Parasitic Oscillations/ Glowing Screens

[Kegger87]

Hello all,

I've recently completed a 1987 build and have it up and running sounding good, however at higher gain I have V4 power tube glowing the screen with the intensity of the music. Only V4 is doing this and swapping the power tubes, the problem stayed with V4, not the tube. Putting a sine wave in the input and outputting to a dummy load with an oscilloscope, I get a clean sine wave until turning up channel 1. The scope goes crazy unreadable before the scope can even show distortion on the waveform. I have 1k grid stoppers between pins 4 and 6 as per the original design. I've ordered 2k2 resistors to try, as I have read Marshall tried these later on. What else can I try to eliminate this screen glow and horrible oscilloscope noise? Suggestions on anything from lead dress tips to anything else I can try are greatly appreciated.

[JMPGuitars]

I'm blunt, don't take it personally...

The lead dress is a nightmare in that amp. ALL AC wires MUST be twisted. That includes the wires to the power switch, and the heater wires.

Please share your schematic and layout so we can comment on your tube wiring and lead dress there.

How are your preamp tubes heated? You don't have pins 4 and 5 connected together on any of them. If they're 12AX7 tubes, whether wired in series or parallel, pin 5 needs to be connected.

Thanks,
Josh

[Kegger87]

Hi Josh,

The heaters are jumped from pin 4-5, it was caught in my earlier troubleshooting (old picture). This is a Modulus kit under their '69 small box model. I've since redone the ground points and I know I need to clean up the wiring mess between the filter cap to board area. All my ac lines are twisted with the exception of the mains wiring to the left of the power transformer ( I was using the layout photos on the modulus site as a guide where they were not twisted). The amp has virtually no hum, deffinately no more than any of the vintage amps in my fleet that are original. I take no offense to criticism I need to learn! I appreciate any and all feedback.

[JMPGuitars]

Ah, okay. Well, you should get us current photos to look at.

There's some questionable solder joints in the photos, but that's not likely to have anything to do with your current issue.

I would suggest starting with using shielded wire on the input jack(s) and gain pot(s).

[Kegger87]

Ok here is the current state of the chassis, and the layout that was provided. Today I'm going to try and shorten up and tidy the off board wiring between the filter caps/ power transformer and the board, I'm not happy with that rats nest there. I should mention any oscillation I'm getting on the scope, is not audible, and the amp sounds really good. What tipped me off to a possible parasitic oscillation was the glowing screen on V4 that changes with the intensity of playing. I'm getting 450 volts on the plates of the power tubes and around 443 on G2.

[JMPGuitars]

The lead dress is a nightmare in that amp. ALL AC wires MUST be twisted. That includes the wires to the power switch, and the heater wires.

This statement still holds. Twist your AC wires to the power switch, I'd twist the standby too.

Fix your heater wires too. They're too long, and not twisted all the way to the socket.

I would suggest starting with using shielded wire on the input jack(s) and gain pot(s).

^^ That too.


Your screen voltages are lower than the plates - which is good, but it might be a good idea to lower the screens a little more.

Higher value grid(1) stoppers are definitely a good idea.

[Kegger87]

I didn't get a chance to work on the amp today, hopefully tomorrow I can get time to get in there. As far as routing things better near the power tubes, are there any long standing rules for routing the OT leads, filter cap feeds and so on? This area of mess im really not happy with. I will remove and rewire the mains lines with twists as well. In the future I'm going to use solid core wire for the heater runs as the stranded wire doesn't stay fully twisted when trying to solder to the sockets. I may bite the bullet and redo these too, not just for performance reasons but asthetic too, to appease my ocd. You mentioned adding grid stoppers to G1, would these be positioned in series between the feed line and pin 5? I received 1k and 2.2k 5 watt resistors today to try.

[JMPGuitars]

I only use stranded wires. I prefer PTFE 20AWG for most of the amp, with 18AWG for the mains and power switch.

PTFE stranded wire holds really well, but I still like to add heat shrink anyway. If you're OCD at all, you'll appreciate this:

jmpguitars-heater-wiring.png

[Kegger87]

Looks great! PM me some more build pics, I can always use good and tidy references of quality builds. I still haven't had a chance to do any rewiring but I'm hoping to at least follow an injected signal through the preamp with the oscilloscope tonight and see if my high frequency oscillation is before or after the phase inverter. I'd like to find out where I went wrong before shotgunning multiple wiring issues and never learn the problem. One thing I'm still unclear on is adding a grid stopper to G1. Is this typically put on the tube socket in series with the wire connection? My turret board is full and I don't have any other real option, other than "floating" the resistor off the socket pins.

