Nightmare On TSL Street

[plexi50]

I have a TSL 100 Triple Super Lead amp that sounds horrible with crackle,low volume and bad note decay
The tubes are biased at 80mv and exhibit a solid and steady bias at idle
When i make a chord or any note the two center tubes start to redplate
As soon as i mute the guitar signal (tone) the redplating disspears immediatley and the bias is back at 80mv

I know that the bias jumps a little higher on all amps when playing but this is abynormal (Frankenstein)
Bias coupling and grid resistors all measure out within thier correct values

Bias coupling caps have not been check yet. I say one of the bias capacitors or coupling caps is bad
If so why would it show a steady bias at idle though

What do you think? TSL 2000 =   

[darryl]

These links may ( or may not . . . ) be relevant to your problem:

http://www.lynx.bc.ca/~jc/TSL122.html

http://www.hullerum.de/Marshall/TSL122repair.html

[plexi50]

Thanks darryl \. Yes this is very helpful info and exactly my problem. I have worked on a lot of these over the past 3 years but none of them have acted as badly as this amp is. Learn something new everyday. Thanks

[plexi50]

Im working on this amp again this morning and i have noticed that the volume is slowly decreasing every time i turn the amp on after doing more checking and poking around in it. I dont know why the two center tubes are redplating and not all of them. Every bias and coupling resistor checks out good. I have been thinking of one other idea that could be the problem

Question: If the output transformer was bad or going bad, would it affect the bias to a redplate degree?
At idle the bias is steady at 80mv. The trimmer pots and board are fine as are the solder points on it
Could a bad OT make the tubes redplate and run away say 3 seconds after i make a note tone or chord?

[phsyconoodler]

Ok,a TSL has the adjustable bias at the rear of the chassis.I bias each pair to 70ma,as 80ma is way too hot.
  The original marshall tubes were likely Svetlana winged C's or EI and they seem to be able to take hot bias for a while but after a year of pounding they expire.
  Also i don't rely on the bias voltage at that test point and always double check with a bias meter.I'm willing to bet those redplating tubes are damaged.
  Pop in a new set and bias for 70ma per pair and see what happens.Never had an issue with a TSL after doing that,assuming nothing else is amiss.

[plexi50]

I used a new set of tubes because the ones that were in it had redplated so badly that the JJ lettering is barley visable
C46 plate cap to ground had shorted and blown apart. The old tubes looked like they were coming apart inside as well
C46 leads me to think the OT may be shot because of the very low volume
That still does not explain the redplating of only the center 2 power tubes
Im loosing it on this one. I have worked on plenty of them and know what to replace on these amps as far as the bias goes
Bias voltage supply diodes/10K bias feed and the bias coupling and grid resistors / Bias trimmer board all resoldered

It just sounds to me like the OT at this point. The bias is rock solid until i play a note and then the tubes start to redline
Stop playing and bias is idling back at 80mv on both sides. I'll adjust for 70 mv from here on. I got the 80mv setting online

Im just thinking along the lines of the OT being shorted internally and possably creating the redplate condition which would blow the tubes if i kept playing the amp  

All power tube pins have been resoldered and C46 area dremeled clean


 

[phsyconoodler]

I've had a few that ate C46 too but they all worked after the repair.Watch the bias voltage on those two tubes,if it's stable it may be the cathode grounds for those two tubes.No ground=instant redplating.

[plexi50]

Cathode resistors are fine and pin 8 is showing ground on all 4 tubes. Me too on seeing many C46 problems but never a OT problem yet. I was looking real close at the pic i took of C46. Does it not look like a glob of solder on what is left of the cap terminal of C46?

I know i think too hard sometimes but this is my possable interpitation of that glob of solder  

Some one read on the internet about removing C46
They sniped the cap off and reconnected both cap leads to the board thinking a connection needed to be complete across the C46 mounting holes. I can picture the cap disenegrating in a puff of smoke. But the board itself where the C46 leads are was chared all the way to the other side. I dremeled it out

Thats what i think anyway. Then again who knows for sure
The HT fuse looks stock and the cathode resistors are fine

[phsyconoodler]

Cathode resistors?You mean screen resistors?

 You still didn't mention what pin 5 is doing when those tubes redplate.
I always replace that cap at C46.

