Arcing Issue

[Wilder Amplification]

Here's one for ya guys.

Basically stock Marshall from the "DC coupled cathode follower tone stack driver" to the speaker jack (modified front end). The issue is in the output section which is bone stock Marshall with a MetroAmp Dagnall clone OPT and the MetroAmp 1203-80 standup PT...100 watter. Amp is biased at approx 35mA per tube (OT primary current shunt reading = 69/67mA on each side of OT primary). Plate voltage = 450V while screens = 445V.

Now the only difference in this circuit is that I have a 0.01uF resonance cap wired in series with a 100K NFB resistor on the 16 ohm tap. In addition to that, I have a 0.02uF cap with a 1M pot wired across it as sort of a "2 stage resonance" circuit for some phat low end. The speaker cab is a Marshall slant 4x12 loaded with Celestion G12H-30 70th Anniversary Series speakers.

Now the issue I'm having is that when I max out the resonance pot, the power tube closest to the end of the chassis seems to arc to the rear panel of the chassis...or it could perhaps be arcing to the IEC Mains Inlet socket on the back...I can't see too clearly which as it happens very quickly. When it arcs there is a slight reduction in volume for a second. It doesn't seem to do it if I turn the pot down and shunt that cap out of the circuit...only when I have the pot maxed out and at "stage volume" does it do it.

Now...my question is could it be that I'm boosting too many lows via the NFB loop causing a severe flyback/back EMF? Possibly a bad power tube (although I personally have never ever seen a bad power tube cause that, I'm not ruling anything out for possibilities)?

Any thoughts and opinions welcome and much appreciated.

[Wilder Amplification]

Arcing Issue

Here's one for ya guys.

Basically stock Marshall from the "DC coupled cathode follower tone stack driver" to the speaker jack (modified front end). The issue is in the output section which is bone stock Marshall with a MetroAmp Dagnall clone OPT and the MetroAmp 1203-80 standup PT...100 watter. Amp is biased at approx 35mA per tube (OT primary current shunt reading = 69/67mA on each side of OT primary). Plate voltage = 450V while screens = 445V.

Now the only difference in this circuit is that I have a 0.01uF resonance cap wired in series with a 100K NFB resistor on the 16 ohm tap. In addition to that, I have a 0.02uF cap with a 1M pot wired across it as sort of a "2 stage resonance" circuit for some phat low end. The speaker cab is a Marshall slant 4x12 loaded with Celestion G12H-30 70th Anniversary Series speakers.

Now the issue I'm having is that when I max out the resonance pot, the power tube closest to the end of the chassis seems to arc to the rear panel of the chassis...or it could perhaps be arcing to the IEC Mains Inlet socket on the back...I can't see too clearly which as it happens very quickly. When it arcs there is a slight reduction in volume for a second. It doesn't seem to do it if I turn the pot down and shunt that cap out of the circuit...only when I have the pot maxed out and at "stage volume" does it do it.

Now...my question is could it be that I'm boosting too many lows via the NFB loop causing a severe flyback/back EMF? Possibly a bad power tube (although I personally have never ever seen a bad power tube cause that, I'm not ruling anything out for possibilities)?

Any thoughts and opinions welcome and much appreciated.

[Billy R aka DynaFreak]

are those 2 caps in series in the NFB loop? If so each one is causing a phase shift and together they certainly could be causing problems at certain freqs to the point of creating some positive feedback. Have you scoped the output with a square wave feeding the phase inverter?

[Wilder Amplification]

Yes they are in series. I have not scoped it yet...but I had in fact thought of the two in series causing a wierd phase shift.

I calculated the equivalent capacitance of the two in series and think changing to a single 0.0047uF across the pot will get me what I'm trying to do with the two.

[Chuck H]

If your scope has a "capture" feature you could set up the scope and induce the problem to see if the lack of feedback is creating a big whompin voltage spike. It would have to be huge considering the insulation resistance of open air compared to typical guitar amp voltages. What is the distance of this arc???

Chuck

[pdf64]

The resonance circuit can't be harder on the amp than running it without global NFB, and it should be able to cope with that.
My guess is that the extra gain is causing more mechanical vibration, thereby causing a bad contact or dry joint to fail mometarily and inducing a back emf spike, or just driving the power amp harder, increasing the transient switchover spike, causing insulation to breakdown.
So for the vibration issue, check all joints / connections related to the OT primary and secondary circuit - it may be something as simple as a poor socket on the speaker cab.
For the insulation issue, examine closely the area around the arc (which I'm sure you already have done).
Whatever, I'd replace the tube socket.

[Steve Conner]

I would say it's probably unstable and oscillating at some ultrasonic frequency.

[R.G.]

I would say it's probably unstable and oscillating at some ultrasonic frequency.

I think that's a very likely possibility. It's certainly something that has to be excluded before you can call it fixed. Time to dig out the 'scope, Jon.

[Billy R aka DynaFreak]

Some of the Peavey resonance/presence circuits work well, IMO. You might want to take a look at the 5150 II schem. The others like the 5150 or EVH120 are good also, but the way the schems are drawn it makes it a PITA to dope it out.

[Wilder Amplification]

Yeah a lot of manufacturers do that...but after a few minutes of tracing the scheme through the various pages you figure it out.

Now while there's obviously a phase shift issue in the NFB loop with the 2 caps, could this be a cause for the original arcing issue I'm having?

[Wilder Amplification]

I think that's a very likely possibility. It's certainly something that has to be excluded before you can call it fixed. Time to dig out the 'scope, Jon.

Yeah I actually should've posted this thread in the Maintenance & Troubleshooting section which I did. The guy who responded mentioned something about the fact that I have 2 caps in series in the NFB loop which could be causing enough of a phase shift to turn it into positive feedback at a certain frequency.

So I figure I'll just change it to a single 0.0047uF cap and see if that helps. It only seems to do it when I dial in the resonance pot (which puts the 2nd resonance cap into the circuit). If that doesn't fix it then I'll get it on the scope and see if I can see what's going on. More than likely I'll change out that power tube socket as well while I'm in there as pdf64 suggested, even though I can't see any visible carbon trace on it.

More to follow.

[Billy R aka DynaFreak]

It certainly seems that the circuit is either the root cause or a contributing factor. If it was me I would scope the output at a low pwr level with your circuit turned all the way up while sweeping the freqs of a square wave input, and look for any craziness at certain freqs, such as large increase in output, ringing, etc. Then I would do the same with a simple NFB loop installed. The other thing that comes to mind is a mechanical problem such as the pot losing contact intermittently due to vibration from "stage volume" sound. You might want to sub in a resistor to see if that is the problem. In any event, 2 caps in series in the NFB loop does not seem like a good idea.

[Wilder Amplification]

Well I ended up replacing that tube socket as suggested by pdf64, put a new set of tubes in and changed the resonance control circuit to a single 0.0047uF across a 1M pot. Gigged with it last night and no arcing whatsoever.

However, the tubes I installed are the JJ KT77s. I have them biased at about 63% dissipation (450V @ slightly under 35mA per tube). At idle I noticed last night that the corners of the plates were glowing red slightly on all 4. Not liking that at all...they're a 25W tube just like an EL34 yet they glow when biased to 63%. I'm gonna do some more testing on these tubes but if they cannot handle 63% of their claimed rated dissipation I'm not gonna use them.

Great sounding tubes though...they seem to be slightly tighter and punchier than the JJ E34Ls I had before but those tubes had been in my amp for almost a year and I've done various circuit tweaks in that time so it's not a valid comparison since I'm not comparing them to a new set of E34Ls.