What is the legal situation in North America regarding this? Is using a 630V poly considered sound practice? In Europe 250VAC X2 are mandatory. X2 as they fail open and 250VAC to have the self healing aspect working as intended.
Last edited:
My own thoughts as far as any legal situation would be that equipment must be repaired using parts approved by the manufacturer.
Service manuals will usually designate such parts with a safety warning.
Self healing caps rely on a low impedance being present (such as across the supply) in order to 'heal' correctly. In other words enough current has to be available to 'fuse' the defect and render it open circuit. Very small transformers (such as in turntables and pre amps) may actually draw insufficient current to do that and so end up with the cap actually failing short due to insufficient current to enable the 'healing' action to occur.
In practice I can't recall any 'Japanese' gear ever having a cap across a switch fail wheras I saw many hundreds of 630 and 1000v DC rated caps across 240v mains fail short circuit (explosively) in UK manufactured products back in the 80's and 90's.
(Self healing caps also pose the question of whether the transients generated by the opening switch would actually accelerate a reduction in the value of the cap as multiple instances of 'healing' occur... assuming the cap has access to sufficient current to enable the process to occur correctly)
Bottom line is use something suitable if this is for your own use, and if it is for a customer then you should use the recommended parts specified in the manual in order to be covered legally. Given that much equipment is 'ancient' using official parts isn't a realistic option and so for a paying customer you should be conversant with suitable recognised equivalents that will not cause a safety hazard or put the customer at risk should it fail.
Service manuals will usually designate such parts with a safety warning.
Self healing caps rely on a low impedance being present (such as across the supply) in order to 'heal' correctly. In other words enough current has to be available to 'fuse' the defect and render it open circuit. Very small transformers (such as in turntables and pre amps) may actually draw insufficient current to do that and so end up with the cap actually failing short due to insufficient current to enable the 'healing' action to occur.
In practice I can't recall any 'Japanese' gear ever having a cap across a switch fail wheras I saw many hundreds of 630 and 1000v DC rated caps across 240v mains fail short circuit (explosively) in UK manufactured products back in the 80's and 90's.
(Self healing caps also pose the question of whether the transients generated by the opening switch would actually accelerate a reduction in the value of the cap as multiple instances of 'healing' occur... assuming the cap has access to sufficient current to enable the process to occur correctly)
Bottom line is use something suitable if this is for your own use, and if it is for a customer then you should use the recommended parts specified in the manual in order to be covered legally. Given that much equipment is 'ancient' using official parts isn't a realistic option and so for a paying customer you should be conversant with suitable recognised equivalents that will not cause a safety hazard or put the customer at risk should it fail.
Thanks Mooly.
Do you consider those failures to have been due to the use of capacitors of inferior quality or to the methodology itself?
I would probably say parts unsuited for the actual use intended. A cap designed for 1000vdc doesn't necessarily cope with 240v AC and any transients.
It was common to see 0.1uF caps across L and N in practically every TV in the UK and the UK made models often saw these parts fail. Classic failure and less scupulous techs would just snip the offending part out. These caps were there as much as anything to prevent noise getting out of the TV rather than in.
They were typically 630 or 1000vdc rated. The Japanese sets of the day never seemed to suffer failure of similarly placed parts. I can't recall even one.
In todays world Class X would be the logical choice but bear in mind that self healing types rely on a substantial current being available to fuse the defect and render it open circuit. Each time that occurs the capacitance is reduced slightly.
This is quite a well written document on Class X and Y
https://sh.kemet.com/Lists/FileStore/EvoxRifaRFIandSMD.pdf
It was common to see 0.1uF caps across L and N in practically every TV in the UK and the UK made models often saw these parts fail. Classic failure and less scupulous techs would just snip the offending part out. These caps were there as much as anything to prevent noise getting out of the TV rather than in.
They were typically 630 or 1000vdc rated. The Japanese sets of the day never seemed to suffer failure of similarly placed parts. I can't recall even one.
In todays world Class X would be the logical choice but bear in mind that self healing types rely on a substantial current being available to fuse the defect and render it open circuit. Each time that occurs the capacitance is reduced slightly.
This is quite a well written document on Class X and Y
https://sh.kemet.com/Lists/FileStore/EvoxRifaRFIandSMD.pdf
Thanks for the reply, Mooly.
