Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Need to find alternatives to this non-existent capacitor!

Status
Not open for further replies.

Protonus

New Member
I need help finding an alternative to 4 (two per board) failed capacitors I need to replace, that are seemingly non-existent! Pics below.

The capacitors are 63 V, 10,000uF, rated for 105 C. Standard Axial Solder Lead style, aluminum electrolytic.
They are made by Nippon Chemi-Con, and they are KMG Series capacitors (their general purpose series).
The plastic disc at the top of the cap is embossed wtih "13D68" - perhaps a date code?

The special thing about these capacitors is their width. They are approx 25.7mm wide (and 50.5mm tall, which is common height). I have not found any other capacitors with the 63 V, 10,000 uF rating, in this width - looking for alternatives on digikey and elsewhere. All of them are either 30mm or 35mm wide. And looking at the pics I'm attaching below - I don't have room for a 30mm wide cap. I bought high quality (low ESR, higher operating hours, etc) replacement caps without even thinking about the package size, and they were 35 mm wide. You can see how much wider they are in a picture below.

These caps are out of an amplifier board, from a SoundBridge brand 12" powered subwoofer that's about 10 years old. I have two of these amplifier boards - both of them had cap failure on these 2 caps - so I'd prefer to get better quality replacements. This sub was only sold in very low quantity and pretty much put on clearance immediately thru Woot.com - you can read a little about it here: **broken link removed**

An interesting note is that printed underneath all the components on these boards is the value of the component soldered there. As you'll see on these boards, underneath the caps after I desoldered them, it said they should be 63 V, 4,700 uF ? That might be a clue here?

I don't know enough about circuit design to know what alternatives I have here?
I'm figuring it might be possible to use a different value cap - I know usually stepping up in voltage or capacitance is OK, but not down? The only other option I came up with is to use one of the larger 30mm wide caps I found, and run leads to it, and mount it perpendicular to the board, floating over the other components (wasn't even sure the best way to do that?). This doesn't seem like a great idea either, especially in a subwoofer amp which will be vibrating all the time.

I contacted Nippon Chemi-Con's USA office and they can't find a KMG series cap in this size either, even in their older paper catalogs. Nor do they have a cap in this value, at 25mm width. I sent them these pics and info to their engineering department so they can investigate further because they're as curious as I am! I'm leaning towards these caps either being fakes, or misrepresented/labeled?

Anyways, thanks to ya'll in advance for your suggestions!

How the board looked to start (you can see one of the caps is swollen enough at the top to tear the plastic jacket on it):
x7JWl8X.jpg


Caps desoldered, showing the 4,700 uF rating underneath:
ZBV7Sli.jpg


Cap pic 1:
uFRzO0V.jpg


Cap pic 2:
Pl5X15f.jpg


Cap pic 3:
CKYTNfX.jpg


Comparison of new replacement caps, being 1 CM wider than the old ones (the old one is in the lower right!)
XbSj7Wt.jpg
 
First, those look like radial leads, not axial.
Second, will the capacitors you have found that are 30 mm in diameter fit?

Capacitance, voltage, temperature, ESR, and ripple current can be important characteristics to consider, but if you cannot find the exact replacement, then matching capacitance and voltage with a good quality substitute will often suffice. Meeting that temperature rating should be easy too. The newer caps may actually be superior.

John
 
First, those look like radial leads, not axial.
Second, will the capacitors you have found that are 30 mm in diameter fit?

Capacitance, voltage, temperature, ESR, and ripple current can be important characteristics to consider, but if you cannot find the exact replacement, then matching capacitance and voltage with a good quality substitute will often suffice. Meeting that temperature rating should be easy too. The newer caps may actually be superior.

John

Sorry, meant radial, thanks for the correction.

No, 30mm replacement's won't fit. It's close, but measuring with my caliper I only have about 28 mm max to work with. I have not found any caps in this rating, that are less than 30mm.
 
Just get ones with the lowest ESR you can afford and retrofit to fit.
**broken link removed**
 
Those Digikey versions are all ≥30mm, so won't fit directly.

