Help with a squeal, Please.

[bcmatt]

Hi,
I built an 18 Watt T-Rex and I am in Revelstoke with my Old Man finishing up the combo cabinet for it today.

However, there seems to be one major problem and I'm wondering if the minds here can help me.
I posted pictures in the other thread about Point-to-Point Wrecks, but this is really something that should be moved out here because this isn't actually a Trainwreck and it is more about trouble-shooting specifically.
Here are the pics:
http://ampgarage.com/forum/viewtopic.php?t=7193

Anyways, if I have the VVR turned all the way off (or almost off), I get crazy squeals on the T-Rex Channel anywhere above 2 on the volume.

The NFB switch does seem to play a part in this squealing game and controlling where on the volume the squeal can be achieved at.

Any thoughts would be great. Is this what is called "oscillation problems"?

[Structo]

I am not a Trainwreck expert by no means but I would think it has something to do with the lead dress of that amp.

By all accounts the Trainwreck is an amp on the verge of instability so any deviation from the layout shown in the files section would probably cause problems such as out of control oscillations.

Looks like you have shielded cable on the input to the grid on V1 correct?

And your grid resistor is right on pin 2?

Other than lead dress or a wiring error I can't help.

Maybe a problem with the VVR?

Where is the VVR, I don't see the board there?

[bcmatt]

Yes, I have shielded cable from the inputs to the grids on V1, as well as from the volume pot to the the grid of V2.

I bought a VVR board for this, but then when my turretboad material got lost in the mail and I was forced to use just terminal strips for the build because of time left till Christmas, I decided that a simple terminal strip VVR would be more appropriate looking for this build. I'll use the VVR board for my upcoming Liverpool instead. Ya, so the VVR knob is in the place of the former Power switch on the very left, and it uses the yellow wires. The VVR components are all mounted on a terminal strip that is right where the Buss bar is grounded.

I think the biggest deviation I have from the original T-Rex Layout (besides not using turret-board), is that I moved the NFB switch to the front panel. Does having a long line from 8 ohm of the impedance selector to the switch make a difference?

[Phil_S]

I think the biggest deviation I have from the original T-Rex Layout (besides not using turret-board), is that I moved the NFB switch to the front panel. Does having a long line from 8 ohm of the impedance selector to the switch make a difference?

I'm unsure, but the problem I have not been able to solve in my t-wreck is the squeal I get when I turn up the presence pot. I have not had the patience to got back through it, rewire and new cap on the pot. I suspect the lead dress on it isn't satisfactory.

[Markusv]

Try reversing the OT Primary leads.

[bcmatt]

Try reversing the OT Primary leads.

That was the first thing I did back when I first finished the amp and turned it on.
The squealing now is just when the 1 channel is allowed a certain amount of signal.

[bcmatt]

I'm unsure, but the problem I have not been able to solve in my t-wreck is the squeal I get when I turn up the presence pot. I have not had the patience to got back through it, rewire and new cap on the pot. I suspect the lead dress on it isn't satisfactory.

Sounds like maybe we have related problems

[Bob-I]

The squealing now is just when the 1 channel is allowed a certain amount of signal.

Sounds like oscillation for sure. I'd start by getting out a chopstick and moving wires around until you hear the oscillation change, then you're most likely in the right area. Typical areas for oscillation are signal wires, like the plate load connections and grids. Also suspect is the NFB leads.

Good luck.

[bcmatt]

The squealing now is just when the 1 channel is allowed a certain amount of signal.

Sounds like oscillation for sure. I'd start by getting out a chopstick and moving wires around until you hear the oscillation change, then you're most likely in the right area. Typical areas for oscillation are signal wires, like the plate load connections and grids. Also suspect is the NFB leads.

Good luck.

Ok, thanks. I'll be chopsticking again in the morning. So, NFB leads can cause oscillations? If that is the case, I should consider moving the NFB back to the rear of the chassis again, and make the front panel hole just a second input jack. I'm trying to make it look pretty, but the faceplate says MV there, so either way it will be wrong.

[keithrick]

bcmat,

My wreck will squeal if it is out of the head box. Have you tried the amp with the back on? Shield and cover that back area and see if that helps.

