Double-check my replacement guesses?

[Ickster]

So I'm here because when my monitor died, I vaguely remembered reading something on the internets about bad caps in TVs--my memory was that a couple of cheap caps failing was causing people to throw their expensive TVs away and buy new. It occurred to me that the same was probably true for LCD monitors, as they're basically the same thing.

The monitor in question is an off-brand el-cheapo special from Micro Center. It's 21", and it was only $99 two or three years ago; I'm probably lucky it worked as long as it did. Economically, it's not really worth fixing (and it's already been replaced), but that's no reason not to try

The caps in question are from the power supply board. The monitor started having problems waking up from power save mode, and eventually refused to wake up at all.

After taking it apart, there was one cap that was clearly bad, but I went ahead and removed all of the smaller ones on the same side of the board. There are two still left on there (see the attached image), but they're glued on and I'm afraid of breaking something before I actually get to try to fix it. If it's really not difficult to remove them and you think it should be done, go ahead and say so.

Anyway, I managed to identify all of the ones I removed; they're called out in the image, and listed in the table below. Each double row in the table lists the one that came off the board first, followed by my guess for the replacement. For the last one I listed, I'm really unsure about the replacement and left it for suggestions.

So, how'd I do? (I'm fully expecting people here to have better suggestions, but I wanted to try before asking.)

Description	Cap.	Volt.	Ht.	Dia.	Lead	Ripple	Imp.	Suggested Replacement
------------------------------------------------------------------------------------------------------------------								
CapXon KF P745	2200 μF	10 V	20 mm	13 mm	5 mm	1750 mA	0.041 Ω	http://www.digikey.com/product-detail/en/UHE1A222MHD6/493-1502-ND/589243
Nichicon HE	2200 μF	10 V	20 mm	13 mm	5 mm	1900 mA	0.035 Ω	
------------------------------------------------------------------------------------------------------------------								
CapXon KF P746	1000 μF	10 V	16 mm	10 mm	5 mm	1040 mA	0.076 Ω	http://www.digikey.com/product-detail/en/UPW1A102MPD/493-1746-ND/589487
Nichicon PW	1000 μF	10 V	16 mm	10 mm	5 mm	1050 mA	0.068 Ω	
------------------------------------------------------------------------------------------------------------------								
CapXon KF P744	220 μF	25 V	12.5 mm	8 mm	3.5 mm	550 mA	0.15 Ω	http://www.digikey.com/product-detail/en/UHE1E221MPD/493-1549-ND/589290
Nichicon HE	220 μF	25 V	11.5 mm	8 mm	3.5 mm	640 mA	0.13 Ω	
------------------------------------------------------------------------------------------------------------------								
CapXon KF P747	1000 μF	25 V	21 mm	10 mm	5 mm	1650 mA	0.045 Ω	?
?	?	?	?	?	?	?	?	
------------------------------------------------------------------------------------------------------------------

[cheapie]

Re: Double-check my replacement guesses?

So I'm here because when my monitor died, I vaguely remembered reading something on the internets about bad caps in TVs--my memory was that a couple of cheap caps failing was causing people to throw their expensive TVs away and buy new. It occurred to me that the same was probably true for LCD monitors, as they're basically the same thing.

The monitor in question is an off-brand el-cheapo special from Micro Center. It's 21", and it was only $99 two or three years ago; I'm probably lucky it worked as long as it did. Economically, it's not really worth fixing (and it's already been replaced), but that's no reason not to try

The caps in question are from the power supply board. The monitor started having problems waking up from power save mode, and eventually refused to wake up at all.

After taking it apart, there was one cap that was clearly bad, but I went ahead and removed all of the smaller ones on the same side of the board. There are two still left on there (see the attached image), but they're glued on and I'm afraid of breaking something before I actually get to try to fix it. If it's really not difficult to remove them and you think it should be done, go ahead and say so.

