Marshall 100w oscillates when a loud note is played

[Blind Lemon]

And also when I probe the phase inverter pin 2, that really sets it off. Change tubes around and shortened up some wires. Has a pretty good hum which I am assuming is a result of whatever the problem is. Seems to be 120hz.

Here are some board shots and voltages.

[Blind Lemon]

More:

[Blind Lemon]

Last 2:

[Raybob]

If this is a new build, try reversing the OT primaries. 

[Colas LeGrippa]

dON'T REVERSE THE ot PRIMARY !

[Colas LeGrippa]

...this is not your problem. I think that a bad cap is responsable for that: check that every stage has it's own filtering cap, that the cap is good, and that the resistors between the filtering caps are good ( not shorted ).
If this doesn't solve your problem, check the grounds. Make sure power tubes cathode and PT center tap  are bolted to the same place in the chassis.


Colas LeGrippa

[shooter]

when I probe the phase inverter pin 2, that really sets it off

By probing, or hard picking
If you can reproduce probing, try a re-tension, re-flow, and re-verify component values.

[Blind Lemon]

New build, swapped OPT primary's first, won't do that again......... what?  I SAID I WONT DO THAT AGAIN. Re flowed all solder joints. Did a cursory look at the components and they look good.


BL

[BetterOffShred]

Hah!  I had them swapped when I fired up my Bassman after I did the blackface mod.. yeah that's the worst sound ever.   

So just a thought, I know a couple of guys were saying that the Silver Mica (the little black M shaped ones) had like a 1/3 failure rate on them.   Maybe if you got a couple of spares try those out real fast.   Just an Idea. 

-Brett

[Blind Lemon]

Nice call. I have had bad ones in the past when I did a Dumble type built. I lifted the leg of one on the brite channel but not the one in the tone stack after the cathode follower.


And just from my experience I have had nowhere near 1/3 failure rate.


BL

[Ed_Chambley]

Also check the cap that is across the Plate wires on the PI.  This is common if one is not installed, well I say common but have seen many amps without them.

[Blind Lemon]

Replaced the Mica Caps different kind and color so as not to come from same batch. No love!

Doug's layout (50w) does not have that cap so I put one on. No love.

I can now probe pin 2 of the PI but not 6 the other plate. Tap it and it sends it into an oscillation, a very loud one. I even changed speaker cabs.

Seems to be very sensitive around the PI. 

BL

[Blind Lemon]

Replaced pre-amp tubes and no love.

Do the PI voltages look right?

I know this is going to be something I over looked. Double check component values and they look right.

When I play through it lightly its not bad, ie: sounds OK. Pluck hard or with the volume up, either channel, it has a very nasty distortion, it wont always go into oscillation but sometimes it does.

Chassis is setting on the bench isolated from the speaker.

BL

[Blind Lemon]

Guess you all are as stuck as I am.

Thanks for the help.

BL

[shooter]

When I play through it lightly its not bad

either channel, it has a very nasty distortion

sounds like you have a gain issue, to much.  Got a scope?

[Ritchie200]

How are you grounding your input shielded wires?  Sorry, could not see in the pics.


Jim

[sluckey]

What is the voltage measurement for V3 pin 2?

[Attila]

I assume you have a EFX loop in that circuit,   try injecting a signal to and from the amp to isolate which part of the circuit is problematic.


Eg.  If you have another amp with a EFX loop  take its OUT and plug into this amps IN to prove the power section,  and the opposite way round to to prove the preamp section.  - this should give some direction.


Some things to try disconnect NFB   and possibly OT primaries polarity... they may be incorrectly color coded.     


 




     

[Blind Lemon]

R200 - At the input jacks,
Shooter - yea but is been a while
Steve - have it written down at home, if I remember right it's about 23v
Attila - Nope - No EFX Loop


BL

[Blind Lemon]

V3 - Pin2 -  24.7V

[sluckey]

Nothing revealing with V3-2 voltage. Now check resistance of pins 2, 3, 7, and 8 with respect to chassis. Pins 3 and 8 should both read 15K. Pins 2 and 7 should both read 1.015M.

Also disconnect the NFB from the OT secondary. I know you've already verified the correct phasing but I'd just like to eliminate the NFB circuit completely. Since the PI tube seems to be a bit touchy I think you should concentrate on that circuit for a while longer.

[Blind Lemon]

3 - 8 = 14.76K
2 - 7 = 1M

NFB will have to wait, I get to pick up my 11 month old granddaughter Eliza from DC. Will check back in a bit.

Thanks Steve.

BL

[guitardude57]

The PI input cap... try loosening the part, heat solder and slacken the part just a touch.
See if the problem goes away.