It may be useful to other amateur diy dabblers like myself to expand on my recent experience as an illustration of what can go wrong.
The switch suppression capacitor of my 1960s Rogers HG88 MkIII valve amplifier is specified as a '0.022uF, 250VAC, polyester'.
As I said earlier, it failed short after 50 years. This event fried its series 220R, 0.5W, carbon resistor, which I found to be in fragments after opening the amp.
The 3A fuse in the plug top performed its safety function, but the 1A anti-surge equipment fuse remained intact!
If nothing else, this tale reinforces the necessity of employing the appropriate fusing arrangements.
It may be useful to other amateur diy dabblers like myself to expand on my recent experience as an illustration of what can go wrong.
The switch suppression capacitor of my 1960s Rogers HG88 MkIII valve amplifier is specified as a '0.022uF, 250VAC, polyester'.
As I said earlier, it failed short after 50 years. This event fried its series 220R, 0.5W, carbon resistor, which I found to be in fragments after opening the amp.
The 3A fuse in the plug top performed its safety function, but the 1A anti-surge equipment fuse remained intact!
If nothing else, this tale reinforces the necessity of employing the appropriate fusing arrangements.
Failure modes can be very very weird... I've seen SMPS and other derivative type supplies with annihilated front ends, bridge rectifier short, semiconductors blown to pieces, print lifted off the board and yet the mains fuse is intact, conversely I've also come across supplies with a totally blasted fuse and yet the PSU and equipment works fine when the fuse is replaced and its all powered up.
Makes life interesting
Makes life interesting
As Mark mentioned earlier, Quencharc is specifically designed for arc suppression. Rifa and Kemet are some other brands that manufacture RC snubbers.
In diy, not all of us can afford the cost. If a simple capacitor is used instead, one must take into account when it fails. Appropriate safeguards should be considered.
Regards
Mike
In diy, not all of us can afford the cost. If a simple capacitor is used instead, one must take into account when it fails. Appropriate safeguards should be considered.
Regards
Mike
Attachments
Just those clear cased, long-obsolete RIFA film types were indeed crap. Other grades of similar RIFA caps would crack open too, regardless of voltage and even in high precision, high reliability applications like telephone system filter boards and their power supplies, etc. The plastic type/grade itself was obviously a major part of the problem.
Since those days, X and Y rated caps have been over-moulded or potted in polypropylene or similar opaque, high temperature plastics. Much the same inside though.
Here's an interesting tale but there are plenty more: Old RIFA capacitors and a disaster story - Page 1
Since those days, X and Y rated caps have been over-moulded or potted in polypropylene or similar opaque, high temperature plastics. Much the same inside though.
Here's an interesting tale but there are plenty more: Old RIFA capacitors and a disaster story - Page 1
Trust me.
I'm 67 now, semi retired, worked about 45 years in consumer audio/video repair shops.
I could fill your bathtub to the rim with defective Rifa caps.
And every time I spot one, it gets yanked and replaced.
I remember asking this product safety guy about MOVs, what they did, why they were in the computer power supply. After explaining to me, he said you could take a set of those and put them in your fusebox - protect your whole house from line transients. I looked at the gauge of the wires feeding the breaker box, then the little leads of the MOVs he gave me - and gave up on the idea. Perhaps one of the more valuable components in a PC powersupply discard.
I was once running a lab test using this device made by Keytek. Line to line, neutral to ground, line to ground; rinse and repeat. Getting bored, I thought "This test doesnt do anything..." Push the red button and KaBlammo! Huh, looks like we got a "failure".
Turns out an etch pattern near a standoff made a pretty series of triangular spiked points - toward the screw head. They had to re-spin the AC input board. This KeyTek device tested the effectiveness of the MOVs in the powersupply, among other things.
I was once running a lab test using this device made by Keytek. Line to line, neutral to ground, line to ground; rinse and repeat. Getting bored, I thought "This test doesnt do anything..." Push the red button and KaBlammo! Huh, looks like we got a "failure".
Turns out an etch pattern near a standoff made a pretty series of triangular spiked points - toward the screw head. They had to re-spin the AC input board. This KeyTek device tested the effectiveness of the MOVs in the powersupply, among other things.
