Sparks flying in power tubes...

[markmalin]

Hey gents,


I won't be able to get to this until Wednesday, but it's bugging me so thought I'd post early before I start diagnosing.


I bought a Marshall JCM 800 Lead combo "non working" (no sound).  The owner claimed it worked fine 3 months ago when he put it in storage, so I figured how bad can it be?  Here's what I've done so far:


1. Speaker cable disconnected (you never know :) so I reconnected it.  Powered the amp on, tubes lit, so I switched Standby to on.  No sound.
2. Checked fuses, the "HT" fuse had no continuity, so I replaced it with a spare 1 Amp (non-slow blow) fuse.  Powered it up, heaters looked like they were working.  Switched Standby to On, heard a pop and the HT fuse blew.
3. I thought maybe it needs a slow-blow, so I replaced it with a 2A slow blow I had on hand.  Powered the amp up, switched Standby to on and heard loud pops and saw sparks bouncing around inside one of the power tubes.  Switched to standby immediately...then one more quick switch from Standby to On and again, sparks. (HT fuse still good)


My plan is to pull the chassis and get it on my bench when I get home Wednesday night.  I thought I'd start with pulling the tubes and powering it up to read some voltages, then start from there.  I'd appreciate any thoughts on how to approach this.  Basically as soon as you take it off standby you get sparks.  I haven't worked on a Marhsall before, it seems from the internet this is a common ailment with these, but would appreciate anyone's input before I start.


As always,
humbly,
Mark.

[HotBluePlates]

I'd just pull the output tubes, put in the rated fuse (I thought 500mA was typical for these?), and try powering on.

Sparks in the tube sounds like arcing, and could be specific to the tube itself. I personally wouldn't invest the time pulling the chassis out of the cabinet unless I ruled out issues (like bad tubes) which can be solved externally.

That said, there may be an issue causing the fuse to blow (shorted cap, B+ short, etc), but it would be nice to know if the tubes are associated with the fuse-popping.

[markmalin]

Sounds good.  Thanks, I'll give that a try first.


Mark

I'd just pull the output tubes, put in the rated fuse (I thought 500mA was typical for these?), and try powering on.

Sparks in the tube sounds like arcing, and could be specific to the tube itself. I personally wouldn't invest the time pulling the chassis out of the cabinet unless I ruled out issues (like bad tubes) which can be solved externally.

That said, there may be an issue causing the fuse to blow (shorted cap, B+ short, etc), but it would be nice to know if the tubes are associated with the fuse-popping.

[jjasilli]

Take voltage readings with no tubes IN.  Make sure there is bias voltage.  LO or NO bias voltage will give you spectacular pyrotechnics inside your power tubes, for a second or two.  You may have already guessed that I found this out the hard way. 

[markmalin]

Thanks, jjasilli.  That's good to know!

[Willabe]

I would pull the chassis on an amp I just bought, even if it was working fine.

Like JJ wrote to check bias and to see if anything looks like it's on it's way out, like filter caps. Also to see if anybody was in there before and hacked/boogered something. You never know unless you look.

 

[Willabe]

3. I thought maybe it needs a slow-blow, so I replaced it with a 2A slow blow I had on hand.  Powered the amp up, switched Standby to on and heard loud pops and saw sparks bouncing around inside one of the power tubes.  Switched to standby immediately...then one more quick switch from Standby to On and again, sparks. (HT fuse still good)

This is dangerous to randomly pick a different value and switch to a slo-blo fuse when you don't know what's supposed to be in there, when you already know it's blowing fuses. 

And not be a jerk but why would you flip the stand by switch back on after you saw sparks flying inside 1 of the power tubes??????? It's not going to magically fix it self after turning it off for a few seconds.   

[jjasilli]

why would you flip the stand by switch back on after you saw sparks flying inside 1 of the power tubes???????

That's an easy one. . . to watch the show in the other tube!   

[markmalin]

This is dangerous to randomly pick a different value and switch to a slo-blo fuse when you don't know what's supposed to be in there, when you already know it's blowing fuses. 

It wasn't entirely random.  The amp says 1A for the HT fuse.  I put a 1A in there and it blew, so one possibility was that it required a slow blow.  I didn't have any 1A slow blows on hand, so I tried a 2A slow-blow figuring if it's going to blow a 1A, it would blow the 2A.  Was that a wise thing to do?  No.  Stupid, but not random.

