When a capacitor doesn't fit physically

[Junkrepair]

Hi--

I'm sure this question gets asked sometimes, but I haven't been able to find a clear answer.

I'm repairing the power supply of a Magnavox TV (37MF337B/37). It blew the tops of four CapXon GF-series capacitors.

I would like to use some caps with higher voltage ratings and longer lifetime ratings than the bad ones I'm replacing.

If I do that though, they won't fit unless I leave them above the surface of the board: I won't be able to pull the leads through the holes so as to leave the bottoms of the caps snug against the board.

Would that be likely to cause any electrical problems?

The CapXon specs are as follows:
13mm x 25mm, 5mm lead spacing, 1000uF, 35V, 20%, 105 degC, ripple: 2780 mA @105degC and 150 MHz, impedance@20degC: 0.019 Ohms, 5000 hrs. lifetime.

I would like to replace that with a 50V Panasonic part EEU-FR1H102:
16mm x 25mm, 7.5 mm lead spacing, 1000 uF, 50V, 20%, ripple: 3320 mA@105degC and 150 MHz, impedance@20degC: 0.019 Ohms, 10,000 hrs. lifetime.

Since the lead spacing and the diameters are different, the new caps couldn't rest directly on the board. The leads would need to be say 5/8" long between the board and the bottoms of the new caps.

Would this different way of mounting the new caps be worth the improved specs of the larger parts?

Or would it cause problems, for example shortening the lives of these or other parts?

[diif]

Re: When a capacitor doesn't fit physically

Just stick with 1000uF 25v Panasonic FR. They still exceed the spec of the original CapXon without the issue of fitting.

[junktv]

Re: When a capacitor doesn't fit physically

I have no clue what i'm talking about so listen to others but you should be able to find a higher quality cap of same or better spec that fits. Tons of info here and i don't think capxon is in the quality list.

Being used to replacing low quality caps i thought i'd have to rebuild a psu that had UCC caps after 10 years of 24/7. They looked great so staying in for now. That being said I have followed others info and moved caps away from heat sinks down the trace on another psu. So far so good. When you look at many smaller caps the leads are bent to fit. How much that would change resistance on a bigger cap i don't know.

[jazzie366]

Re: When a capacitor doesn't fit physically

Just stick with 1000uF 25v Panasonic FR. They still exceed the spec of the original CapXon without the issue of fitting.

Do this, this is the easiest and best option. With the known reliability of Panasonic, then you're probably never going to see these caps fail.

[jetadm123]

Re: When a capacitor doesn't fit physically

1000uf caps rated at 50V are humongous! Just stick with the Panasonic FR (FR series are considered "long life" caps) 1000uf 35V and you'll save yourself a lot of swearing.

[Junkrepair]

Re: When a capacitor doesn't fit physically

I made a little mistake on the CapXon spec. I should have stated: ripple: 2780 mA @105degC and 100 kHz.

@diif: Actually the 35V 1000uF FR part (EEUFR1V102) doesn't meet the spec for ripple. It's only good for 2600 mA @ 105 degF and 100 kHz. The EEUFM1V102 gives 3190 mA and 0.015 ohms, but it's rated for 7000 hrs endurance rather than the 10000 hrs of the 50V FR.

Maybe the title of my post was misleading. It's not hard to make the larger parts fit in the box---it's just that they wouldn't be sitting directly on the board.

I'm not rejecting the several admonitions to use parts of the expected size, but I can't say I've learned anything about running the caps with long leads.

[sam_sam_sam]

Re: When a capacitor doesn't fit physically

Maybe the title of my post was misleading. It's not hard to make the larger parts fit in the box---it's just that they wouldn't be sitting directly on the board.

I'm not rejecting the several admonitions to use parts of the expected size, but I can't say I've learned anything about running the caps with long leads.

It not a problem as long as it fits standing up or laying down in the (case or cabinet) as long as it can air around the cap

I have done all the above

I hope this helps

[ChaosLegionnaire]

Re: When a capacitor doesn't fit physically

I should have stated: ripple: 2780 mA @105degC and 100 kHz.

