Live voltage on a capacitor can - why?

[lasseo]

It is not broken, it seems very much intentional. Reasonable amount of effort seems to have been spent on connecting the capacitor minus leg to the protective canister. What's the benefit of that? A stab at reducing EMI?

Here's a funky looking Sprague capacitor off a Tektronix 2213 oscilloscope power supply where exactly this has been done, complete with warnings about dangerous live voltage stamped on the can. A mounting ring with solder lugs at the bottom of the case is soldered to PCB in three locations around the bottom, and a strip soldered to the - (minus) pin of the capacitor. The + pin is connected to the positive rail. It appears to be a filtering cap, connected right after a bridge rectifier feeding a switch mode power supply. This scope was first released in the early 1980's, so it would have been designed in the late 70's. That's not old enough to be hangover from days when multiple capacitors were stuffed in a single can. It also was a product that was designed to a price point, so going away from standard parts would have been done out of necessity; the bean counters would have seen to that.

So, to what end would anyone want to do this?

[lasseo]

I'll post some photos once I can, for some reason I can't seem to upload them right now. Yes they are below 4 MB in size.

[lasseo]

[tggzzz]

It is not broken, it seems very much intentional. Reasonable amount of effort seems to have been spent on connecting the capacitor minus leg to the protective canister. What's the benefit of that? A stab at reducing EMI?

Here's a funky looking Sprague capacitor off a Tektronix 2213 oscilloscope power supply where exactly this has been done, complete with warnings about dangerous live voltage stamped on the can. A mounting ring with solder lugs at the bottom of the case is soldered to PCB in three locations around the bottom, and a strip soldered to the - (minus) pin of the capacitor. The + pin is connected to the positive rail. It appears to be a filtering cap, connected right after a bridge rectifier feeding a switch mode power supply. This scope was first released in the early 1980's, so I would have been designed in the late 70's. That's not old enough to be hangover from days when multiple capacitors were stuffed in a single can. It also was a product that was designed to a price point, so going away from standard parts would have been done out of necessity; the bean counters would have seen to that.

So, to what end would anyone want to do this?

Your first port of call is to get the service manual (from any of the usual places), identify the capacitor, and locate it on the schematic.

That will show you the all-important context, and should give big hints about why it is connected that way.

Feel free to post the relevant bits of the schematic, and ask further questions.

[tggzzz]

(Attachment Link)

I don't see any warnings about dangerous live voltages on the can.

As with all capacitors, I do see a statement of the maximum rated voltage across the capacitor's pins.

[ArdWar]

That's not old enough to be hangover from days when multiple capacitors were stuffed in a single can.

While this particular one isn't a multiple caps it's definitely manufactured in similar form factor. CAN NEG is quite common for that style (originally/usually COM NEG or CAN COM NEG).

[coppercone2]

I don't get it. It looks like a RF ground for the capacitor can. That seems normal.

Are you saying that tab is connected to mains and the pin to ground? If its a +- rail, then pos would be ground

[SeanB]

Most electrolytics have the can connected via the electrolyte to the negative end. For axial capacitors there is a hard connection via the internal tab of the negative foil, but for radial capacitors this is mostly incidental. For this one the manufacturer decided to bring out the negative lead separate from the case, likely because they were reusing an old 3 terminal leadout gasket and phenolic base, and wanted to not have a long tail that was folded into the capacitor, which reduces the volume internally available for the paper and foils, and thus decided to have the extra volume filled, and use short tails to connect. However the OEM wanted a capacitor with case connected to negative, so they simply went and modified the outer washer used on the crimp, there to provide an extra layer of rigidity to the seal during closing, and which also has the case mounting pins or tabs on the steel. So now it has a tab that is soldered to the negative pin, so connecting to firmly to the aluminium case, but the small steel tab also connects negative to the 3 mounting pins as well. After sealing they simply pressed the tab down over the small length of pin in the insulation, soldered it down, then trimmed it off, looking like it was ground off automatically with a small carbide wheel.