I have had the leads tight on that cap, and when powered...and tapped with a chopstick,
have a sympathetic ring, that turns into an oscillation at the same tone.
Loosen the part a tiny bit...it goes away.


I always tap around the board after first power up with tubes...looking for part/wire contributed noises.


This can also happen at the cap
entry to V2 as well.


Also check lead dress on nearby offending wires too close or not at 90 degrees.


Peace

[Blind Lemon]

Ok, just to get it down. Lifted the NFB and still had the same problem lots of hum and when I probed pin 2 (or was it 7) it went into wild loud oscillations went to SB to stop the run away flipped back on all good.

Damn hummmmm, pulled first tube still there, pulled V2 and gone. Hum gone now lets see if I can still make it oscillate. Easier to get to the turret that pin 7 is connected to -  same same. Run away oscillation hit power switch.

Next did what the Dude suggested, but lets just replace the PI input cap, same type Mallory 150, different color - different batch.

Went to turn the amp on and nothing, check the mains fuse and its blown, stick another in it blows - pull the B+ fuse and put in a new mains it blows again. Well shit!!!! Out comes the light bulb and it looks like the PT has gone south.

 

I think while I'm waiting on a new PT I'll just pull the board and look at it.

Guys, I have built a lot of amps, Plexis, Rockets, all the big and little Tweeds, BF Fenders, Dumble types, slucky's Revibe (sweet). This thing has kicked my butt and it's not a complex build. 

Bet you a dollar to a donut its something simple that I'm missing. Cost me a PT.

Suggestions always appreciated. 

BL

[Blind Lemon]

Just for grins I checked my power tubes for shorts on my little Sencore Mighty Might, one of them showed "Bad" for grid leakage. Everything seemed to have biased up ok, I had checked the temps on all the power tubes early on and they were all close in temps. ???????

BL

[sluckey]

Your filter caps don't appear to be wired like a typical 100W Marshall. You should have a total of six dual 50 cans. Five of those cans should have the 2 caps strapped together.

The plate node should consist of two dual cans wired parallel/series for a total of 50µF. The PT center tap connects to the junction of these two cans. That looks OK, but you don't need balancing resistors across these two cans. The CT connection does the balancing.

The screen node also consists of two dual cans wired parallel/series for a total of 50µF. These cans have two balancing resistors across the cans.That doesn't look right.

A fifth can should be located underneath the IEC power connector, but you have a cover plate over the mounting hole??? This can has the caps wired parallel for a total of 100µF.

A sixth cap can should be mounted under the board near the preamp end. (Some people use individual caps rather than the sixth can.

It looks like you mostly have a Hoffman board with some modifications for the bridge rectifier. But it's not clear which layout you followed for the cap cans. I'm attaching Hoffman's schematic for the Marshall 100W and also providing a link to a nice layout that may/or not be useful.

     http://sluckeyamps.com/2203.pdf

[Blind Lemon]

As I'm pulling the board I notice I have 56K grid resistors not 5.6K.

BL

[Blind Lemon]

Filters re done.

The pot is the bias circuit. I've used this setup on quite a few amps, never had a problem.

BL

[Blind Lemon]

Its been a bit, I've been out of town and putting the amp back together.

I have the Marshall layout now for all the PS smoothing caps. Found a joint that was not soldered along the PS dropping circuit. Replaced the one bad (?) power tube and put in a pre-PI master. Also changed to more of a Marshall layout for the bias circuit. Changed a little bit of the way this amp was grounded and it is dead quiet. I changed no values of the components, I did change a couple of the chick-let style caps to Mallory 150s.

This amp sounds good at normal playing volume (MV @ ~1/2 and normal/brite volume 1/3), but, with ear plugs in, I crank the MV up and strum the open strings it still goes into a nasty osculation/distortion that is louder than hell and the only way to stop it is to hit the SB switch.

Load is 1 8ohm speaker with the impedance selector on 8, if I go to 16ohm it seems to delay the nastiness. Also seems to delay the onset of nastiness if I roll the Presence all the way to the left.

Thoughts?

BL

[Blind Lemon]

Youtube Vid.

https://www.youtube.com/watch?v=2yzGo69SLjs&edit=vd

Notice ear plugs.

[Blind Lemon]

Man tough crowd. Not even a "your video sucks"


BL

[sluckey]

Your video sucks!

Wish I had an idea but I don't. Could be almost anything. I'd make it start that sound then run a scope forward from the input to try to get an idea where the noise first shows up. Look at the B+ nodes with the scope also. May be tough though. It could be that once it begins, it shows up everywhere in the amp.