@diif: Actually the 35V 1000uF FR part (EEUFR1V102) doesn't meet the spec for ripple. It's only good for 2600 mA @ 105 degF and 100 kHz.

its actually fine. caps have a ±20% tolerance so anything within 20% of the original would be fine. if the replacement were a junk cap, i'd be wary about the 180mA drop in ripple but FR is a quality part so the loss of 180mA of ripple shouldnt affect operation (maybe longevity?) and should still be within tolerance.

[Junkrepair]

Re: When a capacitor doesn't fit physically

It not a problem as long as it fits standing up or laying down in the (case or cabinet) as long as it can air around the cap

Thanks. By that I guess you mean the cap has to be either horizontal or vertical---not slanted?

[diif]

Re: When a capacitor doesn't fit physically

I made a little mistake on the CapXon spec. I should have stated: ripple: 2780 mA @105degC and 100 kHz.

@diif: Actually the 35V 1000uF FR part (EEUFR1V102) doesn't meet the spec for ripple. It's only good for 2600 mA @ 105 degF and 100 kHz. The EEUFM1V102 gives 3190 mA and 0.015 ohms, but it's rated for 7000 hrs endurance rather than the 10000 hrs of the 50V FR.

Maybe the title of my post was misleading. It's not hard to make the larger parts fit in the box---it's just that they wouldn't be sitting directly on the board.

I'm not rejecting the several admonitions to use parts of the expected size, but I can't say I've learned anything about running the caps with long leads.

The hours quoted are with the capacitor at 105c. I know the backs of TVs get warm, but nothing like 105c.
The power supply did not blow the tops off the capacitors due to over voltage. They failed due to being poor quality. There really is no need to go to 50v for the extra lifespan. Fit FM rather than the FR if you're worried about the slight discrepancy in ripple.

I often see caps from the factory laying down on a board with the legs at 90 degrees, secured with glue to stop them flapping about.

[sam_sam_sam]

Re: When a capacitor doesn't fit physically

Thanks. By that I guess you mean the cap has to be either horizontal or vertical---not slanted?

You could slanted them but I find that most of time that dose not work it will not fit for some reason or another

[Junkrepair]

Re: When a capacitor doesn't fit physically

I often see caps from the factory laying down on a board with the legs at 90 degrees, secured with glue to stop them flapping about.

Thank you.

Speaking of glue, I notice that a couple of diodes (ZD906 (a P4KE250A-E3) and D924 (a UF1007), for those interested) next to each other on this power supply board had a blob of opaque white somewhat-putty-like glue mostly covering them. Besides that glue, there was a kind of soft translucent white stuff somewhat like a stringy or dried grease or glue on the board near these diodes. I understand that these two diodes often get replaced in an overhaul. The board surrounding them seems to be discolored from heat.

I was surprised to see glue covering diodes. My guess was that the glue was supposed to silence noise coming from the parts. Is that ever a reason for using glue?

[retiredcaps]

Re: When a capacitor doesn't fit physically

The hours quoted are with the capacitor at 105c. I know the backs of TVs get warm, but nothing like 105c.

Re the quoted hours at 105C, for every 10C drop, hours goes up by double.

So

7,000 hours at 105C
14,000 hours at 95C
28,000 hours at 85C
56,000 hours at 75C
112,000 hours at 65C

Let's assume the TV runs at 65C, that means roughly 112,000 hours on the capacitors. The CCFLs are likely dead or need replacing at 40,000 to 60,000 hours.

In short, use Panasonic caps from a reputable distributor like Mouser or Digikey (not ebay or aliexpress) and they will likely last the lifetime of this TV.

[Junkrepair]

Re: When a capacitor doesn't fit physically

Re the quoted hours at 105C, for every 10C drop, hours goes up by double.

Wow ... just about like the rates of chemical reactions. Thanks ... best thing I learned all week.