Nichicon does make a 6800 uF cap with other ratings the same (https://www.mouser.com/ProductDetai...=sGAEpiMZZMvwFf0viD3Y3YYRLScya6Y2nRmQGNugVvA=). It might work, as the pcb is labeled 4700 uF. Cornell Dublier also makes an 8200 uF, 25-mm diameter cap, but its temperature rating is only 85°C.

Alternatively, I would consider mounting 10,000 uF caps that are 30 mm in diameter displaced from the board (i.e., leads bend in a Z) and have Teflon or other insulation tubing over the extended leads.

John
 
Nichicon does make a 6800 uF cap with other ratings the same (https://www.mouser.com/ProductDetail/Nichicon/UVY1J682MRD/?qs=sGAEpiMZZMvwFf0viD3Y3YYRLScya6Y2nRmQGNugVvA=). It might work, as the pcb is labeled 4700 uF. Cornell Dublier also makes an 8200 uF, 25-mm diameter cap, but its temperature rating is only 85°C.

Thank you. So it it should be electrically OK to exchange these caps out for ones with a smaller capacitance rating, as long as the voltage rating is the same?

My knowledge of circuitry is rather pragmatic, I don't know enough of design to know when things like that are OK. I normally try to replace everything with identical values when I repair boards.

Alternatively, I would consider mounting 10,000 uF caps that are 30 mm in diameter displaced from the board (i.e., leads bend in a Z) and have Teflon or other insulation tubing over the extended leads.

That was the "solution" I was leaning towards, but I wasn't sure the best way to pull it off. The leads on most of the caps are short, so I'd have to solder in longer leads. What kind of leads to use, out of what material was my first question. I am assuming I should use solid leads not strand wire?

The other concern is that these caps are going into a subwoofer cabinet, which is bound to shake/vibrate a lot. I was concerned about caps that are "floating" above the board not anchored to anything. But wasn't sure of the best way I should secure them, if at all?

Got any pics or suggestions on this sort of retrofit?
 
Generally you want to keep the capacitance as close as possible to the original. However, since these appear to be filter caps the capacitance is not as critical. A lower capacitor will give you more ripple, so if you go too low you may hear a 50/60Hz "buzz" from the mains (if it's powered directly from mains) or a higher frequency squeal if it uses a switching power supply. But this would only happen if you go significantly lower than the 10,000uF original (i.e. a few hundred microfarads, perhaps).

Did you specify the lead spacing somewhere? Not sure I saw it. Anyway, here's what I'm seeing at Mouser:

https://www.mouser.com/Passive-Comp...zh229Z1yzh252Z1yzh24zZ1yzh0b4Z1yzh206Z1yzh179

Perhaps one of those will work?

Matt
 
The value printed on the PCB is 4700 uF. Perhaps by experience that was found to be too low. Or, maybe the manufacturer just got a good deal on the "10,000 uF" capacitors -- either real or mismarked. Since the 105° reflects resistance to an environmental temperature, not a running temperature typically, the 8200 uF @ 85° might be a good compromise between what's available and what was there originally, particularly if it is in a relatively large enclosure, which is why I mentioned it.

As for vibrations, low temperature hot-melt adhesives or other adhesives with body will keep the parts from moving. Besides hot melt, I have used Goop, E6000 (same manufacturer), electrical grade silicone, and Loctite GO2 glue for similar functions. Goop/E6ooo are good flexible adhesives, but shrink considerably. Silicone is quite flexible but generally has poor peel strength, which can be an advantage or disadvantage. GO2 was a new experience for me. It is marketed as a "hybrid" polyurethane/silicone, which is surely mostly marketing buzz. Chemically, it is similar to a silicone, releases methanol rather than acetic acid on curing, and seems to have better peel strength than silicone caulking, but nowhere near the strength of Goop/E6000. I used it for the first and only time to keep battery packs from moving around the the bottom an enclosure. It worked well and was still removable.