Just a decision made by the OEM, on order from the buyer, who wanted a semi custom capacitor with high capacitance, high voltage withstand, but in a smaller can than standard for that value and voltage, and also wanted to have lower ESR and higher ripple current spec than standard. Sprague was quite willing to do semi custom orders, especially for the manufacturers who had large exclusive order volumes with them, as this also menat a large degree of vendor lock in, and as well the OEM had a now non standard part they could sell into the repair market, knowing they could charge a premium for a part that was, to them, a very low cost item.

[CaptDon]

Yes, can as negative is very common! As for voltage warning, That capacitor could be part of a line operated voltage doubler circuit common in older 'line operated transformerless' television sets. Usually the 'hot' can had a cardboard sleeve around it and these were also classed as a 'hot chassis' television. The capacitor was in fact at line A.C. voltage as well as a D.C. voltage referenced to chassis. Those 'cans' were also usually mounted on a bakelite wafer to insulate them from the chassis. A very interesting thing about the 'series capacitor voltage doubler' on the Sylvania / G.T.E. D19 chassis is that all of the line current to operate the T.V. was going through the capacitor and being of a very sturdy construction with a vent hole in the bottom which got occluded from crystallized dry electrolyte would cause the can to explode violently due to heat from E.S.R.! I have seen them snap the neck off of the C.R.T. and a saw one where the can broke near the bottom flange and shot the empty can straight up and completely through the plastic back of the television cabinet!! The owner (a family friend B.T.W.) carried in the T.V. and the 'missile' wondering what in the hell happened and would it happen again? He said the bang was so loud he nearly crapped his pants and was afraid to go near the T.V. due to the smoke and stink!

[lasseo]

I don't see any warnings about dangerous live voltages on the can.

As with all capacitors, I do see a statement of the maximum rated voltage across the capacitor's pins.

You are right of course. I must have been staring at the high voltage warnings on the main board for so long that the component specs on the side of the can read like "Danger Will Robinson!" in my head. 

[lasseo]

Your first port of call is to get the service manual (from any of the usual places), identify the capacitor, and locate it on the schematic.

That will show you the all-important context, and should give big hints about why it is connected that way.

Feel free to post the relevant bits of the schematic, and ask further questions.

Yes Sir. Unfortunately, none of the usual suspects seem to have scanned the portion where the schematic for this sub-board would be shown. It is, however, shown in the block diagram of the scope, and the service manual states that either the current limiting board is fitted, or the preregulator board that my scope has, but not both. I've highlighted the section in the excerpt of the block diagram:



Another clue comes from a field service bulletin that describes retrofitting power supply improvement ("Option 48") to existing equipment, to "make it less sensitive to power line transients". That makes sense, at 6.1 kg this is a relatively light scope that service personnel was meant to be able to take with them, so power line quality may have been compromised. Looks like mine came fitted with that option straight from the factory.

[lasseo]

Here's a pretty good photo of the location. The Sprague is located at the very back, lurking in between the two pink connectors can be seen at the top of the image on rkagerer's scope. The preregulator board is installed above the main board, with the components hanging down:

https://www.eevblog.com/forum/repair/tektronix-2213-repair-leaked-capacitor/msg2685186/#msg2685186

As you can see it sits relatively close to a transformer. According to the schematic, it's T942, which is "Transformer,RF:driver saturating pot core". I don't know what that is, but maybe that could be something that the cap has to be protected from like coppercone2 suggested?

[coppercone2]

if it saturates that means that the field is at maximum. I think they leak more and more when they get saturated because its not like the magnetic field disappears, it just can't go through the core anymore

Like you increase current past saturation... its still increasing the current through the loop right? then its like a over heating air core inductor I think.

[lasseo]

Just a decision made by the OEM, on order from the buyer, who wanted a semi custom capacitor with high capacitance, high voltage withstand, but in a smaller can than standard for that value and voltage, and also wanted to have lower ESR and higher ripple current spec than standard. Sprague was quite willing to do semi custom orders, especially for the manufacturers who had large exclusive order volumes with them, as this also menat a large degree of vendor lock in, and as well the OEM had a now non standard part they could sell into the repair market, knowing they could charge a premium for a part that was, to them, a very low cost item.

