"Black Flag" JTM100 - the build has begun!

[shakti]

OK, this stuff is the weirdest...I really need all the help I can get. Here's a few more hints:

- the bias oscillates a small bit (say +/- 1 mA) when switching the standby switch and no signal in, all preamp tubes in, all knobs on 0.
- turning the tone controls up a bit helps stabilize the bias.
- removing V3 makes the bias totally stable, regardless of tone controls.
- removing V2 does not have any effect on the problem. Tone controls still affect the bias stability.
- with V3 in, taking a reading of the B+ line does not affect the bias, up to and including the PI plate node. Taking a DC reading on the other end of either PI plate resistor (82k or 100k), makes the bias oscillate heavily, with a range from around 22mA to 50mA (all tone controls on 0). And here's the really funny part: TURNING THE TONE KNOBS SEEMS TO AFFECT BOTH THE RATE OF BIAS OSCILLATION AND THE RANGE OF OSCILLATION! This is with no tube in V2. Turning the tone controls can even make the bias stable again, but then at another idle point than before (usually a lower mA reading, i.e. colder bias).

WHAT ON EARTH IS GOING ON??

[shakti]

Putting another 47pF fizz cap in parallell with the existing one does not change anything. I didn't solder it in, just using alligator clips, but that shouldn't make a difference.

I'll take any suggestions at this stage.

[Kevin]

This is some wacky stuff. Have you tried swapping in new output caps yet? I thought you did the forming in another amp. If you did have the B+ on those caps while forming, then you very well may have smoked them. I don't necessarily know that the amp would behave like that, but it's worth a try, if you haven't already...

I'll take any suggestions at this stage.

It could be your board material. You might wanna swap that out for some new stuff....

[shakti]

Some further clues(?):

With no V2 in, turning the bass knob to 10 immediately stabilizes the bias (all other controls at 0). Turning treble up narrows the range of oscillation (all other knobs at 0). Turning up the mids (all other controls at 0) does not affect the oscillation in any way. Putting in V2, the mids works similarly to turning up the bass.

Edit: scrap that about the middle control...it doesn't seem to affect the oscillation, with or without V2 in.

[shakti]

In fact, with bass at 10 (everything else at 0), the oscillation does not happen at all, whether from standby, or "started off" by reading DC at V3s plates.

[shakti]

This is some wacky stuff. Have you tried swapping in new output caps yet? I thought you did the forming in another amp. If you did have the B+ on those caps while forming, then you very well may have smoked them. I don't necessarily know that the amp would behave like that, but it's worth a try, if you haven't already...

I'll take any suggestions at this stage.

It could be your board material. You might wanna swap that out for some new stuff....

I did the forming of the mains filters in another amp, as those are wired in a way that makes it impossible to form them "Larry-style". Then I put them in this amp, and proceeded with forming the rest of the caps in this amp on first fire-up. At that stage the output coupling caps were probably subjected to more than 400V.

Yeah, gimme some real Paxolin, not this tired crap!

[shakti]

Finally:

- while playing, the bias sometimes drops below idle in between peaks. I guess maybe this is normal, I never noticed on my other amps, but I thought I'd mention it.

- At higher volumes, I'm still getting some nasty crackles and fart-outs in the bass.

I gotta hit the sack now, and expect to wake up to a solution suggested by one of you fine gentlemen!

[neikeel]

Sorry I missed all this excitement/hassle my email has not been sending 'new message' memos?

Regarding the NFB and mixers on the original although not black flag it is probably one of the earliest dual rectifiers: http://s76.photobucket.com/albums/j2/ne ... Superlead/

Mixers:


NFB (27k on 8ohm tap):


Regarding my Black Flag I found it took a long break in (I use yellow labe Mullard ECC83s and, IIRC in the new build xf4s) I used a 5000pF tubular ceramic bright cap (like my 67) and found that changing the choke (or was it because it was finally burned in) worked the trick and it sounds pretty good now (I used new 47uF RIFAs on the board - as per the pics on first page and did not reform them - preamp can is one of Rich Ms).

Your first issue (hitting a big open chord) sounded like a poor solder joint in the preamp (most likely) or faulty preamp valve.

The current issue sounds like one I had (on a 68SL clone) which drove me mad until SDM suggested (quite rightly that I had not run a dedicated earth back from the bias circuit - I got the modulation, unstable bias, weird noises whenever I increased the bias current above 10mA).

Just a thought

[shakti]

Regarding my Black Flag I found it took a long break in (I use yellow labe Mullard ECC83s and, IIRC in the new build xf4s) I used a 5000pF tubular ceramic bright cap (like my 67) and found that changing the choke (or was it because it was finally burned in) worked the trick and it sounds pretty good now (I used new 47uF RIFAs on the board - as per the pics on first page and did not reform them - preamp can is one of Rich Ms).

Your first issue (hitting a big open chord) sounded like a poor solder joint in the preamp (most likely) or faulty preamp valve.

The current issue sounds like one I had (on a 68SL clone) which drove me mad until SDM suggested (quite rightly that I had not run a dedicated earth back from the bias circuit - I got the modulation, unstable bias, weird noises whenever I increased the bias current above 10mA).

