New lead dress, new caps, same old problem!

[Shad]

and I will read through your post more tomorrow

That would be great! I did a lot mor probing and picture taking. However, when I posted the pics I named them incorrect. I know which is whic, but will take few moments to correct. Too tired right now, very very late.

Sadly I did not get all the way through the amp before the Marshall Power Brake on loan went up in a puff of smoke!!
WHat the heck is that about. Could it be I have weird oscillation that fried it, or just simply blew it out?

I powered off within 2-3 seconds tops. Can that short of a time without a load (I can only assume when the smoke came the load was on the way out) toast my OT?

Do you have your O'scope set on AC or DC coupling?

AC

I will post the pics and notes (I took good notes on all settings this time) tomorrow asap.






[/code]

[csbradley]

I powered off within 2-3 seconds tops. Can that short of a time without a load (I can only assume when the smoke came the load was on the way out) toast my OT?

You're OT will be fine.

the Marshall Power Brake on loan went up in a puff of smoke!!

What power was the Power Brake? When your running your amp into clipping, you are putting out significantly more energy than a same amplitude sine wave. The crest factor of the clipped wave is close to 1. A sine wave (typically used for power measurements) has a crest factor of approximatly the square root of 2. So for a load for your amp, you should at least oversize the load by this factor of 2 (square root of 2, squared, [since this is power]) if you are going to crank your amp. Of course, you would think Marshall and other attenuators would factor this into their ratings, because thats what they were designed for.

However, now I am able to observe that there is still a rasp on the notes. I think this is the core problem. I can hear it, and see it on my scope as sharp spikes that occur with the note. I think when hit really hard there is enough of them that is blends into a buzz like sound that is harder to ID, but is there kinda like a snare spring or something. Rasp or buzz seems to fit the description best.

That sounds like an oscillation is occuring. The first three places I would expect is lead dress, grounding, or the power supply.

Does this only occur with lower notes? Just a note, the lower notes have a much longer rise and fall time. This means that the signal is drawing current from the power supply for a longer period of time before the capacitors are able to recharge. Might be helpful information, might not.

Well, I got to get back to work. I would still look at the wires from your OT. It looks like they are crossing over all your input wiring to your power tubes. It's just a guess though.

Back to work. More later...

CSB

[Shad]

That sounds like an oscillation is occuring. The first three places I would expect is lead dress, grounding, or the power supply.

I will explore all three. Quick question, by grounding do you mean I should try a different grounding approach, or that I should check my grounds for resistance? All grounds check out OK but I don't dispute it could be that, or that I need to re-think the grounding approach in general.
How can Igo about testing my power supply? I read some Jack Darr stuff that explains how to probe the power lines for parasitics....I'll check inot that again and perhaps post a pic or two.

Does this only occur with lower notes?

I'll have to check.

I would still look at the wires from your OT. It looks like they are crossing over all your input wiring to your power tubes. It's just a guess though.
Will do for sure!

Could you peek at some shots of my scope just to hopefully rule some things out for sure, and to help me learn a little about what I am looking at?

Here are some scope shots from last night with some notes. Hopefully this can tell more. I will try the other suggestions you've made but really want to get your eyes on these pics.

Here is my V1 output:


Here is V2 Input (Pin 2) right at the socket. If I scope just before the 470K grid R's then I do not get this deformation. Is this normal?


V2 Output (pin8):


Then *poof* goes the power brake!

I am going to put the amp inot my resistive load. I have heard these will sometimes mask an oscillation, but I know what that output rasp looks like. If it is still there with that load I will post a pic of it. I was intending on working my way to the output but I think that shot needs to be posted along with these. I will post the power line pics too if they looks like the pics in Darr's book.

[Shad]

PS the power brake was a PB100....It's leaking something now. Maybe it blew a big cap......project for another day!

[rjgtr]

If the PB is leaking open it and remove the cap and clean it now. Don't wait for it to become a gooey mess....

[Shad]

[/quote]
If the PB is leaking open it and remove the cap and clean it now. Don't wait for it to become a gooey mess

Ick! I hadn't thought of that....I'll post these new pics and voltage readings and then I'll open it up and see what fried and clean up the muck......

[csbradley]

by grounding do you mean I should try a different grounding approach, or that I should check my grounds for resistance? All grounds check out OK but I don't dispute it could be that, or that I need to re-think the grounding approach in general.
How can Igo about testing my power supply?