And not be a jerk but why would you flip the stand by switch back on after you saw sparks flying inside 1 of the power tubes??????? It's not going to magically fix it self after turning it off for a few seconds.

because I'm an idiot sometimes...and I get impatient.  I was tempted to not add that self incriminating detail to the original post, but there it is.

[Willabe]

Slo-blo fuses used to with stand any inrush current at turn on, that's why they are used on PT's. When a tube amp PT is turned on there's an in rush of current because the tube filaments are cold and draw ~4 to 5 times (?) their normal current for a few seconds. Then as they warm up they draw less and less current until their warmed up and draw their normal current.

And at 1st turn on the filter caps will cause an inrush current draw on the B+ wind until their fully charged up.

Fast acting fuses are normally used for the OT primary fuse. Because there's no in rush current spike there.       

[HotBluePlates]

My suggestion for power-up with no tubes, chassis in the cabinet, arises from this experience:

I had some nice RCA blackplate 6L6GC's I'd been using for a while in a Fender amp. One day I flip the amp from Standby to Play and the main fuses pops. Scratching my head, I put another fuse in, turn to Play, fuse pops. I pull the output tubes, install a new fuse, everything seems fine. On a hunch, I test continuity from pin 2 to pin 3 (heater to plate) on the 6L6's, and one of them is shorted (internally).

I threw that one tube in the trash, popped in a new 6L6 in its place, and the amp worked fine from then on. Sucks when you have an expensive output tube crap out like that...

[PRR]

Sparks are fun.

The only real question: YouTube link??

> figuring if it's going to blow a 1A, it would blow the 2A.

What if the load is 1.9 Amps?

2.0 Amp fuse will sit there forever.

Meanwhile the transformers go up in smoke. Done that, got no T-shirt.

You did, in this case, "prove" the short is more than 2 Amps.

If this is the 50-Watt two-big-bottle series, "normal max" B+ is maybe 0.3A. If 100W, 0.6A. 1.0A is a proper size.

> continuity from pin 2 to pin 3 (heater to plate)

I had short G2 to K. Not the same problem, but (if you persist) makes the same smoke.

How much could it be? One side of OT is, what? Say 50 Ohms. But the PT/rects have impedance also. Say 100 Ohms total. B+ is say 400V. 400V/100r is 4 Amps. Yeah, that'll blow 1A/2A fuses pretty quick. Keep going to 4A/5A, you have 1,600 Watts in a contraption which is good for 100W-200W of juice flowing through.

[Ritchie200]

Having worked for a fuse company and personally involved in the 1/4" and 5mm glass (and other) tube design and manufacturing, let me add this: One misconception is that any current draw over the rated value of the fuse will cause the fuse to blow. Yes, and No! Yes, it will eventually open - but it may take some time. It takes about 200% over-current (to rating) for the fuse to blow instantaneously. Let's say your amp is drawing about 30% over-current. It would take an hour for the fuse to open. The performance graph ramps up fairly quickly, but there is a time factor. Does not matter if the fuse is a fast blow or slo blow, CONSTANT over-current performance is identical. Bottom line - if you blow a fuse, you have a problem! Check it out.  1/4" fuses with quality issues that would cause it to blow under the rating are basically non-existent. They are all individually tested before they get packed. This is true for any domestic supplier (Bussmann, Litttelfuse, etc) as they must meet UL, IEC, and British Standard requirements.  A Chinese no-name and all bets are off.

Jim

[Jack_Hester]

Check the tube pin sockets terminal tension.  I had a power tube to start arcing inside the glass, after using a bias adapter.  Seems the pins on my adapter spread the socket open enough for it to make poor contact.  I'm sure that it was arcing down in the socket, but I killed the power too quick to notice.  It had no issues, until I used the adapter.  I have two different adapters, and one of them has very fat pins.  That's the one that caused the issues. 

I tightened the pin sockets and re-installed the tubes.  No more arcing.  I now snug up all the pin sockets after using any adapter. 

Jack

[tubenit]

Be sure to check the tube socket IF there has been a chance or arching. You may find traces of the arching on the socket.  My understanding is that the ceramic sockets are somewhat less prone to arching.

with respect, Tubenit

[markmalin]

Thanks guys.  I'm hoping to look at it tonight - I had to order some 1A slow-blow fuses and will pick them up over lunch today.


