Strange issue, debugging help

General discussion area for tube amps.

Moderators: pompeiisneaks, Colossal

User avatar
Malcolm Irving
Posts: 139
Joined: Wed Jan 14, 2015 12:06 am
Location: Derbyshire, UK

Re: Strange issue, debugging help

Post by Malcolm Irving »

WRC34 wrote: Tue Apr 18, 2017 1:28 pm I have chopstick'd the amp a few times, hoping to turn up the issue as simply being a cold solder joint. This was discussed earlier in the thread. And I consider myself more than competent to work on a live chassis - how else would I have obtained the voltage readings I posted?
Apologies. I should have noticed the voltage readings imply that. I'm just a bit wary of advising things like that. Chop-sticking was mentioned earlier, but sometimes that just infers moving wires around a bit. As well as solder joints, do give resistors, tube sockets, capacitors etc. a few big hits in different directions - it can identify a crackle like this.

The type of crackle you describe is usually a series of minute electrical make/break contacts being made somewhere (including possibly inside a component). This can be stimulated electrically or mechanically, so it might be worth trying the chop-sticking with a guitar signal present as well.

Not trying to question your competence - just explaining what I would do myself in the circumstances.
User avatar
WRC34
Posts: 169
Joined: Fri Dec 19, 2014 3:40 am
Location: United States
Contact:

Re: Strange issue, debugging help

Post by WRC34 »

No problem, Malcolm. I appreciate your input in any case. Thank you for taking the time to reply.

I have just re-chopsticked the amp fairly aggressively. The only components that make any noise whatsoever are input grid stopper to the input tube, 6SJ7, which I feel is natural (?) and the 6SJ7 screen bypass cap, which is only noisy when tapped hard on one side. Tube sockets are all quiet and unresponsive to tapping w/the trusty chopstick. Plate resistors are all quiet. They're all new-to-this-build 1W rated carbon comps.

I actually swapped out the V1 grid stopper a week or so ago in this process to see if that would help - it has not.

In addition, I just swapped out the phase inverter filer cap with no difference/results.
User avatar
sonicmojo
Posts: 669
Joined: Thu Mar 03, 2011 3:41 am
Location: Oahu, HI

Re: Strange issue, debugging help

Post by sonicmojo »

Does the issue still only occur in the head cab as in your first post description? That seems like a big clue if the case.

If so, perhaps check that your shielding in the cab(if it has any) is close or making contact with your power or standby switches. If it has mesh, a stray strand could be floating around. I had an amp where the aluminum tape got crumpled a bit and made intermittent contact to a switch.
---------
Bryan
User avatar
WRC34
Posts: 169
Joined: Fri Dec 19, 2014 3:40 am
Location: United States
Contact:

Re: Strange issue, debugging help

Post by WRC34 »

sonicmojo/Bryan - the issue now persists regardless of whether the amp is in the head cab or not. I once owned a D Z Route 66 which was giving me a similar problem and it ended up being that the chassis bolts were not tightened enough. However, in this case the issue has now proved itself to occur even when the chassis is removed from the head cab and upside down on the work bench on the other side of the room from the cabinet it's driving (aka free from contact interference and vibration).

The reason I swapped out the filter cap which fed the phase inverter first is because it is a 22uf 500V F&T cap, which is the exact value F&T I had an issue with brand new last year. I will move on with the filter cap replacement theory and swap each one out one by one, if this yields no result I will find another course to pursue and keep y'all posted. This is the most brain busting issue I have ever had on a scratch build.