By grounding and power supply, I didn't mean anything wrong with them, per say, but that those are good places for noises and currents from other areas of the amp to couple into other parts of the circuit where their not supposed to be.

Lead dress can cause signals to couple via the air from one point in the circuit to another. Currents in your power supply/ground can be modulated over any impedance and couple into other areas of the circuit that share the power supply.

All of your O'scope plots look normal to me to far. From what your describing as far as the sound, I sorta expect the signal to show up as some high frequency crap riding on the tops of each clipped sine wave. Then as the signal started to die down, it will look as if the oscillations will be inside a larger sine wave, like the sine wave is being modulated by the oscillation frequency. Thats when you'll hear that buzz. Like when you play a note and you hear the snare rattle in the room. Does that make any since. I think you used the snare referance. Very good analogy.

CSB

[veets]

Maybe I haven't noticed, or maybe you haven't said, but does this "farty" "raspy" noise only occur when the amp chassis is sitting on the speaker cabinet? Or does it happen if the amp is not physically attached to the cabinet? You have also mentioned that the noise is at it's worst on low frequencies.
When I built my 18, I had some terrible raspy rough farty noises too- but only if the amp was turned up past about half, and it was worse when sitting on the cab. Lower frequencies made it worse too (biga$$ power chords) It turned out to be a bad filter cap in the power supply. Changed it out and the amp was fine. The cap tested OK too, but changing it cured the problem. It was the vibrations that made it short out intermittantly, I guess. But I was able to locate the fault by tapping on the case of the capacitor with a chopstick---have you tried that?
BTW, I'm surely no "guru"--but a fresh set of eyes doesn't hurt.
I hope you get the problem sorted out soon-they sound incredible when they are workin' right!!

[Shad]

Here are my output shots.

@100hZ:


@1khZ:


Unfortunately, I cannot see the spikes with my resistive load!
However, that kink is blocking distortion isn't it?
ur12 pointed this out in another post, and just want to verify that is what I am seeing.
Can that actually be caused by the lead dress? I will be experimenting with that as promised after a couple of other observations.

Ok and this observation is bugging me, tell me it is normal so I can go eliminate it from my thinking, please.

That V2 P2 scope picture. It is normal for my positive half to be clipped like that, and the negative half be pushed farther?
On this same pin, if I have the gain at zero I get .895-VDC, and if cranked I get a ful 1.5-VDC!
Isn't this what is pulling my negative half down, and chopping the top?
Sorry if I am obssesing over nothing.

Going to look at my power lines now.

[Shad]

By grounding and power supply, I didn't mean anything wrong with them, per say, but that those are good places for noises and currents from other areas of the amp to couple into other parts of the circuit where their not supposed to be.

I see. I do have some rather long ground leads to the busbar. This was because of an article I had read by Aiken, not to blame him though, I am sure I did not fully understand the article (mental note to read again).
I will try to shorten those up, and fool with them while fooling with the lead dress.

Lead dress can cause signals to couple via the air from one point in the circuit to another. Currents in your power supply/ground can be modulated over any impedance and couple into other areas of the circuit that share the power supply.

So ground leads have as much potential to need careful dressing as well?
I guess I should be looking at all my grounds to the pots as potential for problems as well?

All of your O'scope plots look normal to me to far.

Rats, even V2P2? That bugger with it's funky voltage has got my panties bunched!

I suck. I had followed the amp from the back to front the first time, but didn't take notes or pics. When I went in again I went from front to back.
The output waveform into the reactive load I think woudl tell a lot.

From what your describing as far as the sound, I sorta expect the signal to show up as some high frequency crap riding on the tops of each clipped sine wave. Then as the signal started to die down, it will look as if the oscillations will be inside a larger sine wave, like the sine wave is being modulated by the oscillation frequency. Thats when you'll hear that buzz.

I'll cop to it, I can't really visualize what your describing, but I am willing to bet that is what I was seeing yesterday into the reactive load.
I will get my hands on another one asap (not a PB100 though!) so I can get a shot of that. It was very weird, sorta normal waveform, but then these spikes would occur, and then they were doing this weird phasing thing, appearing on this side, then on the other....actually now that I think about it I think what you describe is exactly that!

Like when you play a note and you hear the snare rattle in the room. Does that make any since. I think you used the snare referance. Very good analogy.

Exactly, don't know where they put the lever to lift the springs off my amp but yeah the distortion snare is tight!!

Gotta laugh

Maybe I haven't noticed, or maybe you haven't said, but does this "farty" "raspy" noise only occur when the amp chassis is sitting on the speaker cabinet?