Anyway, I managed to identify all of the ones I removed; they're called out in the image, and listed in the table below. Each double row in the table lists the one that came off the board first, followed by my guess for the replacement. For the last one I listed, I'm really unsure about the replacement and left it for suggestions.

So, how'd I do? (I'm fully expecting people here to have better suggestions, but I wanted to try before asking.)

Description	Cap.	Volt.	Ht.	Dia.	Lead	Ripple	Imp.	Suggested Replacement
------------------------------------------------------------------------------------------------------------------								
CapXon KF P745	2200 μF	10 V	20 mm	13 mm	5 mm	1750 mA	0.041 Ω	http://www.digikey.com/product-detail/en/UHE1A222MHD6/493-1502-ND/589243
Nichicon HE	2200 μF	10 V	20 mm	13 mm	5 mm	1900 mA	0.035 Ω	
------------------------------------------------------------------------------------------------------------------								
CapXon KF P746	1000 μF	10 V	16 mm	10 mm	5 mm	1040 mA	0.076 Ω	http://www.digikey.com/product-detail/en/UPW1A102MPD/493-1746-ND/589487
Nichicon PW	1000 μF	10 V	16 mm	10 mm	5 mm	1050 mA	0.068 Ω	
------------------------------------------------------------------------------------------------------------------								
CapXon KF P744	220 μF	25 V	12.5 mm	8 mm	3.5 mm	550 mA	0.15 Ω	http://www.digikey.com/product-detail/en/UHE1E221MPD/493-1549-ND/589290
Nichicon HE	220 μF	25 V	11.5 mm	8 mm	3.5 mm	640 mA	0.13 Ω	
------------------------------------------------------------------------------------------------------------------								
CapXon KF P747	1000 μF	25 V	21 mm	10 mm	5 mm	1650 mA	0.045 Ω	?
?	?	?	?	?	?	?	?	
------------------------------------------------------------------------------------------------------------------

Those all look pretty good. For that last cap, a Rubycon ZLJ 25V 1000uF should be close enough. A closer replacement would be nice, but it should work. You can get them here.

Those other caps, if they look okay, could probably be left for now. Primary cap failures are rare, and little tiny caps don't die often either.

[PeteS in CA]

Re: Double-check my replacement guesses?

The date code on the main O/P transformer is "0746", the "P7XX" for the caps you listed is the parts' date codes, and I saw the date code on the cap that is laying down (and glued) is P804. So that power supply is probably 5-5.5 years old. Judging by that cap being mounted that way, I'm going to guess that height is a tight dimension where that P/S is located in the monitor, and cooling is probably marginal. Nichicon's PW and HE series are both excellent series; in that environment, I would recommend using PW rather than HE (and looking at United Chemicon's LXZ and Panasonic's FC series if you can't get PW parts with the right values and case size). HE should probably do just fine, especially if you don't drape shirts over your monitor, but PW would probably last a bit longer still (a difference, in the application, between excellent and reeeeally excellent).

[Ickster]

Re: Double-check my replacement guesses?

So that power supply is probably 5-5.5 years old.

Wow. I amazed the monitor is that old. I know that there was probably a greater-than-usual lag between maunfacturing and sale because it seemed like one of those weird lots that discount retailers sometimes get, but still.

Judging by that cap being mounted that way, I'm going to guess that height is a tight dimension where that P/S is located in the monitor, and cooling is probably marginal. Nichicon's PW and HE series are both excellent series; in that environment, I would recommend using PW rather than HE (and looking at United Chemicon's LXZ and Panasonic's FC series if you can't get PW parts with the right values and case size). HE should probably do just fine, especially if you don't drape shirts over your monitor, but PW would probably last a bit longer still (a difference, in the application, between excellent and reeeeally excellent).

You're correct; the power supply is mounted in the rear of the monitor case. It's a fairly tight fit, and all passively cooled. Because of the fit, case size was the final constraint that put me down towards the caps I chose. My basic method in narrowing down my choices was to use the Digikey site and filter by lead diameter (exact), ripple (>= spec), voltage (exact), capacitance (up to 20% above spec), and impedance (+- 10% from spec). That pretty much always left me with no more than four choices, and size was the final determining factor.

