Marshall 100w oscillates when a loud note is played

[Blind Lemon]

Well guys the mystery has been solved.

Thanks to Silverfox's questions, it got me thinking about the wattage of the screen grid resistors and how I had used those exact resistors in that exact place on the power tubes before. But, there was twice the load on those resistors because I was only using 1 resistor for a pair of screens. Then it hit me, the amp would behave itself with just 2 power tubes because those screen weren't tied together. (Thanks Silvergun) I needed 4 screen grid resistors for 4 tubes, one for each tube.

Bingo!!!! I had used the Hoffamn layout for the 50w, it works but you need to add a few more parts for a 100w.

Maybe someone smarter than me can confirm or explain this but i think that the problem was an interaction between the tube pairs on the same screen grid resistor. The 2 screens were tied directly together after the SG resistor. I'm thinking that with each tube having its own SG resistor the they are isolated from one another, no problem.

What have I learned? If your going to deviate from a proven design, think it through and be ready to solve problems. I have built enough amps that I expect to have to trouble shoot a new completion, if I don't I am pleasantly surprised.

Also when I'm stumped and looking for help, to listen and think about every question posed. There were quite a few people trying to help and all with good suggestions.

This is a great community of knowledgeable amp techs, I use to hang out here quite a bit in the early days, but life happens and you drift. Hopefully i can hang around more and maybe help someone.

I want to thank you for the helped, this amp was kicking my butt. But as with so may other things, with hind sight it was really an easy problem to see.  Right????

BL

[sluckey]

This time the problem wasn't right in front of you!   

[jojokeo]

(Thanks Silvergun) I needed 4 screen grid resistors for 4 tubes, one for each tube.

The 2 screens were tied directly together after the SG resistor. I'm thinking that with each tube having its own SG resistor

So know we know... SG = silvergun Screen Grid   


To be honest I had no idea about how many "silvergun" resistors you were using or their values & wattages? 

[HotBluePlates]

... I was only using 1 resistor for a pair of screens. ... the amp would behave itself with just 2 power tubes ... I needed 4 screen grid resistors for 4 tubes, one for each tube.

... i think that the problem was an interaction between the tube pairs on the same screen grid resistor. The 2 screens were tied directly together after the SG resistor. ...

Were the new resistors mounted on the board, or right at the sockets?

[Blind Lemon]

No room on the board, mounted them on the tube sockets.


BL

[Ritchie200]

(Thanks Silvergun) I needed 4 screen grid resistors for 4 tubes, one for each tube.

The 2 screens were tied directly together after the SG resistor. I'm thinking that with each tube having its own SG resistor

So know we know... SG = silvergun Screen Grid   


To be honest I had no idea about how many "silvergun" resistors you were using or their values & wattages? 

I don't know about you, but I feel a resistance every time SG posts to bash my boy....

Jim

[Ritchie200]

Seriously this is all great stuff to put in our back pockets for the next project or problem that crops up on the forum or at home.  Unfortunately BL had to suffer through it to find the fix.

Now we need sound clips with this baby dimed and rockin!!!!

Jim

[HotBluePlates]

No room on the board, mounted them on the tube sockets.

There's a good chance then the resistors acted as "screen grid stopper" resistors.

Supposedly, "high slope" (high transconductance) tubes are more prone to oscillation due to long screen grid wires.  EL34's certainly fall in the category.  And the screen grid acts as a low-mu input terminal.  Could be simple bad luck with long screen grid wires too close to plate/OT wiring, and just the right inductance/capacitance/phase relationships to oscillate.

It may be the best plan to always install screen resistors right at the tube socket (even if there is board space for them).

[jojokeo]

It may be the best plan to always install screen resistors right at the tube socket (even if there is board space for them).

Doesn't everybody?! 

[Ed_Chambley]

It may be the best plan to always install screen resistors right at the tube socket (even if there is board space for them).

Doesn't everybody?! 

Not everybody every time.  Would you like to know when I don't and why,  or are you good? How about a discussion?  I would bet you already know.

[jojokeo]

It may be the best plan to always install screen resistors right at the tube socket (even if there is board space for them).

Doesn't everybody?! 

Not everybody every time.  Would you like to know when I don't and why,  or are you good? How about a discussion?  I would bet you already know.

Yes please Ed. I was sort of kidding but do practice what we preach here.

[sluckey]

I mount screen resistors on octal sockets. But I also have built a lot of EL84 amps. These tubes don't have free pins to use as tie points so I put screen resistors on the boards for those.

Hoffman put screen resistors on the board for all his amps except the recent Plexi 6V6.

[jojokeo]

I mount screen resistors on octal sockets. But I also have built a lot of EL84 amps. These tubes don't have free pins to use as tie points so I put screen resistors on the boards for those.

Understood, however it can be done using a single resistor such as on this organ donor build. The layout and proximity made it easy and a no brainer.

[sluckey]

It's really a personal choice. Placement of screen resistors is not all that critical. The whole reason we are even discussing screen resistors is not because of improper placement of these resistors but because two of the resistors were omitted.

[HotBluePlates]

... The whole reason we are even discussing screen resistors is not because of improper placement of these resistors but because two of the resistors were omitted.

True.

But coming from that angle, the worst you'd expect is that double-screen-current went through a single resistor, causing a little more compression.  But there was oscillation...

Once you're looking at a 100w build, the area taken up by the output tubes is bigger, and the wiring to those sockets gets longer.  Long "grid wires" in general would be a risk of oscillation (granted, usually for G1 where the tube is much more sensitive).

[davebolden44]

a general curiosity from an intermediate amp builder, would the lack of the 2 screen resistors directly result in the catastrophic PT failure mentioned earlier?

[HotBluePlates]

... would the lack of the 2 screen resistors directly result in the catastrophic PT failure mentioned earlier?

If the screen resistors contributed to failure of any transformer, it's only because an oscillation at full amp-power burned something up.  In other words, indirectly if at all.

There are many tube amps with no screen resistors at all, and nothing burns up.  Screen protection for guitar amps is a special case.

[Blind Lemon]

Ok been out for a few days and am kind of surprised this didn't die. More traffic now than when I was having the problem.


Any way good discussion.


Now there is something I omitted, just forgot to post. The PT did not open, in my moment of anger and haste I miss diagnosist the problem. I had forgot to post that information, it made me happy for about two seconds and I went on.


HBP - my reasoning that the power tube pairs were interacting causing the oscillation was because of the lack of oscillatiion with only 1 pair installed, in either the inside or outside positions. No other changes were made so the wire length was the same. But, that being said I couldn't say one way or the other for sure.


I wish I would have taken a pic of the amp but I was building it for someone else. I had talked him out of it 3 times over the last 5 years, but he got his wish on the fourth time he argued with me. He is suppose to bring it back in a week or two so I can go over it again with fresh eyes and a better attitude.   I'll get a pic then of how I did the SG resistors.


BL