John
 
Last edited:
I don't know enough of design to know when things like that are OK. I normally try to replace everything with identical values when I repair boards.

Electrolytic's are almost always VERY non-critical, and have huge tolerances as well.

Anything 4700uF and above (that will fit) would be fine.

Don't worry if they are only 85 degrees, there's no need for 105 degree in this low spec. application - where you NEED 105 degree caps (and good quality ones) is in switch-mode supplies where the frequency is a great deal higher.
 
Generally you want to keep the capacitance as close as possible to the original. However, since these appear to be filter caps the capacitance is not as critical. A lower capacitor will give you more ripple, so if you go too low you may hear a 50/60Hz "buzz" from the mains (if it's powered directly from mains) or a higher frequency squeal if it uses a switching power supply. But this would only happen if you go significantly lower than the 10,000uF original (i.e. a few hundred microfarads, perhaps).

Gotcha, that makes a lot of sense, thanks for the explanation.

Well, 60Hz would be very audible out of this subwoofer (it's more of a power woofer, doesn't quite go deep enough to be a "real" subwoofer). So I could see a buzz being bad.

You mention significantly lower, then say a few hundred uF. Wouldn't 6,800 uF down from 10,000 uF be a few hundred lower? Or perhaps I'm not following?

Did you specify the lead spacing somewhere? Not sure I saw it. Anyway, here's what I'm seeing at Mouser:

https://www.mouser.com/Passive-Comp...zh229Z1yzh252Z1yzh24zZ1yzh0b4Z1yzh206Z1yzh179

Perhaps one of those will work?

Sorry, I forgot. I'm measuring the lead spacing to be 11mm on each of them, measuring from the center of each lead to the other center. So I figure in reality they were probably spec'd at 10mm? I could probably even make 12.5mm work if I bend the leads out.
 
Gotcha, that makes a lot of sense, thanks for the explanation.

Well, 60Hz would be very audible out of this subwoofer (it's more of a power woofer, doesn't quite go deep enough to be a "real" subwoofer). So I could see a buzz being bad.

You mention significantly lower, then say a few hundred uF. Wouldn't 6,800 uF down from 10,000 uF be a few hundred lower? Or perhaps I'm not following?

Nope, down to a couple hundred uF, not down by a couple uF. In other words, replacing a 10,000uF capacitor with a 100uF capacitor is probably a bad idea. However, replacing a 10,000uF capacitor with a 6800uF capacitor probably won't be as bad. I suggest picking one around the 6800uF range (they are quite common) and if you start hearing a hum or buzz, then you know you've gone too low. I don't really see that happening, but it's something to keep in mind.

Sorry, I forgot. I'm measuring the lead spacing to be 11mm on each of them, measuring from the center of each lead to the other center. So I figure in reality they were probably spec'd at 10mm? I could probably even make 12.5mm work if I bend the leads out.

Good to know. Just a note, if you go with 12.5mm spacing you'll need to bend the pins in, not out :p

Try using the Mouser parametric search. I have found them to have a much larger variety of electrolytics than Digikey (though perhaps I haven't been in the right category.... :D ).

Good luck!
 
Ok, thanks everyone. This is the list of 25mm wide, 63 V, 6,800 uF caps that Digikey offers:
https://www.digikey.com/product-sea...1&stock=1&quantity=0&ptm=0&fid=0&pageSize=500

6,800 uF being the highest value at 63V and 25mm wide that they sell. And as you can see there are a bunch, 1 from Chemi-Con and 7 from Nichicon. Since the failed caps are all from Chemi-Con, I'm inclined to use Nichicon instead heh.

Turns out (after looking at lots of pictures) the type of caps I'm replacing are actually Snap-In style, not just normal through-hole radials. That explains why the leads were so thick and reinforced and the measurement being a bit "off" on lead spacing.

My basic understanding of cap ratings, is that higher Ripple Current, and higher operating hours @ rated temperature, are both desirable for long term reliability. And I'd sure like to go longer without having to fix these amps again.