[plexi50]

Ok. Amp biased at 80mv per side. Guitar jack plugged into input. All EL34 tubes show 39mv per tube on pin 5 of each tube
Volume is wide open on the clean channel. When i touch the tip of the guitar cord to my finger i get a rise of voltage to 54mv on two tubes both on the same side and next to each other. There is a nasty distorted buzz

When i touch the tip of the guitar cord to my finger i get a rise of voltage to 41mv on the two tubes (opposite) both on the same side and next to each other. There is no audio signal to be heard. This pair of tubes does not redplate. I have switched the tube pairs around with the same results and pin 5 test positions

Yes i ment all screens are good. The cathodes of the power tubes are grounded
Man this is interesting! I still hate these amps

[punkykatt]

Plexi50, is it possible without too much trouble to perform a turns ratio test on the output transformer?  Discharge filter caps. You will need to disconnect the OPT ctr tap from the B+.  Pull all the power tubes, find a couple of nails or something like it (same size as tube pins) to install in pin 3 tube sockets inners or both outers. Put a 1/4" jack plug in speaker jack, Attach DMM set for highest AC voltage to pins 3 of tube sockets, apply small AC voltage 10v to jack plug terminals and see what you get for readings. Be carefull.   I forget off hand what the formula is for doing turns ratio. If you dont have it I can look it up for you.
Good luck  Punky

[plexi50]

Plexi50, is it possible without too much trouble to perform a turns ratio test on the output transformer?  Discharge filter caps. You will need to disconnect the OPT ctr tap from the B+.  Pull all the power tubes, find a couple of nails or something like it (same size as tube pins) to install in pin 3 tube sockets inners or both outers. Put a 1/4" jack plug in speaker jack, Attach DMM set for highest AC voltage to pins 3 of tube sockets, apply small AC voltage 10v to jack plug terminals and see what you get for readings. Be carefull.   I forget off hand what the formula is for doing turns ratio. If you dont have it I can look it up for you.
Good luck  Punky

Punky i never learned how to measure an output transformer the right way. I have measured the CT and primary leads to see if the ot windings were reading ohms fairly close in relation to the center tap. I quess i have been flying blindly all these years. Most of my bad OT finding have been nothing more than an educated quess. But i have not been wrong so far. It's sort of like if all else is in spec and working then all that is left is the OT. Even so if this turns out to be the case it would be the first time that i have seen a pair of tubes redplate on one side upon making a note and the tubes returning to normal idle bias upon muting that note

Some bias swing is normal when you are playing an amp but nothing like what  have going on here with this amp

[punkykatt]

Did you try setting the bias way cold and see what happen? Are you checking the currant draw with a bias probe like Doug sell or are you going by the 1 ohmers in the amp ?

[plexi50]

I  set the bias by adjusting the trimmers on the back panel in mv
I was measuring pin 5 directly of each power tube using my Fluke DVM in mv
No i dont have a bias probe right now. I need to make or buy another one

[sluckey]

I was measuring pin 5 directly of each power tube using my Fluke DVM in mv

Pin 5 ain't mv. It's negative volts. Most likely about -40V.

[plexi50]

I was measuring pin 5 directly of each power tube using my Fluke DVM in mv

Pin 5 ain't mv. It's negative volts. Most likely about -40V.

Yes pin 5 was measured in negative volts. You got me*

[plexi50]

I am working on this amp again. The bias is stable at idle. What would cause a new set of tubes to redplate as a note runs out?
Mute the signal and the bias is back at a steady -77mv per side. The bias rises to near -250mv+ while a note runs out
This is a crazy amp and it's making me crazier  

[punkykatt]

Whats happening to the plate voltage when that happens?  Did you try setting the bias way cold?

[plexi50]

Whats happening to the plate voltage when that happens?  Did you try setting the bias way cold?

I will be working on this thing tommorrow. I will let you know what the plate voltage does as the bias runs away. No i have not set the bias any colder that 70mv per side

The thing that gets me is i have worked on more than 15 of them and this one is psycho

Click on Monkey to get the psychotic connection. Click again to regain sanity

[LooseChange]

The pc board is bad and it's conducting voltage from the plate supply. Big problem with these. Marshall got a bad batch. You will need to lift the bias rail off the board and connect directly to the socket. This fixes the problem most of the time. The real solution is to replace the board.

[Willabe]

Hey Plexi, you got that monkeys phone #? I want to start a band with him.


                   Brad     

[plexi50]

Thanks LC. Yeah i have been reading all about this particular batch of boards that Marshall never fessed up to being bad.
The Monkeys phone number is 1-800-BANNANA

[Willabe]

The Monkeys phone number is 1-800-BANANAA

Thanks.       