Good luck and Happy New Year!

[bcmatt]

bcmat,

My wreck will squeal if it is out of the head box. Have you tried the amp with the back on? Shield and cover that back area and see if that helps.

Good luck and Happy New Year!

Yup, I've done that. It didn't help though, unfortunately. Good idea though. Definitely worth it to try and easy addition.
Thanks.
Happy New Year

[Bob-I]

Ok, thanks. I'll be chopsticking again in the morning. So, NFB leads can cause oscillations? If that is the case, I should consider moving the NFB back to the rear of the chassis again, and make the front panel hole just a second input jack. I'm trying to make it look pretty, but the faceplate says MV there, so either way it will be wrong.

Yes, NFB can oscillate. Before I start moving things I might just connect it dirrectly from the speaker jack to the PI and see what happens.

I see some leads between V1 and V2 that look fairly close together and almost parallel, plus some of the leads in that area look quite long. I'm a fan of color coding leads so I know which is cathode, plate and grid. I can't tell with this pic because I can't see tube socket but typically plate and cathode can run parallel but grid cannot.

Another thing to check is put a scope on the various stages and look for the oscillation. I typically appears as a squiggle on top of a sine wave input, but a few times I've seen it all by itself as a HF wave. Since yours is audible I'll bet you could see it easily.

[bcmatt]

Yes, NFB can oscillate. Before I start moving things I might just connect it dirrectly from the speaker jack to the PI and see what happens.

I see some leads between V1 and V2 that look fairly close together and almost parallel, plus some of the leads in that area look quite long. I'm a fan of color coding leads so I know which is cathode, plate and grid. I can't tell with this pic because I can't see tube socket but typically plate and cathode can run parallel but grid cannot.

Another thing to check is put a scope on the various stages and look for the oscillation. I typically appears as a squiggle on top of a sine wave input, but a few times I've seen it all by itself as a HF wave. Since yours is audible I'll bet you could see it easily.

I believe that my problems are mostly around V2. It is exclusively gain-stages for the T-Rex Channel. The typical 18 watt normal channel is completely problem-free and does not have anything to do with V2.

Thanks for info about which wires can run parallel or not. That's the kind of knowledge I really need to hear. I think from now on I will color code my wires like you mentioned. I will also avoid mounting components to the tube sockets like this. I will likely only use eyelet boards from now on too.

I was going to move the NFB switch to the rear anyways, and it did help significantly with the squealing. Flip the switch one way and there is squealing, the other way is ok. There is still significant hum on the T-Rex channel (around V2 I suspect). I would sort of like to move the components away from the socket and onto terminal strips, but I am afraid there is a real lack of real-estate to work with.

I don't have a scope, so that is out. As far as hooking up the speaker jack to the PI, I am unsure what you mean by this and how and why.

I appreciate the help. I better go back to work. I am going to install a second T-Rex input to fill in that faceplate hole. It will likely look better than the NFB switch anyways. Although, the faceplate says MV there. An input jack marked MV and with numbers is probably slightly less confusing than a mini-toggle-switch marked as MV (that switches between 2 and .

[Zippy]

I was going to move the NFB switch to the rear anyways, and it did help significantly with the squealing. Flip the switch one way and there is squealing, the other way is ok.

So what is your negative feedback switch doing in each position? It could be that the amp needs the requisite amount of feedback for stability. This amp is already unstable enough. Why would you need less negative feedback?

That kind of mod in this kind of amp is looking for trouble. Sounds like it found it.

If you need to tinker with the response of the amp, consider the fast/gradual switch. That'll give you something to put in that hole in the chassis.

[Bob-I]

I was going to move the NFB switch to the rear anyways, and it did help significantly with the squealing. Flip the switch one way and there is squealing, the other way is ok. There is still significant hum on the T-Rex channel (around V2 I suspect). I would sort of like to move the components away from the socket and onto terminal strips, but I am afraid there is a real lack of real-estate to work with.

I don't have a scope, so that is out. As far as hooking up the speaker jack to the PI, I am unsure what you mean by this and how and why.

I meant to hook the NFB direct, skip the switch. In fact try running the amp without any NFB. If there's no squeel then reverse the plate leads on the output tubes.