They are not 'long obsolete', they are in current production, and they are not crap except in equipment that has lain around unwed, when they dry out and crack and let the smoke out when finally powered up. I've certainly replaced my share of them too, but in equipment being resuscitated from a period in the garage.
Datasheet: https://content.kemet.com/datasheets/KEM_F3011_PME271M_X2_275.pdf
PME271 caps are metallized paper, saturated in insulating oil. The clue as to what type of oil, lies in the spec. "self extinguishing" to a simple flame and afterburn timing test. This probably means the oil was a silicone or quite likely, a chlorinated biphenyl type compound and water is not going to be an issue there without obvious signs . If the latter, it would have been changed when these compounds were banned in the US in the late 1970s but not necessarily elsewhere until as late as 2001.
That suggests some doubt as to what may have been the cause but in any case, who would knowingly keep on selling products that were a ticking time-bomb and a safety/liability issue? They'd be obliged to issue warnings to leave devices powered on or powered up regularly as in normal use and to dispose of the caps safely at the end of the device's working life. This would also apply to spare parts that end-users may also be holding, well beyond warranty periods.
A question over drying out and cracking is what dries out and cracks? Clear thermoplastic moulding materials like the casing don't "dry out" unless they are chemically unstable and the case is unsealed. That leaves you with an unstable dielectric material which comes down to a paper and oil issue. If the composite dielectric dried out of some component it would not be water, as that would invite corrosion of the metallizing and affect stability. If the plastic case was breaking down and releasing a flammable monomer, it would shrink rather than expand. On the other hand, a build-up of this gas pressure with the periodic ignition of a 250VAC surge might bring on enough force to explode the case but there's hardly enough internal volume for the oxygen needed to make a significant blast, I'd think.
At the end of the day, there was a problem as others report, here and in other forums. The product, by today's standards, is crap. Maybe in its day it performed better than any other type for a limited period but we do make progress or find ways to side-step the problems, over the years
PME271 caps are metallized paper, saturated in insulating oil. The clue as to what type of oil, lies in the spec. "self extinguishing" to a simple flame and afterburn timing test. This probably means the oil was a silicone or quite likely, a chlorinated biphenyl type compound and water is not going to be an issue there without obvious signs . If the latter, it would have been changed when these compounds were banned in the US in the late 1970s but not necessarily elsewhere until as late as 2001.
That suggests some doubt as to what may have been the cause but in any case, who would knowingly keep on selling products that were a ticking time-bomb and a safety/liability issue? They'd be obliged to issue warnings to leave devices powered on or powered up regularly as in normal use and to dispose of the caps safely at the end of the device's working life. This would also apply to spare parts that end-users may also be holding, well beyond warranty periods.
A question over drying out and cracking is what dries out and cracks? Clear thermoplastic moulding materials like the casing don't "dry out" unless they are chemically unstable and the case is unsealed. That leaves you with an unstable dielectric material which comes down to a paper and oil issue. If the composite dielectric dried out of some component it would not be water, as that would invite corrosion of the metallizing and affect stability. If the plastic case was breaking down and releasing a flammable monomer, it would shrink rather than expand. On the other hand, a build-up of this gas pressure with the periodic ignition of a 250VAC surge might bring on enough force to explode the case but there's hardly enough internal volume for the oxygen needed to make a significant blast, I'd think.
At the end of the day, there was a problem as others report, here and in other forums. The product, by today's standards, is crap. Maybe in its day it performed better than any other type for a limited period but we do make progress or find ways to side-step the problems, over the years
You can find many posts of these crap Rifa capacitors swelling up, shorting, and just plain causing issues on other websites.
The use, or non-use, of equipment has nothing to do with their life-expectancy.
I've seen it myself - customers have brought in equipment to my shop that was in regular use over the years, with swollen Rifa caps.
So judging from my years of experience, I conclude that they're a problematic component and must be replaced by something better.
I'm not familier with that specific amplifier, but other amps that use that amount of filtering, and even more, do not adhere to that way of thinking.
And I've never, ever, seen a 0.1uF cap factory installed in any amp I've serviced.
I've serviced hundreds, maybe thousands over the years to say that.
The usual is 0.01 to 0.047uF - at least here in the USA.