New Marshall JTM-45 clone build making noise on high gain.

[PfeifferElectronics]

My new build is working now that I have the OT wired correctly. However, when I drive it hard on the high treble channel with the presence knob all the way up, I get popping and i can see an arc inside the KT66s. I also get an arc when I flip the standby switch to on.

With all the preamp tubes removed, it goes into machine gun mode as soon as I flip standby to on position. KT66s arcing each time it pops. I checked all my voltages and here is what i get on the power tubes.

Pin 3 = 460V, Pin 4 = 450V, Pin 5 = -54, Pin 6 = 448

Thanks for your help!

[MWJB]

Describe this arcing, is it distinct flashes of white/yellow light? Blue haze, that might flash on, on standby, is normal.

Your situation with the preamp tubes removed is not normal. Can you list the rest of the dc voltages for the 12AX7s please. What plate current do you have?

Can you post pictures, bear in mind that even Marshall struggle to build a perfectly working JTM30/1962, they are prone to oscillations at high treble & presence settings, so you might well have a few areas that need a tweak.

[PfeifferElectronics]

It's a blue haze that flashes and goes on and off repeatedly.

V1: pin 1 = 235V, pin 3/8 = 2.0V, pin 6 = 236V
V2: pin 1 = 198V, pin 3 = 1.3V, pin 6 = 328V, pin 7 = 197V, pin 8 = 198V
V3: pin 1 = 252V, pin 3/8 = 42V, pin 6 = 245V

Plate current for the 12AX7s?

Thanks.

[MWJB]

No just plate current for the KT66.

Sounds like you have some motorboating, check integrity of the grounds, see if you can measure any ohms from grounded eyelets to the chassis, check all pots/jacks/ground lugs are nice & tight.

Seems odd that your voltages at V3 are only fractionally higher than V1, despite the elevated cathode voltage, please measure the voltages at the junctions of the plate resistors feeding pin 1 & 6 for V1 & V3.

Still would like to see pictures.

[PfeifferElectronics]

Plate currents are 46.7mA and 47.6mA.

Voltage at the top of 100k resistors feeding pins 1 & 6 of V1 = 327V
" " & 82k feeding pins 1 & 6 of V3 = 379V

[MWJB]

"Here are some pictures too. This is my first build so it is not very pretty" Therein lies your problem, this is a circuit that I am very familiar with, yet it's taking me a lot of effort to decipher exactly what is what in your amp. Your layout is bad, you have way too much wire in there, the wire to your 2nd power tube pin 5 (next to the rectifier) is rediculously long - these wires need to be short.

Your power tubes are running too hot, but I doubt that has anything to do with your machine gun sounds. Pull out the board & start again, look at a genuine Marshall JTM45 or 5F6A bassman (or Doug Hoffmans layouts for these amps) & pay attention to layout & wire routing. What I do is scale drawings of the circuit...then I think about how I can improve things, do some drawings & post here for opinions.

[PfeifferElectronics]

So you are saying that because the wire to 2nd power tube is too long, I'm getting motor boating and machine gun noise? I realize my layout is terrible, and I will most likely rebuild the board, but I need to know if that is what is causing my problem.

Thanks.

[MWJB]

That wire stands out to me principally, but many of your wires are too long and parts are not laid out on the board in a manner that aligns them with where they terminate/feed. The original amps that your design is based on had issues with the layout (Marshall still do today...Fender too, but not with the 5F6A), your layout is not as good as them. Save yourself weeks & months of pointless tweaking, strip out the board & wiring now & set about revising it...you could go on for months doing it wire by wire, part by part.

It may sound like I'm being harsh, but many of us here recognise such mistakes because we have made them ourselves in the past, learned from them & don't make them anymore. No one is born knowing this stuff, such issues are not obvious until you learn about amp layout. Spend the time on the planning, then build once (well, ideally...the more & longer you do something, the more ways you find to improve it).

[Alex R]

What he said. The problem you're experiencing is likely to be internal electromagnetic feedback - wires from later in the circuit radiate signal and earlier wires pick it up and you get the feedback, motorboating, whistling, howling, whatever. Pay particular attention to speaker output wires on the one hand, and preamp grid wires on the other, when working out where to route wiring. It is well worth sticking to the original layouts, they were indeed designed with this issue in mind.

[greekie]

The JTM45's I've seen have almost all had a severe oscillation problem with the newer KT66's. Install a 5.6k resistor at each output tube grid to stop this. It can very well lead to your tubes being killed in the end!

YM2C

Jake

[Jag]

I agree with Alex and MWJB, you'll have major frustration chasing your tail with that layout. It's not the sort of thing to fix once you get rid of the bugs. That layout probably IS the bugs.

[Jag]

The only suggestion I would have regarding the amp, as is, would be to check that you don't have the filter for your bias supply wrong (+ should be to gnd for that one). Don't ask me why I make this suggestion, it comes from presonal esperience.

[PfeifferElectronics]

Jag, thanks for you advise, I was scratching my head on the orientation of the bias filter caps too. I had to find a picture of a completed build to confirm they are + to ground.

[Steve Conner]

I do think it would be worth trying greekie's suggestion before scrapping the layout. Grid stopper resistors can make the circuit more forgiving of layout. Remember they need to be mounted directly on the tube sockets. If they don't help, THEN scrap it.

For me this is the real voodoo of tube amps. A good designer makes a circuit that is tolerant of layout. The performance of the circuit is determined by the components you can see, and the "hidden circuit" only has minor effects.

But the guys who made the classic tube amps were not good designers, otherwise they wouldn't have been designing amps: the Cold War-era military industrial complex would have snapped them up. And they were under extreme cost cutting pressure, too. Who knows what crazy stuff is going on in there with stray capacitances, coupling through rails and so on. It's not on the schematic, that's why it's called the hidden circuit.

[greekie]

... military industrial complex ...

Great Eisenhower-quote!

Jake