I’ve just done a brain transplant (new control board) on our dishwasher. Of course I couldn’t resist opening the old one and it looks like two large capacitors are bulging.
formatting link
If that’s the only thing wrong with them it seems reasonable to change them before the second of the two dishwashers dies. (It’s a “twin-drawer” machine).
I’m struggling a bit to find replacements that are 100V, 1000uF, 85C and no more than 20mm high and 25mm in diameter. The originals were made by Jamicon but I can’t seem to find them.
Rubycon have some 25x25mm snap-ins, which can be found on Mouser:
100MXC1000MEFCSN25X25 (Mouser links don't seem to work without a session tracking ID, so you'd have to search mouser.com for the part number)
Also they have a 25x25mm from CDE: SLP102M100C1P3
Jamicon also have a 25x25:
formatting link
I'm guessing a 25mm high is no good?
I'm not finding anything shorter than that, but their catalogue is hard to search. Does the existing one have any other markings on it? I'm looking for the series, eg LT or LF in the above. Is it wire ended or a snap-in?
A manufacturer seems to have acquired "D sized case" capacitors, you need to do a swap, and with the stated parameter set, all you can find are "E sized case" electrolytic capacitors in your catalogs. I bet every motherboard I own here, has that situation on the bulk caps near VCore.
The only exception might be the substitution of OSCONs in place of the commodity originals, and the parametrics (ESR,ESL) as well as ripple current rating (4x normal) are quite different. And if this is part of an SMPS, some slight tweaks on the analog components next to the switching chip are required. A lot of SMPS data sheet recommended designs, are designed with commodity ESR and ESL values in mind.
Circuits have some tolerance to slight reductions in capacitance, so maybe a solution would be an 800uF cap instead of a 1000uF cap. Then look at the X7R versus whatever, and see if you can justify your choice based on tolerance alone.
Why can't the designers leave some air-space for the next-sized cap ???? Why do they select "interference hole sizes" for the cap leads, so it's impossible to get out, even with an expensive vacuum desoldering machine ???? I worked for one company, that provided plenty of clearance on the legs, and they used zigzag lead forming on the pick%place, so the caps would not fall out during the trip through the solder tunnel. Those were a pleasure to remove. Came right out on the first try. An actual, maintainable, PCB.
I don’t think so but I’ll have another look at the enclosure. It’s just possible there might be room. I’ve tried closing the enclosure with and extra 5mm of cardboard inside to simulate a longer capacitor and it just closes (with probably just a little pressure on the circuit board).
It does. “LP” (low profile maybe?) and “735K0(MM). (Not clear whether it’s an O or an 0).
Pass. I’m not familiar with the differences. Could I tell without desoldering it first? The legs are 10mm apart.
Hmm, 17 years-ish?
It might seem a lot of trouble for such an old machine but it was an expensive machine in its day.
formatting link
It still works well and has been otherwise utterly reliable. It’s also beautifully designed for easy service and repair (apart from these capacitors!) You can access the vast majority of the components either from inside the machine drawer or by removing the front panel, neither of which require the use of tools.
I must admit that much of this is way over my head. ;-) What I have discovered though is that in any capacitance near 1000uF, there is nothing in a 20mm long/high package.
It looks like a 25mm cap could be squeezed in (at the expense of any overhead clearance) and with probably a bit of pressure on the PCB when the enclosure is shut. How big an issue that is likely to be I have no idea.
Howsabout cutting holes in the cover and adding a "power bulge" to accommodate the replacement caps? ... or mounting them somewhere else and running wires back to the original location?
Which may indicate that it was not actually a 1000uF capacitor (or not
100V). If none of the reputable manufacturers can fit a 1000uF,100V capacitor into a can of that size is seem unlikely that Jamicon could have done so.
My general experience with replacing failed electrolytic capacitors is that the originals which fail quickly have been from no-name manufacturers and have been in cans smaller than the replacements from Rubicon, Panasonic, Vishay etc. The replacements have always lasted longer than the originals and always to the end life of the items.
That did cross my mind but I don’t like the idea of possibly compromising the fire resistance of the control board case (assuming it has a degree of fire resistance of course!)
If it's bulging, that might take up an extra mm or so?
Would there be space if the enclosure was cut/shaved down a little? Is it plastic or metal?
The only LPs they have are branded Teapo not Jamicon, and are too big:
formatting link
Snaps have short stubby legs stamped from metal, like the picture on that datasheet. They don't need cutting and the tips poking up from the solder will be uniform and flatter. Wire-ended capacitors show the end of the cut wire, which will be a sharp peak.
That sounds well into the 'capacitor plague' era, and maybe a cheap brand like Jamicon was a bit overzealous in their published specifications? It is possible they have more 'enthuiastically' marked their product compared with the competition. Which might also explain why it failed.
Another thought is to study the circuit and see if you really need that voltage or capacitance rating. eg if it can get away with a 63V cap then you can probably find one to fit.
Definitely worth keeping these things running, especially for a few pennies of components. If you can find them!
It would be unusual, for them to leave room for you to do that.
Or they would have done it, themselves (pair of 470uF, in place of single 1000uF).
For filtering, there is the value of capacitance, but there is also a thing called the "ripple current rating". The circuit would work better with four 470uF, because there would be a higher overall ripple rating. Capacitors can have *multiple amperes* of ripple current flowing, which can make them warm.
This is why motherboards have two banks of caps, multiple caps in each bank. Ripple in each bank.
They might have seven caps shoulder to shoulder - that's not to get 7x the capacitance, that's to get 7x the ripple rating. If they simply wanted 7x the capacitance, they would use a single capacitor with a 7 times higher value printed on the sleeve.
Params:
capacitance \___ stamped on sleeve WVDC / May list temp, but not hours at that temp. (Need specific summary family catalog for this.) life@temp / Using Arrhenius, you can re-rate at the application temp. Could be 2000 hours at 105C for example. Arrhenius is not the chemlab 7C value! It is curve fitted to capacitor lab data. Making it 256000 hours at 25C. You don't want them running at the rated temp. You want cooler air than that. Much cooler air.
ripple current --- spec easily available (if the value is not embarrassing).
ESR \___ spec can be a PITA to get ESL / dont know exactly why they play this game application circuits have already taken "lax" values into account, so it's not like it matters :-) SMPS sample designs take "shit caps" into account.
That looks like a three phase DC motor controller of some kind.
It seems like there is probably space to lay some long and thin caps, eg
18mm d x 35mm h 1000uF 100V, above the microcontroller chip at the top left. I would try to keep them away from the transistors at the top right as they may get hot, but the MCU area probably won't. Then wire them to the existing cap PCB holes.
I am assuming that these are radial capacitors (both connections on one end)? If so, you can also look at the axial range (one lead out of each end), which can be installed on end, with the "top" wire insulated and brought down the side of the capacitor and down to the PCB.
It may not help, but it could give you a wider range to choose from.
I think we're at cross-purposes! You said that you thought that it might not actually be a 1000uF capacitor. In your OP, you said you'd put in a new control board, so I was suggesting you unsolder one of the new (non-bulging) capacitors from that and measure it to see if a new one was really 1000nF. If it was only, eg 470uF, one of that value could be used to replace the bulging caps, and being only 470uF might fit.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.