So bearing everything I just mentioned in mind, and my limited knowledge here, I selected this cap:
https://www.digikey.com/product-detail/en/LLS1J682MELA/493-7293-ND/2548902

They are Snap-In style (one of only 2, the other being the Chemi-Con one). They have the highest operating hours of any of those, at 3,000 hours @ 85 C. And they have the highest ripple current rating, at 4.52A. I'm not sure that any of this really matters, but when you have 8 to pick from, I needed some way? Also, the price was nice, $4.33 each.

I don't know that these are the best choice though, out of those 8. And my order won't be fulfilled for a while, so, if you feel I should have selected one of the other 7 caps, LMK and I'll change my order.

Thanks again for all your help!
 
Just following up here. The amp is fixed! Working great, so happy to have bass again!

The new 6,800 uF caps I ordered above, worked great. So thank you very much for the advice here on choosing a suitable replacement. In fact, as you read on, these 6,800 uF are likely overkill now.

An interesting addendum to the story, I have two of these amps, so 4 of these caps. I recently picked up an Peak ESR70, so I could start measuring ESR (and capacitance) for checking boards for failing caps. I decided to try it on the old caps, the "10,000" uF ones.

The visibly failed cap, the one that was bulging, measured at only 3,600 uF, which is certainly low. It was certainly bad.

The other 3 caps though, that looked OK? All 3 of them measured within 10% of 4,700 uF! I think it's no coincidence, particularly since the board itself was marked with 4,700 uF. Which would mean that all of these caps were badly mislabeled (they claimed to be 10,000 uF)!

I'll consider that mystery solved, though it seems awfully terrible to me.
 
All 3 of them measured within 10% of 4,700 uF! I think it's no coincidence, particularly since the board itself was marked with 4,700 uF. Which would mean that all of these caps were badly mislabeled (they claimed to be 10,000 uF)!

Sounds like the amp was built with cheap Chinese knockoffs.

FakeCapacitor1.jpg
 
Sounds like the amp was built with cheap Chinese knockoffs.

FakeCapacitor1.jpg

Contrary to the image, though, they would not have used the top-brand "Rubycon". More likely Samxon or Capxon or somthing like that.
 
OMG I've never seen/heard of anything like that! That's ridiculous!

I think I'll open up one of these tonight! Now I'm even more curious!
ingenuity unlimited. is what I call their attitude to manage anything like fakes.

as regards
audio hum (100Hz) on the woofer,
i would suggest to add 0.01uF capacitors on the 4 diode elements of the bridge and it would disappear.
Even two would suffice but , then you have to select across which of the 4 diodes.
I had enough practical experience on this in my service and hobby either.
 
I opened the failed cap. Sadly, no nesting dolls found. Just expected roll construction. Interestingly there was still a lot of liquid electrolyte left, despite quite a bit having leaked out and the case being expanded. It was wet enough I didn't unroll it to look for failure.
 
ingenuity unlimited. is what I call their attitude to manage anything like fakes.

as regards
audio hum (100Hz) on the woofer,
i would suggest to add 0.01uF capacitors on the 4 diode elements of the bridge and it would disappear.
Even two would suffice but , then you have to select across which of the 4 diodes.
I had enough practical experience on this in my service and hobby either.

I dont understand how adding 0.01uF will reduce the 100Hz ripple. At best it will attenuate the buzzing harmonics at 100Hz rate.
 
I dont understand how adding 0.01uF will reduce the 100Hz ripple. At best it will attenuate the buzzing harmonics at 100Hz rate.
I meant same and surely not the post bridge filter. It is noise generated while the diode trasnsits around 0.65V either in raising or trailing mode. ( forgot the name for this-- ringing noise or some such thing . t is surely double the mains frequency.
A momentary tart of conduction , i suppose makes up a small spike like.
During 60s attending transistorised radio receivers, working from in built power supplies,
the 100Hz hum was disturbing even on shotwave bands and once these parallel caps are placed across diodes, it just disappears.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top