No, I keep the chassis well enough off ot the side of the cab, and also place a large copper shiled over the chassis.
I suppose I could move it onto the cabinet and see if it gets worse. It's quite loud and the chassis is vibrating quite a bit just in the room.
I'll put it on the list to try. I suppose that would really bring it out if mechanical.

It turned out to be a bad filter cap in the power supply. Changed it out and the amp was fine. The cap tested OK too, but changing it cured the problem.

I replaced the capcan, the onboard filter caps, and also the cathode caps.
JJ capcans, sprague board filters, some unknown brand of cathode caps. All test ok with my cap meter, though it isn't a real fancy one with ESR or whatever... I 'think" I am ok on the caps, but could be wrong.

I was able to locate the fault by tapping on the case of the capacitor with a chopstick---have you tried that?

Yes many times. Also tried moving tubes in the sockets to see if any noises there. None to be found. Naturally I have swapped tubes one at a time, and even tried a new quad of power tubes yesterday, just to rule out the possibility that I fried two, or more, at a time somehow.

BTW, I'm surely no "guru"--but a fresh set of eyes doesn't hurt.

No worries. I'm asking for help from anyone that has any to lend. Fresh eyes is exactly what I need. It is hard to get perspective with ny nose right in there all the time. Myself also being so noob doesn't help me much either, though I am trying to learn all I can.

I hope you get the problem sorted out soon-they sound incredible when they are workin' right!!

I have an 18 a friend built. Sounds amazing! Guy can't solder for squat (it was here because a main lead came loose!), lead dress is very scary, there is little rhyme or reason to it, few rules obeyed, sounds friggin' amazing! .......................................Pisses me right off!
Seriously though it is because his sounds so damn good that I keep at this. I know when I get it right I will be really really happy with the sound.

[csbradley]

Rats, even V2P2? That bugger with it's funky voltage has got my panties bunched!

Looks normal to me. Others may want to comment. When your signal is flat, you are probably clipping your V2 output. Pushing it into saturation. During this time, current will flow through the grid (P2) and distort the signal that you are seeing. You should be able to turn the volume down until you no longer see clipping on the Plate (P1), and the signal on the grid (P2) should look normal.

Probably anything you see other than high frequency crap riding on the tops of your signal will be normal.

very weird, sorta normal waveform, but then these spikes would occur, and then they were doing this weird phasing thing, appearing on this side, then on the other....

Yea, I'm not really sure how to describe it so it makes sense. But I think we're thinking about the same thing.

Can you give us a shot of the whole board and how your running your ground bus or star ground?

Are you running shielded wires for your grids. Its worth a try. If it get better, then it's your grid leads picking up signals radiated by other parts of the circuit.

CSB

[csbradley]

Ok, a small grounding ramble. Put it in a different post so I don't add confusion. Just me rambling....

I think I may have different views on grounding than a lot of people, but remember, all of the current that is flowing through your tubes and signal lines is flowing from your B+ to ground. Where you have large currents and any impedance, its likely to radiate and be picked up elsewhere to cause oscillations.

Star grounding is good for audio frequencies, but it is rather poor for high frequencies. If you have high frequency oscillations, star grounding is not the best. Also, when people typically do star grounding, they forget all about the positive side, and only think about the "ground". Your B+ needs to appears as an AC ground, so it is just as important as "ground".

The current is flowing from B+ to ground. Large current loops between B+ and ground increase its capabilities of radiating any voltage potential that shows up with those currents. Also, large current loops increase the ability of the circuit to "receive" stray signals, such as signals generated by your circuitry, radio stations, CB radio, and magnetic fields such as those generated by you power transformer, or even your house/club AC wiring.

In order to decrease this with "star grounding", use the same methodology with your B+ lines. Ground your circuit right near your B+ line (negative side of power supply cap that B+ comes from) and run your B+ and ground lines together. Even twist them to reduce your loop area and cancel out any potential radiated signals. Same idea that you use when running your heater wires. You have 6.3 AC Volts on your heater lines. When your running your amp flat out, you are likely to have similar AC voltage differences between your B+ and ground in some areas.

Ok, most likely, you will not need to do this with amplifier circuits. In most cases a ground bus or typical star grounding is fully adequate. But if you have problems with high frequency oscillations or RFI (radio stations coming through your amp) this reduces your amps ability to radiate signals that can cause oscillations, and reduces your ability to receive radio signals that can be demodulated and come through as noise or AM radio in yout amp.