Last night I pulled the chassis and briefly put it on my bench.  It seems "wet" inside, I assume from something like lubricating contact cleaner.  Like a faint film of oil.  Nothing "looks" amiss on first inspection, no burn spots or bubbled caps.  But you can see a faint film of lubricant or whatever on the chassis.  In some places where you lift a wire you can see where it accumulated.


Anyhow, just an observation.  Will put a fuse in the HT tonight and power it up without tubes first and check some voltages.  I'm hesitant to put those power tubes back in.  I don't have another Quad of 6550's on hand, just some old 6L6's.

[mresistor]

Is it wet? As in dripping with oil and if you place the chassis on it's side, do you see any of the wet stuff running? Would like to know just "how" wet. Does the stuff smell?  Does it smell like WD40?

[markmalin]

Is it wet? As in dripping with oil and if you place the chassis on it's side, do you see any of the wet stuff running? Would like to know just "how" wet. Does the stuff smell?  Does it smell like WD40?

just lightly oily.  You can see the oil especially where wires touch the chassis (like it wets between the wire insulation and the chassis).  But it's not like really wet.  It feels like WD40.  I'm guessing someone sprayed fader cleaner or something into all the tube sockets

[markmalin]

Here's an update, gents.

Chassis on bench, tubes out, inspected the sockets (no signs of arcing), installed correct 1A slow-blow fuse and 4A slow-blow, applied power via my variac/isolation transformer, turned it on and turned standby off and nothing out of the ordinary.  440 V DC on the 6550 plates, about -56 for bias voltage.  Seemed normal.  Powered it off.

I next installed the pre-amp tubes and powered it up again (on for power switch, on for standby switch).  No sparks (obviously because the power tubes were not installed).  Powered it back off.

I measured continuity between pins 2 and 3 on the power tubes, no shorts there.

Originally it was the two inner tubes of the quad that were sparking.  So I retensioned the power tube sockets, then installed the two outter power tubes.  Powered up the amp, left it in standby and watched the heaters glow.  Took it off standby and it seemed fine.  No sparks.  Powered it back off.

I then put the two outer tubes in the inner sockets, powered it up as above, no sparks.  Turned it back off.

After this I verified the socket pin tensions looked good, installed all 4 power tubes, turned the amp on and left it in standby for a few seconds, switched to on then off, no sparks.  Turned it to on again and left it on, no sparks.  Turned the amp back off.

I then plugged a guitar in, brought the amp back up watching the power tubes.  Turned the volume up slightly and actually got sound.  In about 10 seconds, though, the one inner tube started to light up inside.  No sparks, just started getting a crackling sound and I could see that one tube flickering blue.  I turned the power off at that point.

Best guess is these are the original GE tubes.  I have an old set of 4 6L6's from a Twin I restored years ago.  They tested poorly, possibly grid leak on one.  I'm wondering if I should:
1.) try that set of 6L6's to see if I can run the amp for a while without any issues
2.) try changing the order of the power tubes and see if the badness follows the tube or the socket.

Any other thoughts?  I don't want to buy a new set of 6550's and risk damaging them until I know the amp is OK.  At this point I'm thinking either the socket pins were loose, or the tubes are getting flakey.

Mark.

[Willabe]

....... installed correct 1A slow-blow fuse and 4A slow-blow, ........

The HT 1 amp fuse should be a fast acting type, not a Slo-Blo type.

[shooter]

I could see that one tube flickering blue

I would move it, I would also monitor the bias on the *pair* that is *suspect*, both before and after you move it.  If it follows the tube, walla, put in the 6L6s, re-bias, and see/hear.   I would also have all my meters n scope probes hooked up so I can *see* if things look dicey. 

[Paul1453]

This is not a recommendation of what you should do, it is just what I would do.

I would test the four old tubes you have and select the best two to use for testing.

I would use these 2 tubes to replace a Push Pull pair of tubes currently in your amp.

If you swap out one pair of tubes and the problem still exists, I would put the 1st pair back in and replace the second pair with two testing tubes.

If the problem is now gone, at least one or possibly both of the second pair of tubes is bad.

OK, so let's say the amp is now working properly with one set of old tubes and one set of new tubes.

Now I would take one of the tubes from the pair that when they were removed the problem went away, and replace one of your old tubes with that tube.

If the problem returns, then obviously that tube is no good.  Trash it!

I would now put the old tube back in and verify that the pair of old tubes still works properly with the good pair of new tubes.

