JCM 900 2100 w/ potential blown PT

[jamesmafyew]

It's an autoranging DMM and does read negative volts DC without switching the leads. I've used it to check lots of other amps' voltages. The possibilities I've reasoned are that 1) the series 47nF cap is open, or that there's an issue with the two 10uF electrolytics so that there's just half-rectified AC on the grids instead of a constant DC voltage. The meter reads a steady 0VDC and has a meaninglessly fluctuating VAC value.

[Chuck H]

Ok.
Check for DC on both legs of C15. If you have no voltage there then there's no voltage getting to the circuit for some reason.

Chuck

[Prairie Dawg]

That, or decomposing garlic. They were quite common at one time in the early to middle fifties and between the Sarkes Tarzian folks and the Federal folks they made a lot of them. They also put off a lot of heat which is why they have the fins.

Selenium is nasty stuff. Best to avoid it whenever possible. It's found in fairly high levels in the soil of some western states and concentrates in sinks where irrigation runoff concentrates. Excellent book by Tom Harris called "Death in the marsh" describes the process quite nicely.

[jamesmafyew]

Should there be DC there, since that's before the diode? I would have assumed there would be AC before and after that cap.

[g1]

You should have AC voltage on both sides of C15. It is probably open. Have seen some with 250V cap. there, recommend 400 or 600V.
Possible scenario: cap. opened, no bias, output valve fuse blew, someone tin-foiled or overfused output fuse, blew power transformer.
Sound far fetched? Yeah, no one ever puts in bigger fuses or foil!
I only mention it because I have seen it in rental JCM900's.

[J M Fahey]

1) I refer to orientation on the schematic, the actual part on PCB might be pointing anywhere.
2) Measure AC voltage on the right and left pins of C15.
On right should have close to 340VAC;
if you don't, follow the track which should feed it;
on the left one, probably around 50 VAC; if you don't, maybe C15 is open or D1 is shorted (diode test it).
If you doubt C15 parallel it with a similar capacitor.
2) Diode test D1. then switch to DC, measure on both ends of D1.
On right your meter should indicate some "DC"; it really is AC but quite unsymmetrical, which means it has a "DC component". A needle meter works best here.
On left you should have around -40V; if you don't either D1 is open or C14 is shorted.
That's enough for today; good luck.
PS: do the diode test first, with amp off and unplugged and all capacitors discharged.

[Chuck H]

+++ AC not DC. Caps don't even pass DC. Just thinking/typing too fast.

Chuck

[jamesmafyew]

Thanks all. This has been enormously helpful.

C15 was indeed open. I replaced it and now I get something resembling the appropriate bias voltage on the grids (around -40V). I'm still working with the amp with no tubes in it though, because I'm seeing that the main supply caps charge up even though I'm not bringing the amp out of standby --- so there is high voltage on the plates and screens even with the standby off. It takes about 30 secs for the HT to reach 490V. As the caps charge up, I see the bias voltage drift down.

I don't think this is normal... or is it? Am I seeing half-wave rectified DC on the plates from the one side of the HT tap that's connected? I don't want to put tubes in the amp without knowing whether this is normal... or if there's something else I need to be tracking down.

[J M Fahey]

Hi James.
*if* you see around -40V biasvoltage on output grids (that's normal, that's what you expect to have there) and up to +490V on all plates, that's normal too, because there is practically no load at this point.
When you put tubes there it will calm down to about +450V .
Put all tubes there, hook a speaker. put every control on "0" and turn it on.
Watch carefully for sparks, arcs, smoke or redplating for about a minute or two and then measure that everything is OK.
Why not measure first?
I'm afraid you'll be with your eyes glued to the meter and ignore, say, redplating.
I would not trust the old tubes, use known good ones, at least some pulled from another similar head.
PS: anyway, I *would* have another spare head, just in case. (Even an SS one).
You do not take chances when Touring, period!!

[jamesmafyew]

Thanks JM. Agreed about touring --- this guy has an amp to use because he's using MY backup. I've had this happen before and it was no fun executing a repair with only a Radio Shack to run to. I learned my lesson, hopefully this guy has as well.

To be clear about what I was asking before: should the B+ be present at all if the standby switch is off? I would have thought from the schematic that it wouldn't. It appears slowly over about a minute with the amp in standby (as opposed to immediately when I flip the standby switch to "on"). But if that's normal, then otherwise everything checks out. I'll throw the spare tubes in and see what happens.

[g1]

Wondering if the plate voltage in standby mode could be due not having tubes installed. Maybe it will fall off appreciably with the power tubes in. Can you check the plate voltage again (standby) with tubes in? Also curious to know what the AC voltage is across C15 in standby with tubes in. By the way, what was the voltage rating of the cap. which failed?
If any one would care to comment, why is class x specified for C15?

[jamesmafyew]

After putting tubes in, the amp appears to be fine. Nothing blew up and sound comes out. THANKS for all the help!

To address each of g-one's questions:

1. Probably. It dropped to about +120V on the plates when in standby after putting tubes in. While in standby WITH tubes, the bias measures a constant -62V and does not drift down over time. As soon as the amp comes out of standby, everything goes to exactly where it should be --- Plates at +450V, screens at +440V, bias at -40V (which is what it stabilized at in standby with NO tubes).

2. With the amp in standby and tubes installed, C15 has 260VAC across it (just measured this). When I flipped the amp on, it shot up to 270VAC and then settled a little lower.

3. #2 is absurd because the cap I pulled from C15 is rated for 250VAC! The new one I put in is rated to 425VAC, 1600VDC.

4. Googling "Class X capacitor" brings up some stuff about line filtration. The takeaway from my reading there basically amounted to "needs to handle high AC voltage and transients"... which is apparent for C15 from the schematic and from measurements. The thing is, the cap looks different from all the other nonpolar caps in the amp, which makes me wonder if it failed before and was replaced. This amp has obviously had a lot of repairs done over the years, some for evidently catastrophic failures.

It seems likely that if a line voltage transient caused this whole mess in the first place, C15 would be one of the first casualties.

[MstrB]

Yup, Tweed Fenders used Selenium

[MstrB]

James, about the C15, what type of capacitor did you use and what kind was in there? The one I am working on has a Sprague Orange drop 600vdc.

[jamesmafyew]

Whatever was in there wasn't marked well except for the 250VAC rating. It was a gray rectangular film cap. As I said above, it looked different from all other caps in the amp -- so it's possible that it was replaced before (but really, who knows). I replaced it with Digikey part # P10515-ND. To report more up-to-date, the amp continued touring after the repair and has not had any further problems.