[plexi50]

My suspicions were right about the output transformer. DOA. I put top chassis mounted OT from an old peavey 100 watt Heritage that i had laying around and it is screaming again. I heat shrinked off the ultralinear taps that it had. Bias does not run away like it did before. So this supports my theory that a shorted OT can make the bias rise extremley high on a note and then at idle be normal again. The only bummer about the OT that i used is that it only has a 2 & 4 ohm tap. I settled on the 4  ohm of course

Surprised i didnt blow at least one or more of the power tubes that were in it before

Now to conquer the bias drift problem by isolating the power tube grid pins from the board.  Wiring the grid resistors now
(soon to be isolated from the board like in last the pic below) on all 4 El34 tubes

Does it seem like were going backwards in some amp manufacturers products? Crazy man

[plexi50]

Im finished using a dremel to isolate the grid pins from the board and the resistors are all wired up
Tommorrow is another day

[LooseChange]

Really nice work and very time consuming. Hope you make a lot on this repair!

[plexi50]

The customer gave it to me after i explained that the board was contaminated and needed Heart,Brain and Output surgey. Believe me i would not get into doing this for anyone. This is too time consuming and to be honest makes me feel like im working on a toy from Walmart. At this point it is a conquer and see what happens kind of thing. I just finished 8 amps in the past 2 weeks so i have a little time until the next round comes in

[jerrydyer]

Loosechange nailed it, my buddy came up with a mod to lift the bias rail away from the b+ there they are way to close and cause probs in those amps
richard at
lingenberg sound .com
or richard@lsound.com   might be able to help

[dpm309]

I worked on a DSL about a year ago that had a hole blown through the board next to one of the power tubes.  There is a lot of discussion on the Marshall forum about these boards.  The DSL and TSL use the same main board.  I was able to buy a replacement for a little over $200 through Antique Electronics (special order) and the owner has not had any problems since.  He was experiencing the bias drift problem before it blew and decided to replace the main board which saved a lot of labor cost.  Apparently the original boards had some issues due to the dual sided PCB boards shorting or becoming conductive.  I have attached some pictures of the board.

Dan

[plexi50]

Thanks jerrydyer. I may try and get in touch with him just to see what he has confirmed. Id like to see a pic of what he has done.
I have it finished and stable. Bias is steady at 70.1 with a little drift up and down but nothing like before where it would increase within minutes from 70mv to 81mv. So far the amp has been on for an hour and it is still at 70.1-.4mv on both sets of El34 tubes

I notice a big change when i replace the bias supply cap (C42) and it got even more steady

Amp sounds great but will leave it on for 3 hours to let it get a good hot workout

[plexi50]

phsyconoodler what is the function of the plate/screen C46 cap? Oscillation?

[plexi50]

I also raised the PI plate resistors off the board. I blew cold air on them and the voltage dropped.  If this amp had a fan i think it would further help the bias drift keeping the compoments cool. What a PITA amp. Still running steady at 70 mv per side using the 3 prong bias pins on the trimmer board. I found this link to an outfit in the UK that sells the newer TSL boards for $96.00 US

http://www.hotroxuk.com/marshall-tsl100-main-board.html

[Willabe]

What a PITA amp. I found this link to an outfit in the UK that sells the newer TSL boards for $96.00 US

I wonder why Marshall is selling replactment PCB for this amp?     


                        Brad     

[plexi50]

I had to lift the bias rail off the board feeding both sets of EL34 tubes and wire them to the coupling caps to get a solid and stable bias that does not drift at all. I was going to order the new board and i thought i would give it one more try. Running 2 hours so far and stable with no drift AT ALL FINALLY! at 70mv per side. Im surprised at how fast the bias rises now and stops stable @ 70mv per side. Will take pics of what i did later tonight after i shut this thing down

[jerrydyer]

your work looks really good there.

[plexi50]

Thanks  jerrydyer. It would have been better had i not gotten carried away with the nail polish at the bias caps

[Ryteone]

Thanks for all the posts.  I just got a 60w TSL that just stopped working.  I had worked on it for a customer three years ago.  I changed out the tubes, noticed the screen resistors and the caps right off them looked overheated.  I swapped all those out and biased the new tubes and sent the amp on its way.  Well two weeks ago he was practicing and the amp just stopped no sound.  Well when I opened up the amp I couldn't find the two screen resistors, they had unsoldered themselves.  I was reading all over the internet about the problems with these amps.  I didn't just want to swap out the parts again and have it fail even in a worse way.


This one hasn't gotten as bad as yours but was just going to follow what you didn't (as long as the customer wants to spend the money on the time its going to take).


Scott...

[HotBluePlates]

... when I opened up the amp I couldn't find the two screen resistors, they had unsoldered themselves. ...

Oops!!