[Blind Lemon]

1.Checked components
2.Re-flowed solder joints
3.Pulled board completely out and checked it
4.Re-installed PS to mirror Marshalls
5.Have swapped out most caps on the board.

 

Do you think it could be the OPT?

My o-scope is crap.

I got it oscillating and I left it that way so I could move some wires around and it started to smoke the screen grid resistors. So I put some true 5W cement block resistors in it and put my DVMs on the screen side of each one and got the thing screaming and it started dropping voltage and pulling big current again. I don't think I could troubleshoot it screaming like that without melting the tubes down.

Don't know how to troubleshoot this thing at this point without just shotgunning parts at it.

This sucks ...............................

[sluckey]

Pull the preamp tubes and rig a jack so you can connect the guitar directly to the PI input cap. Can you make it oscillate?

[Blind Lemon]

I can play it all day long at low volume.

At high volume,,,,, MV almost all the way up and volume at 10 o'clock is when it starts. Light strumming-good, Medium strumming you might hear some distortion (got ear plugs in) bare down on it and it runs away.

I say that to say I can try your PI hookup but I don't think the signal will be large enough to make it happen.

What do you think? 

All this and NASTEP going on too.

[Blind Lemon]

As requested, not enough drive. Could not make it start screaming even with hot humbuckers.

Didn't take long, used the aux. speaker jack as input. Ugly but worked.

[Blind Lemon]

Got a New Fender Twin OPT, won't be able to for a couple of days, but I think I am going to slave it in.


If that doesn't work would you replace the power tubes first  (friggin quad of EL34) or the power supply caps first?


At least that's were I think I am heading.



BL

[Willabe]

I'd replace the tubes 1st before replacing filter caps or OT. (Because their already new caps and new OT.)

[shooter]

Mighta missed it;
what is the bias without tubes, any fluctuation?
with tubes?
what happens to the bias when it starts making bad noises?

[Blind Lemon]

shooter - ur talkin bias voltage? Steady without and with, of coarse current goes up when strumming and way up when screaming.


willabe - I got the caps and transformer now, got to order new power tubes, these power tubes were new also.


Thx guys, someone's is going to hit on something.


BL

[SILVERGUN]

Have you pulled a power tube off of each side yet to test them in pairs?
 
 

[jojokeo]

Good call SG, I am going with power tubes. There's only so much that can go wrong and through process of elimination iron, caps, and resistors don't work then not work then work again on their own like this.

[Ritchie200]

Exactly. Mix and match to see if you can ID the possible offending tube. Also early on you mentioned you relplaced the preamp tubes. Did this include the PI?


Jim

[silverfox]

At this point I would suggest taking another amp and cross wiring it. In other words, the other amps preamp goes to the amp under question, power amp, the preamp from the amp under question goes to the power amp of the known good amp. That will indicate a preamp or power amp problem in the amp under question or it will indicate a systemic problem in the amp under question if the problem fails to appear when the two sections are isolated.

I'm going to guess there is a short at the higher power. Check the rectifier and the power caps. Can't remember from all the posts: Have the filter caps been swapped out? That video sounds like an arc.

silverfox.

[Blind Lemon]

Nice thought on testing in pairs. Will try that next. - silver gun


Got a new quad of EL34s coming.


I have a Rocket that I may be able to slave in to it but I'm betting a dollar to a donut it's a PI and out problem.


All preamp tubes have been swapped multiple times. PI included.


Already swapped out the reciter diodes, have not swapped out the JJ 50/50 caps.


Thank you Gentlemen.


BL

[silverfox]

I just had another thought. what are the value of your screen resistors? sluckey mentioned something about the screen supply and now I'm curious if there isn't some sort of interaction between the screens and G1 at higher power levels- and that could be why the screen resistors smoked. sure they were underated? But why didn't they smoke in the beginning.

silverfox.

[Blind Lemon]

Just played the amp with 2 power tubes in PP and it was fine, with all 4 being played (in pairs) to make sure it wasn't a tube. It's not a tube, they all perform perfect at high volume.

Read silverfoxe's post about screen resistors and wattage. Per the schematic they should have been good,,,,,,,,,except for one thing,,,,,,, I'm only using one screen resistor per pair. Used a 50w layout.  Look at my board in the first pics of this thread. Should be using a screen resistor for each OPT.

Give me a few minutes to do a little rewire and test.

BL

[Blind Lemon]

 

[sluckey]

Did you replace those 56K grid stoppers too?

[Blind Lemon]

Yep, I'm cleaning this mess up. Seems that took care of it.

I'll post more a little later.

BL