This explanation seems quite plausible. As you can see it's pretty tight on the board, with the can very close to the rectifier bridge:

[lasseo]

The question follows, should I try to replicate the set-up when substituting the Sprague for a modern capacitor? Of the reputable suppliers, at least Chemi-Con and Rubycon have suitable replacements, and the can size is considerably smaller than this one. But how best to go about it.

First option, I could conceivably take a leaf off vintage radio enthusiasts' book and stuff the old can with a modern cap, still utilising the old can as the presumed RF shield. There would be an air gap between the old and the new can, so there is at least some risk of overheating. If I fill the gap with potting compound, I will unlikely be able to re-use the can as a shield again. That could well be an in-fifty-years-time type of problem, but still

Second, I could just recycle the phenolic base of the old cap as a socket, making for easy connections to the double-sided board, and figure out some other way to shield the cap. Suggestions how to do the shielding are most welcome.

Third, my least favourite for the time being, would be to just drop in a modern capacitor, add a couple of jumper leads and hope that the resulting EMI or noise degradation will not be noticeable in use. It's just that hoping for the best doesn't ring like a robust fix to me...

Thoughts?

[tggzzz]

As you can see it's pretty tight on the board, with the can very close to the rectifier bridge:

You probably want to replace that RIFA, before it lets everybody in the room know it needs replacing.

[lasseo]

You probably want to replace that RIFA, before it lets everybody in the room know it needs replacing.

Yes! There's four of them on that board, all in various states of developing cracks, and a fifth slightly bigger one on the main board. From what I gather the flash-bang can be quite spectacular.

[MathWizard]

This is a reminder to me to not touch anything inside mains powered stuff with hands or fingers, or gnd clips, etc. You can't just assume it's safe. At some point you might, but better check 1 or twice even on working stuff.

And with old vintage equipment, watch out for previous DIY repairs that might be dangerous.

[lasseo]

And with old vintage equipment, watch out for previous DIY repairs that might be dangerous.

Yes. I have seen some - luckily not dangerous, let's just say craftsmanship-challenged.

Curiosity got the better of me, so I opened the capacitor can. Carefully pried open the edge and removed a retainer ring, the three-pronged leg, and the phenolic base, to be treated to a whiff of something that smelled like a combination of sulfuric and citric acids.

[lasseo]

Does anyone have experience with this or a similar product? It is a "re-enterable" potting compound, which supposedly means it is possible to get components out of it if a need arises. I should think using a filler is much better thermally than leaving a couple of mm air gap between the can and the infill replacement capacitor.

https://www.raytech.it/en/product/low-voltage/fillers/gel/magic-gel

[coppercone2]

what we need is capacitors that have a line voltage switch on top. so you twist it to wind it tighter or looser to set the voltage

it might be possible

[lasseo]

what we need is capacitors that have a line voltage switch on top. so you twist it to wind it tighter or looser to set the voltage

Oohhh, a bit like the tuning capacitors on vintage radios, but with a screwdriver slot for adjustment?

[Gyro]

Does anyone have experience with this or a similar product? It is a "re-enterable" potting compound, which supposedly means it is possible to get components out of it if a need arises. I should think using a filler is much better thermally than leaving a couple of mm air gap between the can and the infill replacement capacitor.

https://www.raytech.it/en/product/low-voltage/fillers/gel/magic-gel

It's being used increasingly in outdoor mains jointing boxes. You can get them pre-filled so that the connector sinks into the top and bottom halves when clamped closed, but here's an example of using the raw fluid compound (scroll forward to 13:10 if you want to save time)...   



You could always do what the Chinese fakers do and just fill the space with dry sand. To be honest though, just inserting enough non flammable packing, or a couple of blobs of neutral cure silicone, to prevent the new cap from rattling is probably fine.

[coppercone2]

I thought the sand dampen microphonics?

I imagine the ultimate capacitor would have a damper of some kind around it, like a rubber shell.

[amyk]

Curiosity got the better of me, so I opened the capacitor can. Carefully pried open the edge and removed a retainer ring, the three-pronged leg, and the phenolic base, to be treated to a whiff of something that smelled like a combination of sulfuric and citric acids.

Likely boric acid.