I do not have a schematic because I built from my "mind" I guess. It's a cathode biased power section using an AC30 or Trainwreck Rocket type power transformer only instead of a quad of EL84s it uses a pair of EL34s. 6SJ7 input pentode into a 6SL7 long tailed pair phase inverter. Kind of like an octal version of EF86 into 12AX7 ltp into push pull power amp. PT secondary is rated for 260 mA DC which I think should be more than enough to handle a pair of EL34s cathode biased. 6.3V winding is a whopping 8.8A and 5V winding is 3A. Using a 5AR4 rectifier and a 5691 in V2 spot as phase inverter. These pull more filament current but behave similar to 6SL7. Since I have the filament current to spare I figured this would be fine. Using approx 3.9A of the 8.8A available.
User avatar
sonicmojo
Posts: 669
Joined: Thu Mar 03, 2011 3:41 am
Location: Oahu, HI

Re: Strange issue, debugging help

Post by sonicmojo »

I feel your pain. How about the upside down canned air freeze test on some parts. I had a bad carbon comp resistor once that was very hard to chase down. Found it that way. it was the fried egg sound thing. Don't overdo the freeze but it might help you find a bad component or solder joint that the chopstick does not. Just be careful as it tends to add some condensation as it evaporates.
---------
Bryan
User avatar
Malcolm Irving
Posts: 139
Joined: Wed Jan 14, 2015 12:06 am
Location: Derbyshire, UK

Re: Strange issue, debugging help

Post by Malcolm Irving »

Have you thought about building a ‘signal sniffer’? That is a very small amplifier with a speaker (maybe based on an old radio or whatever). The input is protected from the DC in the amp by a 0.1uF cap on the ‘hot’ signal side. This can be used to probe different points in the amp to see where the crackle signal is present. It’s not quite as good as a ‘scope, but almost.
For example, is the crackle present at the output of V1? You can infer some things by listening to the crackle sounds. For example, the crackle may be present on one or more B+ nodes as well as in the signal path. Is the ‘tone’ of the crackle affected by the tone controls of the amp? Is it affected by volume control(s)? Obviously, the crucial question is where the crackle first appears, as you trace the signal path from input jack onwards.
User avatar
WRC34
Posts: 169
Joined: Fri Dec 19, 2014 3:40 am
Location: United States
Contact:

Re: Strange issue, debugging help

Post by WRC34 »

Thank you again Malcolm and Bryan for your ideas.

I will try and describe the crackle in fine detail - Bryan, yes I like your choice if words, the "fried egg sound" it follows immediately after a note is played, and ends very quickly i.e. it's never a sustained crackle and it is much more noticeable on lower notes especially palm muted power chords like a low E or D. In fact, I have been working on this so long that now whenever I make a change and power the amp back up to test I just chug-a-chug a few palm muted chords which tells me immediately if there's any improvement. The brief time it seemed fixed it was acting fine with the head right on top of the cab at very high volume. My home tests were good and semi-extensive, and I felt confident to bring it to practice. In fact, it actually lasted throughout an entire 2 hour rehearsal cranked and only at the next rehearsal did it start acting up again and that was 1 hour into it. When the issue returned, as mentioned before, it seemed that a change of output tubes solved the problem, but of course since the problem had returned I wanted to examine further, so back on the bench it went. Since then the issue has been constant on the bench, regardless of any change I've tried. What is so confusing is that my first "fix" which was swapping out the first coupling cap, solved the problem entirely and the amp lasted for a grand total of about 4 hours of sustained high volume playing before the problem returned.
Firestorm
Posts: 3033
Joined: Fri Jan 25, 2008 7:34 pm
Location: Connecticut

Re: Strange issue, debugging help

Post by Firestorm »

It might help to isolate the problem to a specific stage: attenuate the post-PI signal to see if it's exclusive to the power stage; attenuate the pre-PI signal to see if it's just in the preamp.