Ok, done rambling.

CSB

[Shad]

Here are some output shots running guitar into a real cabinet.
For some reason I couldn't sweep my generator into making the spikes today like yesterday. I wonder if the cap that blew in the PB was failing and screwing with what I was seeing.

So yeah, anyway the eyes and ears tell me those little spikes are the offenders. I can see them appear as I hear them.
Oh and my KACK sound has returned again and doesn't want to go away with the same fix.

Quick questions.
Why would these spikes appear with the guitar, but not the generator?
Perhaps I need to fool with it some more...
Am I right in thinking it might be ideal to get these spikes to appear on my scope, using the generator, and working attenuator, then try to rework my lead dress into making them go away?

I think I'd go nuts trying to do it with the guitar. I don't have anyone to play it for me, and I don't want to be messing about with a guitar in hand while poking around in there. Even I know ths much.

Can you give us a shot of the whole board and how your running your ground bus or star ground?

Will do.

Are you running shielded wires for your grids.

For V1 and V2 I am.
For the PI and to the output tubes, I am not....

Wow, ok readig your last post on grounding my brain if starting to smoke like that PB did last night. My brain is alrady reeling in too many directions. I can follow what your saying for the most and all, just not at THIS moment. LOL!

I need to give the noggin a break tonight. This thing is killing me.
I will get board shots tomorrow first thing.

BTW I did not see any rasp on my power lines with big signals pushed through. It did change, and wiggle a lot, but no rasp as described in Darr's book.

Brain going bonkers in all directions. Gotta stop thinking about it for a spell or I won't get focused.

Thanks for all the help. Check in tomorrow and I'll have shots of the board and maybe some more info. Will try lead dress again when it won't piss off the neighbors.

Hmmm, B+ and grounds.....I do have some rather long power leads, and I haven't fooled with them yet.....No must stop now or I might hurt myself.

Till tomorrow then. And again huge thaks for taking time out of your day to assist a stranger with a strange amp.

[csbradley]

When you see funky stuff going on, try narrowing the time (X axis) of your scope to see whats going on. There may be something going on that you can't see with you current setting.

As far as the guitar thing goes. Do you have a looper pedal or delay pedal. Just set it to loop the same thing, or set the delay on infinant decay.

CSB

[veets]

"I have an 18 a friend built. Sounds amazing! Guy can't solder for squat (it was here because a main lead came loose!), lead dress is very scary, there is little rhyme or reason to it, few rules obeyed, sounds friggin' amazing! .......................................Pisses me right off!
Seriously though it is because his sounds so damn good that I keep at this. I know when I get it right I will be really really happy with the sound."

That's what leads me to believe that there is something fundamental and simple about why yours won't work right. Even if a fella solders like crap, can barely follow a cake recipe and yet can hammer one of these bad boys together with a shovel and a piece of chewing gum--and results in an awesome sounding amp. Doesn't seem to make sense, does it, especially with all the care and attention to detail you have obviously put into this build. I'd be pissed off too! Good work practices, and following the rules, stuff like that are important, but what really is important and makes these babies sing is the schematic, not whether or not one uses multi-core solder or the best brand available capacitor. The schematic is good, and if you follow that, your amp will sound good. I certainly don't want to insult your intelligence, and excuse me for mentioning the obvious, but could you have misread a multiplier band on a resistor or something simple like that? We've all heard the stories on here about what THAT can do to an amp's sound!!

[Shad]

When your signal is flat, you are probably clipping your V2 output. Pushing it into saturation. During this time, current will flow through the grid (P2) and distort the signal that you are seeing. You should be able to turn the volume down until you no longer see clipping on the Plate (P1), and the signal on the grid (P2) should look normal.

Sorry, can you explain what you mean by "flat"?
Also, when you say "volume" do you mean the MV or the "gain"?
I ask because when observing this clipping I have the MV on 0, and the gain control crack only a hair. this signal has this deformation no matter how little that gain control is up.
I will perfom this test again and verify that.
If what you say about turning the volume down and seeing the grid look normal is true, and I cannot in fact get that result, then perhaps there may be an issue around this tube?

When you see funky stuff going on, try narrowing the time (X axis) of your scope to see whats going on. There may be something going on that you can't see with you current setting.

Yeah, I have done that often now, for some reason I haven't yet on those output spikes. I guess I was meaning to but the pB blowout kinda threw me. Over that now though, so can take a look at it here ina minute or two.

