Spitfire hum problem

[EtherealWidow]

Hey guys. I've completed wiring up a Spitfire for a friend of mine and am having some troubles with it. First is that the signal seems to get lost somewhere before the PI. If I poke around I can get "pops" coming from the speaker all the way up to the coupling cap just before the PI. I tried desoldering the input jacks completely and poking the pin 2/7 of V1 to no avail.

My second problem, which could be related to the first, is that I'm getting a very nasty 60Hz hum coming from the amp. I installed a hum balance pot with the wiper tied to the junction of two 220k resistors across the first filter cap to elevate the heater voltage as well in an attempt to reduce noise. The noise is worst at one end of the pot, and very much reduced at the other end with almost all the lower frequencies of the hum gone leaving only the harmonics of the hum.

My last problem is that there seems to be an intermittent short on one of the tube sockets. One of the power tubes is completely fine, but if I shift the other, I can essentially "turn off" the filaments (and therefore plate current as well) to all the other tubes in the amp except for the one power tube that is unaffected. I already replaced the socket of the suspicious power tube thinking that it was a bad socket, but got nothing in return.

I've rolled all the tubes and they all seem to be drawing plate current. The sockets are located underneath the board and therefore impossible to measure readings directly from the socket, but I took some other readings from the associated components.

V1
Rk dropping 4VDC
Ra dropping 60VDC

V2
Rt dropping 50VDC

V3/4
Rk dropping 9VDC

I'd be super grateful for any insight. Thanks much.

[frankdrebin]

so,how many volts you have on that pot?

[EtherealWidow]

Good point. B+ is 310V, so that's what's sitting across the pot. Upon realizing that the heaters were elevated to about 155VDC, I changed the "bottom" resistor on the B+ voltage divider to 15k, rather than the 220k. Now the heaters are sitting at 47VDC, but the hum problem hasn't changed any.

[EtherealWidow]

Alright. Couple things have changed. The 4V on the V1 cathode somehow changed to 2V. Cathode resistor was confirmed at 1k5, although the "220k" anode resistor was actually 22k. Haste makes waste.

Noise is still present, and now the PI is acting funny. Rt is now only dropping 30V, Rk is dropping .8V. One Ra is dropping 69V while the other is dropping 198V. I would suspect a shorted triode, but then that would mean that the Rk and Rt voltages would be much larger, right? In any case, the math isn't even adding up right for the tubes to be conducting like that. I triple checked the resistor values in light of the previous errors. Rt is 47k, Rk is 1k2, and both Ra's are 100k.

I feel I should add at this time (although I don't think this has anything to do with any of the problems I'm having) that the PT for this circuit was MASSIVELY oversized in terms of voltage. This is a salvage of a Gulbransen organ (the same that Cygnus used) that originally used EL84's as well (although I don't know how they used them given the PT size) Anyhow, I'm dropping all the secondary voltages with a 50R resistor on the primary side. It's dropping 22VAC. Further dropping B+ with a sag resistor, no tube recto.

[matt h]

(deleted)

[EtherealWidow]

Absolutely. I'll probably get those voltages posted tomorrow. Also,I need to specify there is not 310VDC sitting across that pot There should be somewhere around 6.3VAC with the heaters elevated, but I'll check tomorrow. Thanks for the help and insight so far guys!

[matt h]

(deleted)

[EtherealWidow]

So here's a bit of an update. The 22k resistor was swapped out when I posted that last comment and there was a change in cathode and anode (obviously) voltage. No change in signal though.

Found a cold joint on one of the EL84 cathodes that was the cause of the filament voltage going in and out for the other tubes. No idea why it affected filament voltage, but it did. We're all fine now with that. Still had a nasty hum and no signal until...

...Found that I had wired one of the OT primary wires to pin 6... fixed that and now I have signal making it through the entire amp. WHOO! Still have a horrible hum.

Investigated the weird anode voltages on the PI a bit further. Determined that there's some other source of current going through the anode resistor that's dropping almost 200V. Probably causing the hum too. That's why the tail resistor is not dropping a whole lot. Even checked the "top" of the cathode resistor and the "bottom" of the plate resistor to see how much voltage the tube itself is dropping: 14V. Unplugged the PI and that particular resistor is still dropping almost 200V. Nothing on the other Ra and nothing on the tail or Rk either. Unplugged all the other tubes too and that resistor is still dropping almost all the B+. The 22k B+ dropping resistor also has voltage across it. Had to desolder all the wires from the tube sockets to the board to rip it out. Found another cold joint on pin 7. Reflowed that, although I'm not sure it's going to fix the problem. I don't get how a bad joint on the grid would cause the anode resistor to draw current and drop most of the B+. Should I be looking elsewhere?

[EtherealWidow]

So after unwiring the board and taking a good holiday break from it before rewiring it back up, I'm still having the same problem. Reflowed any suspicious looking joints on the PI socket and I'm still dropping about 192VDC on the same anode resistor with the tube plugged in or unplugged. Chopsticking makes no difference either. I'm stumped.

[EtherealWidow]

Just found something else interesting. The .02uF power tube coupling caps. The one that's connected to the Ra with the messed up voltage has only 45VDC across it while the other has 175VDC across it. The 175VDC seems more correct to me because one end is connected to an anode and the other end at ground potential. Is the one with 45VDC across it leaking by chance?

[matt h]

(deleted)

[M Fowler]

Here is Ceriatone's Spitfire voltage chart

[EtherealWidow]

Alright. After a very busy work week and the holidays finally being over, I finally got around to putting some very long 18ga solid core leads on another 9 pin socket so that I can measure the voltages on the pins without having to rip up the board again. Here they are.

V1 (parallel triode gain stage)

1: 129VDC
2: 0VDC
3: 1.03VDC
6: 129VDC
7: 0VDC
8: 1.03VDC

V2 (PI)

1: 173VDC
2: 13.6VDC!!!
3: 31.4VDC
6: 46VDC
7: 27.6VDC!!!
8: 31.4VDC

V3/4 (EL84's) voltages were close enough to just bundle them together

1: .1VDC
3: 9.6VDC
7: 295VDC
9: 298VDC

[matt h]

(deleted)

[EtherealWidow]

I know that the grids get drawn down by the DMM, which I would imagine makes it even worse! I mean... if 13VDC at the least is getting drawn down...

I don't even know what to look for now. I've pulled up this board so many times and had to rewire it. I know that the PI has been a huge problem for this entire voyage. It's a very compact chassis, so I don't doubt that something is shorting out against something else. Looks like I'm going to pull up the board one last time and bend the tube pins away from each other in hopes that it will be fixed. My friend didn't build it, rather, I'm building it for my friend. The PI layout looks fine, according to the ceriatone one. I guess over the next couple days I'll just try and get this done and do another update then.