That said, it sounds very like a somewhat less noticeable effect that occurs in the reverb channel of many BF and SF Fenders where the length of the tone stack leads permits parasitic coupling to other components. That one sounds like a fizz riding the note or chord and ends abruptly as the signal decays.
User avatar
WRC34
Posts: 169
Joined: Fri Dec 19, 2014 3:40 am
Location: United States
Contact:

Re: Strange issue, debugging help

Post by WRC34 »

Firestorm, pardon my ignorance, but how would I attenuate the post/pre PI signal?
danman
Posts: 33
Joined: Sun Apr 20, 2014 6:34 pm
Location: Richmond Va

Re: Strange issue, debugging help

Post by danman »

I have had a bad standby switch give me similar problems on a particular amp of mine.
User avatar
WRC34
Posts: 169
Joined: Fri Dec 19, 2014 3:40 am
Location: United States
Contact:

Re: Strange issue, debugging help

Post by WRC34 »

danman, thanks for your input!

This amp has no standby switch. Power switch only and 5AR4 rectifier :)

That being said, could a bad power switch do the same thing??
User avatar
WRC34
Posts: 169
Joined: Fri Dec 19, 2014 3:40 am
Location: United States
Contact:

Re: Strange issue, debugging help

Post by WRC34 »

Also, Firestorm, this amp has only a volume & tone control, no "tone stack" really, but the leads to pass by each other. There is about 1/2" space between the lead from the volume pot to the "top" side input grid of the phase inverter and the leads which connect to the coupling caps feeding the output tubes. Could this be the culprit? The leads themselves do not generate much noise when tapped w/ye olde chopstick.
Last edited by WRC34 on Thu Apr 20, 2017 2:19 am, edited 1 time in total.
User avatar
WRC34
Posts: 169
Joined: Fri Dec 19, 2014 3:40 am
Location: United States
Contact:

Re: Strange issue, debugging help

Post by WRC34 »

Malcolm Irving, the "signal sniffer" you described sounds interesting - obviously not just for this issue but for easier troubleshooting in general - but I'm a little confused as to what exactly it is and how you would implement it. Is it a transistor type amp, like those little "smokey" amps that used to come in cigarette packs? And how would you "attach" it to different parts of an amp/circuit to diagnose an issue? With a problem like this, where the issue is only happening at high volume, how would you separate the sound of "main" amp itself from the signal sniffer?
Groove1
Posts: 25
Joined: Sat Dec 05, 2009 12:37 pm
Location: 230V, Germany

Re: Strange issue, debugging help

Post by Groove1 »

Last week I had a crackle at the low E-string fret #3 and at the G-string fret #7. It was a TW Express. Before the 3rd stage (V2a) I changed the 100K resistor back to the regular 150K. After this step, the issue was gone. Maybe it helps. Good luck.
User avatar
Malcolm Irving
Posts: 139
Joined: Wed Jan 14, 2015 12:06 am
Location: Derbyshire, UK

Re: Strange issue, debugging help

Post by Malcolm Irving »

WRC34 wrote: Thu Apr 20, 2017 2:17 am Malcolm Irving, the "signal sniffer" you described sounds interesting - obviously not just for this issue but for easier troubleshooting in general - but I'm a little confused as to what exactly it is and how you would implement it. Is it a transistor type amp, like those little "smokey" amps that used to come in cigarette packs? And how would you "attach" it to different parts of an amp/circuit to diagnose an issue? With a problem like this, where the issue is only happening at high volume, how would you separate the sound of "main" amp itself from the signal sniffer?
Any kind of amp really - preferably one that is 'expendable' - I use a little plastic transistor guitar practice amp that I got free with a magazine subscription. (I've used it a fair bit and so far I haven't blown the sniffer - but you never know.) The shielded side of the input lead clips to any ground in the main amp. The signal side of the input has a 0.1uF cap to block DC from the main amp and then a probe lead. The probe can be attached more or less anywhere in the main amp (you will get a clunk as the 0.1uF charges or discharges, but usually by turning the sniffer volume right down it will work even where there is a big signal swing). You would need to run the output of the main amp into a dummy load resistor to avoid the sound of the main amp obscuring the issue.

Best to use a battery powered sniffer amp to avoid any ground loop issues.
Post Reply