As far as the guitar thing goes. Do you have a looper pedal or delay pedal. Just set it to loop the same thing, or set the delay on infinant decay.

Excellent idea, but I haven't anything like that on hand.
I did order new caps for the PB though. It was made in '95. So here is to hoping new caps will alllow me to use it with no more issues.
What I did do was just put the cab face down but up on one side a little to let the speakers breath. Set the guitar aways away, tuned a chord on it. Made sure I got spikes on the output, then took these shots of my output grids!

Two things I note right away and wonder what others think. First and foremost, no spikes! Looks like lead dress around my outputs is a place to work on....not you like haven't mentioned this before. Just trying to get my head around this scope, and trying to tie what I see on it, with whatever my problem will turn out to be.
The second thing I note, is that again I see that the top portion of my wave never hits the 10X marker, while the bottom portion dips lower. By about exactly what I am missing from the top. I guess this is back to V2.
I am going to check V2 out again right now, just to verify what I have and how it is working. If that clipping ought to go away on some setting, but mne doesn't....well maybe that warrants a closer look too.






Think I am going to inspect that V2 thing again, shield my grid leads to the output tubes, as you say it can't hurt at this point. See what that does, if anything, play with the lead dress around the outputs.
Have a lot more to read about grounding before I make any changes there.

there is something fundamental and simple about why yours won't work right.

Without a doubt it will in fact likely be the very simplest of fixes. Indeed it always is. It is the finding it that poses all the "fun".

yet can hammer one of these bad boys together with a shovel and a piece of chewing gum--and results in an awesome sounding amp. Doesn't seem to make sense, does it, especially with all the care and attention to detail you have obviously put into this build.

Yeah ok I'll cop to that, though I do think the 36watt TMB is a bit more complicated than the 18trem.

but what really is important and makes these babies sing is the schematic, not whether or not one uses multi-core solder or the best brand available capacitor. The schematic is good, and if you follow that, your amp will sound good.

Schematic? That wasn't drawn from the layout (with errors and all)?
Where the hck is that? The only one I has seen I was under the impression was made from the layout, which had errors to begin with.
I'll search the forums for that schematic but a link to it would be awesome!

I certainly don't want to insult your intelligence, and excuse me for mentioning the obvious, but could you have misread a multiplier band on a resistor or something simple like that? We've all heard the stories on here about what THAT can do to an amp's sound!!

Entirely valid point. To be honest, I bought the board pre-loaded and have only checked the board in sections. Not that that is an excuse for not checking, pointing out that I should have done that the minute I got it.
I feel dumb, but as dumb as I will feel if I do find an error.

[csbradley]

You do have a working 18 Watt, correct. You might want to crack it open, and just make some measurements with your O'scope so you will know what to expect on you 36 Watt. It will give you a chance to see what the signal is supposed to look like and how it sounds, without having to worry if its right or wrong, or how your going to fix it.

CSB

[bevins]

What layout did you use? Post a link to it, and I will zip up a schematic off of it if you think it will help. That way it will be easier to troubleshoot.

Let me know,

Bob

[StratoRat]

Shad - Did you ever resolve this issue and get you 36 watt working well?

[Shad]

Tsshhhh....invalid session!

The long and the short. Don't build it. Buy a GDS 36 watt kit and bypass a lot of headaches.
I have this notion about the 36watter that says someone posted it, others took that work and turned it into short term profits, the poster got no kick backs, the poster decides to NOT give away the fully realized end design.
Why should he? So someone else can make a kit of it? So that the schematic is readily available to every amp builder with an internet connection? Therefore turning any builder into potential competition?
Ritchie has the "secret sauce" (his words) and isn't going to give it up here. I think it has to do with dollar signs strictly. Ritchie can sell his amp to the bands, because one look at the forum will show that few actually seem to pull it off with colors, and he looks golden. Never seen Graydons schematic of this amp, I think he sells it along with his kit. I know from reading past threads that Graydons 36er is NOT the same.
I think if you want a good sounding 36er you will either have to buy one made from Ritchie, or buy a GDS kit, or tweak the basic design for a long, long time
I do not think the schematic that floats around is all it is cracked up to be.
Bitter? Well yeah I was actually. Now though I simply play my deluxe reverb amp I built that required no re-working my lead dress, nor even "chopsticking", nor any of the many, many other levels of hell I went through with two 36ers.
If someone comes up with something 36ish that actually looks like an improvement and not a blatant backwards step in the design maybe I will give that a try. If not I'll just recycle the iron into something that actually has legs and a real track record.