't recall seeing one where the two controls were NOT at opposite ends from the tubes they controlled. When i adjust them, I keep that in mind if I think there is a bad tube. If the screws are on centerline, one could possibly turn the little trimmer board upside down. Unless there are clearance issues. if sits in a very small space betwen the main board and the wall.
Not all in the 2000 series have the little cap. You didn't specify, so since you mentioned DSL and TSL, I included it. it only affects the one tube on the end on the amps that have it.
I never use JCM numbers for that reason, no one knows what amp you are talking about - DSL201? TSL100? Customers never use the model number thay always call it a JCM2000.
|3/22/2006 5:21 AM|
I also didn't saw one with non-reversed bias trimmers...and no, this one don't have a DIN connector.
As I stated in the first post, it's a 100W head, this one a DSL - I have a TSL100 head (bias problem also) waiting a PT re-wind and space on the bench...
After some hours of sleep, I would upgrade C42 to 100V; and still need to make a burn-in with amp on the cab, to make sure things don't get worse as they do with bare chassis on the bench.
To be honest, these things are completely unreliable, especially cause the two are from a rental company, not bedroom player...
Thanks guys for the comments,
|3/22/2006 2:10 PM|
The early versions of the DSL/TSL had wopping big 220k ohm grid resistors. I always yank them out and replace them with 10k. This change helps.
Enzo is right the little cap between the anode and screen of the very end tube always goes faulty and I also replace this cap on spec with one rated at 1KV.
Best of luck.
|Page 1 of 2||Next>||Last Page>>|