Tek 067-0502-01 capacitor explosion!!

[Addicted2AnalogTek]

Yesterday I noticed that there was some noise on the output of my 067-0502-01 amplitude calibrator, so this evening I decided to open it up and have a look... and found a rather fascinating exploded capacitor. This popped my cherry for "fantastic capacitor failure"!  Time to load up the Mouser cart with some more parts!!

Some history:

I bought this item about 3 years ago in very good condition and fully functional, so I left it alone and shelved it for use.  I recall opening it for inspection when I received it, and it looked clean and fully intact. I used it for several months after first receiving it, and then it sat for about 2 years.  I would power some gear on and let it warm up for an hour or so before using it, so it's no surprise that I didn't hear the little firecracker event happen inside my calibrator.

So here she be!

[bd139]

That looks like it went with quite a bang! No vents in them!

I’m genuinely surprised those capacitors last as long as they have. About 5 years ago I got my hands on a Tek 321 which hadn’t been powered up for 30 years and it was full of that variety. It worked fine. Nearly fell off my chair.

[Addicted2AnalogTek]

Oh yeah, it must've been quite a bang! I wish I would have been in the room to hear it!!  It exploded with enough force to break the neighboring ceramic strip, which has diodes with thick leads soldered into it!  (see pic) 

The exploded capacitor was the filter capacitor (C312  --  4uf 200V @ 85C) for the +100V rail, so it's no wonder I was seeing lots of noise on the output. The only remaining capacitor on that rail is a 5uf 150V decoupling capacitor.   

I happen to have a small supply of 5uf 150V Sprague 30D caps, so I'm going to solder one in there and replace the two unhealthy 100uf 12V and 100uf 15V capacitors with 100uF 35V Sprague 516D capacitors for the time being.   Once I figure out exactly which replacements I'm going to use (preferably 105C or 125C rated), I'll post the info.   I'm just going to go for the (nearly) wholesale lytic replacement in this thing since it's in amazing condition and I'd like to preserve it as much as possible.

Capacitors to replace:

C118    2uf 150V     (using sprague 30D replacement)
C312   4uf 200V      (The SM parts list says 4uf 35V...  )
C345   5uf 150V      (Using sprague 30D replacements)
C367   100uf 12V 
C448   5uf 150V       (Using sprague 30D)
C457   100uf 15V

The only one that really NEEDS to be replaced is C312, but the ESR on the remaining ones is near double that of NOS parts (2011 date code), although they measure good on the capacitance meter.  C448 is also browned from heat.   There's two 17uf 150V caps that are seem perfectly healthy, which is no surprise to me. I've yet to run across a failed one of those. 

There's also a 4W precision resistor that has started to brown from heat, so I'll be checking into that as well. I'm going to have to trace out the circuit a bit to confirm which one it is, but I think it's R450 (2.09K) , which is one of the voltage divider resistors for the +100V floating power supply. It should still be healthy enough to keep.

[bd139]

That's quite a bit of damage from a cap. Not sure I'd have liked to have been in the room!

I had a quick look and they still sell the 30D series caps apparently (well Vishay do). Quite impressive considering the dates I've seen on some!

Good luck on the restoration. Let us know how it goes

[Addicted2AnalogTek]

I try to use 30D caps where I've found 30D caps (or same type, diff manufacturer), I'll just increase the voltage rating until the case sizes are similar to to the old cap.  They're just rather pricey caps, typically running at $4-$6 each new from Mouser. Knowing I'll most likely never have to replace them again is enough to make me purchase them, though.

I'm not anticipating that I'll have much to do inside of this unit since it was fully functional, but I'll of course update the thread as things progress.

[Addicted2AnalogTek]

Here's the link to the TekWiki site with the service manual PDF:

http://w140.com/tekwiki/wiki/067-0502-01

[The Soulman]

Be sure to check the rectifier diodes before switching it on again.
One could have easily failed due to the capacitor, going low impedance/short circuit drawing great amount of power and explode.
Or the other way around, diode failed (short) first and that cause the capacitor to explode.

[Addicted2AnalogTek]

Be sure to check the rectifier diodes before switching it on again.
One could have easily failed due to the capacitor, going low impedance/short circuit drawing great amount of power and explode.
Or the other way around, diode failed (short) first and that cause the capacitor to explode.