Just a thought

Thanks for your suggestions, Neil.

Mine has a Metro 352-114 choke already, and Marstran 1202-119.

I guess a poor solder connection in the preamp is possible, but I doubt the cause of my current problems. It's the first time I'm using the split top turrets. I like them when I have only one component attached to the turret, but with more components it gets crowded...still, I think my soldering on this amp is my best yet. I'll reflow them just for the heck of it.

My PT bias tap is grounded at star ground #4, along with the screens and bias cap ground, so that's not the problem.

I'm currently "cleaning up" the wiring of the B+ back to stock Marshall specs and trying to keep everything as short as possible. If that doesn't fix it, I'll check for DC leakage on the output coupling caps. I'm leaning towards those being the culprit. I notice I don't get quite as much range on the bias adjustment as I usually do, so it's possible that the couplers are leaky.

[arledgsc]

I had a strange problem with my JTM45 when new. Compressed/distorted sound, bias OK one minute and the next not. Turned out to be a bad solder joint of the OT primary at the power tube.

I would double check every solder joint in the OT and HT power circuit. If you have carbon composition resistors as voltage dropping resistors in the B+ feed circuits I would change those to something more stable as a test. Also any flakey B+ component can cause wildly oscillating voltages. Popping sounds are common to bad caps while questionable resistors usually cause a random noise. Freeze spray in the past has been good at times for isolating marginal components. Hit them with the spray and listen for changes - better or worse.

Good luck and let us know what you find. Scott

[shakti]

Thanks for all your suggestions.

I've cleaned up the B+ line and reverted it back to stock Marshall layout, but it did not make a difference. I do have CC voltage drop resistors, but I really don't think those are the issue. Neither do I think a poor solder joint anywhere is the problem. I've tapped just about every connection in the amp, but that does not elicit any problems. Yes, there is some sort of random noise in the amp, but I'm convinced that it's just a byproduct of the main problem. And the main problem is:

- bias is unstable when switching from standby, but just by a few mA. If I turn up the bass, the bias stabilizes.
- applying a signal will sometimes make the bias oscillate heavily (meaning it doesn't go back to idle bias after stopping the input signal)
- similarly, taking a reading of DC on the far side of the PI plate resistors makes the bias oscillate heavily. Turning up the treble limits the range of oscillation, turning up the bass makes the oscillation stop.

Next up is to measure for DC leak on the output coupling capacitors. I suspect that may be my problem.

[shakti]

Well, the output coupling caps seem fine. De-soldered from the 220k bias splitters, I measure around 250VDC on the V3 plate side, and 0V on the other side. At initial fire-up I get a reading of a few volts, which quickly (<5 seconds) drops to 0V. That seems fine, doesn't it?

I notice I get very little range on the bias pot with a 27k resistor before the diode, and a 47k in line with the 25k bias pot - only from about -30 to -37VDC. Is this normal? It is exactly the same with the coupling caps disconnected, i.e. this is the range I get from the bias circuit alone.

I'm running out of ideas... However, as turning up the bass pot (i.e. increasing resistance) makes the problem with bias oscillation go away, I'm thinking it could be something in the tone stack. Any suggestions? I really hope it's not one of the 0.022 mustards...but how can I check them in those positions?

[neikeel]

At this stage, having checked all of the things that you have done I would send out an SOS to SDM he is very helpful in these situations if he has time.

ps I have had similar (but not identical) issues with a faulty bias cap, but the interaction with the bass pot leads you to suspect a cap aound the tone stack.

[shakti]

Yeah, that's what I'm thinking too. Removing V3 takes the oscillation away, so it must be something in the preamp. And bias voltage at pin 5 stays steady even though bias reading through the 1 ohm resistors swings heavily. So it doesn't seem likely that the actual bias circuit is at fault.

I did notice in George's 100W instructions he says to ground the bias ground along with the heater CT. On mine I have the heater CT at star ground #1 (along with inputs and V1k). Bias tap ground from the PT is at star ground #4, along with a ground lead from the bias caps, and screens ground.

Funny thing is that since the bias oscillates at a completely steady pace (cycle of around 2 secs), I'm somehow thinking maybe residual AC/ripple is "leaking" into the tone stack or PI? The amp sounds fine, but there's too much farting out in the bass and some nasty transients here and there, particularly while sustaining notes.

[shakti]

Well, regardless of the issues, I do really like the sound of this amp! It really shines with a Strat at medium volume levels, and does a beautiful semi-broken crunch sound. It really *does* sound like "If 6 Was 9" with a Strat! Very alive, responsive, glassy clean sound with girth. I can't quite get that "Catfish Blues" sound yet, it gets a little ratty with a fuzz, and the bass really gets out of control then....

It doesn't sound quite as good really cranked, but again it might be a burn-in issue. A little spitty and buzzy. Probably partly due to the low filtering, but that other "issue" I have may be part of the reason. Bass in particularly gets quite nasty at high volume levels. However, listen to "Outside Woman Blues" on the Cream BBC CD - that really sounds a lot like this amp cranked - spitty and nasty, yet still has the wailing Marshall sound! If I could only get rid of the worst buzz and fart in the bass and sort out the bias issue I'd be really happy!