Thanks again!
Craig

[TriodeLuvr]

You mentioned adding grid stoppers to G1, would these be positioned in series between the feed line and pin 5? I received 1k and 2.2k 5 watt resistors today to try.

Grid stoppers should be 1/2W carbon comp. Any value between approximately 1K and 2K will work well for suppressing parasitics.

One thing I'm still unclear on is adding a grid stopper to G1. Is this typically put on the tube socket in series with the wire connection? My turret board is full and I don't have any other real option, other than "floating" the resistor off the socket pins.

The grid stop should be attached as close as possible to the socket pin. Can you add a small single lug or two lug terminal strip on the socket's mounting ear for a tie point? If not, I would consider it acceptable to solder the wire directly to the resistor and protect the joint with a short length of heatshrink.

Jack

[Kegger87]

I finally found some time to work on the amp this evening. I desoldered the filter caps, standby and mains wiring, output transformer primaries, and grids of the power tubes. Everything is now twisted and rerouted to cross any other wiring at right angles and be as short as possible while still running with some care around the sockets. Firing back up with a signal generator input and the oscilloscope on the output, I was able to keep a nice sine wave without collapsing into oscillation all the way up the volume. This was not possible before. It is now midnight, so a play test after a quick bias check will commence at a more neighborhood friendly hour tomorrow.

[JMPGuitars]

Definitely looking better. I would also suggest at least zip-tying your pin 9 heater wires there so they're tight together.

[Kegger87]

I still want to completely redo the heater wires but for now the amp isn't very noisy at all so once it's running happily I'll dive back in ( at least for cosmetics). I had a chance to play test it today around 6 on the high treble and the screen of V4 is still glowing with intensity of playing. Now I think it is time to start looking at adding/ changing grid stoppers on G1 and G2. As I understand, "factory" values on G1 should be 5.6k and no higher than 2.2k on G2. As mentioned earlier my voltages on the screen vs plates was only about 7 volts different. Which screen resistor position should I focus on first? I have to order the 5.6k so I might up the G2 value from the currently installed 1k to the 2.2k as a first attempt.

[TriodeLuvr]

... the screen of V4 is still glowing with intensity of playing.

Maybe the tube has been damaged by previous issues. Looking at the kit layout diagram, I don't see anything that should cause this. Both screens have 1.5K stoppers, and the screen voltage supply is bypassed by the second cap in a CLC power supply with a value of 50uF. Is only one output tube doing this? What's the voltage across the 1Ω cathode resistors at idle?

Jack

[Kegger87]

Alright I've now changed the g2 5w resistors from stock 1k to 2.2k on both tubes between pins 4 and 6, and added the 5k6 1/2 watt to both pin 5 in series. Plate voltage now 444v with a bias of 37ma on the hotter tube giving approx 65% dissipation. The amp sounds noticeably better, tighter would be how I describe it, but the screen on v4 is still glowing with peaks of playing intensity. It is of note, that the glowing is much less pronounced. I'm beginning to suspect these tubes suffered too much damage while testing, to ever stop glowing again. I swapped the power tubes after a quick play test, and now the tube that was in v4 glows in the v5 spot (v5 never glowed at all prior). The "good" swapped tube now in v4 glows , but far less than its parter did in both the v4 and v5 positions. After a cool off period I'm thinking I'm going to swap in my xf4 mullards, rebias back up and see if these "more robust" old production tubes are affected at all. The one thing I forgot to check was the new voltage at pin 6 after the resitor swap to see how much less voltage is on the screen. Before the tube swap I'll try this measurement along with the voltages across the cathode resistors and report back.

[Kegger87]

Ok so the grids are running at 428-430 volts with the plate voltage at 444. Measuring across the 1 ohm cathode resistors I get 35mv on v5 and 40mv on v4. What should I expect to read?

[TriodeLuvr]

Those numbers are generally OK, but the tubes don't seem to be very well matched. Are the cathode resistors 1%?

[Kegger87]

Yes I thought the tubes were not very well matched either... The cathode resistor values are and they measure very close to perfect.

[JMPGuitars]

Matched isn't that important. ZP would argue you get a better sound with unmatched tubes. There's lots of opinions on both sides of that discussion.

[TriodeLuvr]

The problem with mismatched tubes is the possibility of transformer saturation and distortion at low frequencies. That doesn't generally produce an appealing audible effect IMO, and it ultimately limits output power. If we're looking for more even-order harmonic energy, the phase inverter is the place to do it.

Jack