So let's say the amp works properly again.  Now I would take your remaining new tube and replace the same old tube you did last time.

If your amp still works you have identified and disposed of your problem tube.

If the problem returns, I would pull that tube and put the old pair back in to verify the amps still works with the pair of old tubes and new tubes.

If it works again, then trash the other new tube as you have identified them as both bad.   

If it doesn't work now, the second tube has caused one of your old tubes to go bad.

I return to your old tube stash and take the third best tube and pair it with your other testing tube that hasn't been paired with the bad tubes you identified.

If it works properly again now, OK.  Play it that way if cash is tight, or get a new pair of tubes if a little money is available, or get a whole new quad set if money is not an issue.   

Does my testing procedure make sense to you?   

[PRR]

> tube started to light up inside.

On the inner surface of the glass?

Or the "air space" (should be vacuum) around and between the tube elements?

Glass-glow is a known thing and usually harmless. Stray electrons excite odd molecules of glass and glow.

"Gas glow" in the middle of the tube means the tube is gassy. This WILL lead to major problems, including burned-up transformers.

Tube testers will sometimes find a VERY-gassy tube. But a tube can be "OK" at 150V warm test, only liberate its gas when HOT and HIGH-voltage.

I'd try it once more in a different socket, then deface that tube and set it aside for disposal.

[markmalin]

It was in the air space, I'm pretty sure.  Unnatural.  Not the typical inner surface of the glass glowing purple, but intermittent bright flashes of purple inside

> tube started to light up inside.

On the inner surface of the glass?

Or the "air space" (should be vacuum) around and between the tube elements?

Glass-glow is a known thing and usually harmless. Stray electrons excite odd molecules of glass and glow.

"Gas glow" in the middle of the tube means the tube is gassy. This WILL lead to major problems, including burned-up transformers.

Tube testers will sometimes find a VERY-gassy tube. But a tube can be "OK" at 150V warm test, only liberate its gas when HOT and HIGH-voltage.

I'd try it once more in a different socket, then deface that tube and set it aside for disposal.

[markmalin]

The 80's Marshall JCM 800 100W Lead schematic calls for a type "T" on both fuses, which is a slow-blow. "T1A"

....... installed correct 1A slow-blow fuse and 4A slow-blow, ........

The HT 1 amp fuse should be a fast acting type, not a Slo-Blo type.

[markmalin]

I'm going to move forward on the assumption it's just tubes.  Originally one of the two inner tubes was sparking.  I moved that one from the inner socket to the outer socket and it still sparked.  I marked that one bad.


For the heck of it I populated only the two outer sockets with the pair originally in the outer socket and left the two inner's empty.  After about 10 seconds one of those actually sparked, but rather than conclude there's something amiss in the amp, being that these are the original tubes (1984) and one is flakey, I'm going to assume they are all somewhat flakey and just pick up another set and move on from there.  I can't really come to a solid conclusion with likely bad tubes.


To me the real test was that I installed a quad (albeit mismatched) of 6L6GC's I had laying around.  I let the amp run for several minutes with no unusual activity and was able to play it.  With the old tubes, within seconds I'd get sparks inside the tube.  I ordered a set of 6550's from CE and will install and bias them and hopefully that will be that.  The fact that this 6L6 set works fine leads me to believe it's the tubes.  We shall see.  If I get sparks in the new set, I'll cross that bridge when I get there.

[Ritchie200]

With two tubes you will also have higher voltages all around.  That may be what pushed the two "good" old ones over the edge.  However, if the tubes are that old, you don't know how bad they have been flogged over the years in countless hair bands, it's time to replace them anyway.  If you got it for a good price, this sounds like a great deal!


Jim

[markmalin]

That makes sense.  It's dated 1984, so I'm sure they've been put through their paces.


I actually picked it up to flip, just to generate some income for an amp project.  I picked it up pretty cheap given what it is, so hope to turn it around and make a few 100.  I'm more of a jazz/blues guy.  Granted, in the 70's I played "hard rock" and used a 50W half stack...nothing but a Marshall and a Les Paul could make "that sound".

With two tubes you will also have higher voltages all around.  That may be what pushed the two "good" old ones over the edge.  However, if the tubes are that old, you don't know how bad they have been flogged over the years in countless hair bands, it's time to replace them anyway.  If you got it for a good price, this sounds like a great deal!

Jim