Thanks, I'll definitely check them.  The unit actually still functioned properly after the capacitor had exploded. I must have ran it for a few hours at least after the cap had checked out and didn't notice the noise on the output until I was running a calibration procedure on a 465, so I'm confident that the rectifier diodes are still good. I'll be checking them anyway as a preventative measure, although I'm not going to lift the leads on any of them unless the results of in-circuit testing are suspect.  I'm trying to minimize the risk of lifting the silver lining out of the ceramic strips, as I already have one 13 position strip to source and replace.

I think that the capacitor was just showing it's age, and possibly let some moisture in around the cracking rubber end plug.

[Jwillis]

Learning from experience ,capacitors can pack quite a punch.Last year I had one a little bit bigger explode and hit me in the face just below the eye ball.Punched a hole right into my sinus and took 3 subsurface stitches and 6 surface stitches to plug the leak.Doctor said 5 millimeters higher and I would have been permanently blind in that eye.So anyway I have permanent night blindness in that eye and a peripheral blind spot in light.Not to mention I had to pick up shrapnel for weeks later.Gets into everything.
I can't emphasize the potential danger of electrolytic capacitors enough.Many older caps have no scoring on the top so they go off like projectiles.Newer ones are scored and in theory don't become airborne.I now wear eye protection or remote start whenever I fire up a prototype circuit or open chassis.
Be careful folks and don't do what I did. I was lucky some what.   

[Gyro]

When I was a youngster, I put a dent in underside of the steel chassis of a Garrard SP25 turntable with an exploding overvolted large can 400V electrolytic. 

I still don't know how I hid the smell of the mass Aluminium foil and electrolyte soaked separator that was draped all over the hot valves of the amp from my parents!

[Addicted2AnalogTek]

Learning from experience ,capacitors can pack quite a punch.Last year I had one a little bit bigger explode and hit me in the face just below the eye ball.Punched a hole right into my sinus and took 3 subsurface stitches and 6 surface stitches to plug the leak.Doctor said 5 millimeters higher and I would have been permanently blind in that eye.So anyway I have permanent night blindness in that eye and a peripheral blind spot in light.Not to mention I had to pick up shrapnel for weeks later.Gets into everything.
I can't emphasize the potential danger of electrolytic capacitors enough.Many older caps have no scoring on the top so they go off like projectiles.Newer ones are scored and in theory don't become airborne.I now wear eye protection or remote start whenever I fire up a prototype circuit or open chassis.
Be careful folks and don't do what I did. I was lucky some what.

  WOW!  Good thing it didn't hit your eye! I can see the potential for more damaging effects than blinding..   I'll start making use of my safety glasses that are hanging on the rack and covered in dust... From the sound of it, I should probably just invest in a face shield for those risky situations!   All the more reason for wholesale recapping of old electrolytic caps, too...

Update:

All of the rectifier diodes test good with a DMM while in-circuit.

All of the in-circuit measurable 5% carbon comp resistors are well within tolerance. (The majority are 10%, so I'm going to consider them good as well)

C312 replaced with 5uf 150V 30D (+100V rail filter cap that went boom)
C448 replaced with 5uf 150V 30D (original was brown, moderately high ESR)
C347 replaced with 5uf 150V 30D (original was decoupling for +100V rail)

I'm leaving the remaining capacitors as-is, at least until I have preferred 105C or better replacements. (I'll be wearing safety glasses when I juice this bad-boy up....)

[Addicted2AnalogTek]

So far, so good.

+300V Floating supply measures +299.3V
+125V Floating supply measures +125.1V
+100V Floating supply measures +99.8V
-100V supply measures -102.7V
+100V reference supply measures +98.0V

the -100V supply is spec'd at 2mV and I'm reading 2mV (how bout that, might need to replace a 17uf 250V cap after all..)
the +100V reference supply is spec'd at 30mV and I'm reading 20mV
--Used 7A26 + P6101 probe for ripple measurements--

My diff amp for the 7K isn't currently working, so I can't safely and properly measure the ripple on the floating supplies.


Output amplitudes are correct, however the output is still a bit rough. Especially so without bandwidth limiting turned on, and even more so on the digital scope.  It'll suit my needs for the moment, but I'm going to go ahead and order replacements for the remaining capacitors and see if I can figure out what's wrong with my 7A13 - without having a 7K extender. 

EDIT:
I forgot that the vertical output amplifier IC in my 7854 has an intermittent internal connection and needs to be replaced. The abnormalities of the waveform from that scope are likely from that fault, not a fault with the amplitude calibrator.  The Rigol DS1054Z also enjoys screwing with me, so most likely everything is A-OK inside of the '0502'.   Adding a screenshot from the Rigol just for kicks:

'0502' set at 20mV square wave