Anyway, glad to hear that my choices are generally good. I'll get them ordered and update the thread when I have it put together.

Thanks!

[theOracle]

Re: Double-check my replacement guesses?

I think I would forget lead diameter as a filter criteria

I have yet to work on anything where my replacement choice cap's leads were too large to fit through the hole in the board

there are a lot of other great choices from Digi-Key that would work in this application, such as Chemi-con KZE, Panasonic FR and FM, Rubicon ZLH, etc.

[c_hegge]

Re: Double-check my replacement guesses?

I would recommend using PW rather than HE (and looking at United Chemicon's LXZ and Panasonic's FC series if you can't get PW parts with the right values and case size). HE should probably do just fine, especially if you don't drape shirts over your monitor, but PW would probably last a bit longer still (a difference, in the application, between excellent and reeeeally excellent).

Huh? HE has a longer endurance lifetime rating than PW does. (2000-8000 hrs for PW vs 4000-10000 hrs for HE), and the ripple ratings are slightly higher. I would have thought that HE would last much longer.

[Wester547]

Re: Double-check my replacement guesses?

Huh? HE has a longer endurance lifetime rating than PW does. (2000-8000 hrs for PW vs 4000-10000 hrs for HE), and the ripple ratings are slightly higher. I would have thought that HE would last much longer.

HE uses aqueuous electrolyte, PW does not (it uses an organic solvent that dries up much slower and is more resistant to heat, as does FC and LXZ from Panasonic and NCC respectively). That is what Pete was denoting. However, as far as ripple and ESR ratings go, HE is indeed superior. I would think in heat generators like a LCD monitor that PW would last longer even if it wouldn't perform as well in terms of ripple suppression. HE may also be closer to the original capacitor's ESR and ripple ratings. I wouldn't say "10,000 hours" is much longer than "8,000 hours" (unless you mean that HE would last much longer than KF from CapXon, which would of course be true...), especially if you consider the fact that endurance ratings aren't wear out tests (and are tested within loose parameters to begin with) so they aren't the be all end all of capacitor lifetimes (which are not published).

[c_hegge]

Re: Double-check my replacement guesses?

I realise that about the endurance lifetimes, but I forgot about the electrolyte difference, so yes, you could be right. I was just thinking more along the lines that the HEs would be run further under their maximum ratings than PW.

[theOracle]

Re: Double-check my replacement guesses?

aqueous electrolytes have been responsible for most of the premature cap failures over the years

remember all of the massive premature failures with electrolytes based on aqueous "quaternary ammonium salt compounds" ?

[Wester547]

Re: Double-check my replacement guesses?

aqueous electrolytes have been responsible for most of the premature cap failures over the years

remember all of the massive premature failures with electrolytes based on aqueous "quaternary ammonium salt compounds" ?

I thought the non-aqueous capacitors used quarternary ammonium salts as organic solvent for electrolyte. It still evaporates but much slower than water-based electrolyte. The capacitors from the 1980s and 1990s that would eventually leak from the bottom, to my knowledge, didn't have sufficient bungs. However, the more water in the electrolyte, the more sensitive to heat a capacitor is. Unfortunately, especially lower (and lowest) ESR lyrics (the lower ESR they are) do require more water in their electrolyte, but I suppose certain additives could be used in the electrolyte to minimize the sensitivity to heat and increase the lifespan.

[theOracle]

Re: Double-check my replacement guesses?

yes, I had just woken up and rather mis-wrote that

The quaternary ammonium salt compounds are neither aqueous electrolytes nor organic solvents. The newer low ESR electrolytic capacitors use aqueous electrolytes. The quaternary ammonium salt compounds are not used any more except in electric double layer capacitors、

Organic solvent capacitors tend to show lower ESR and sometimes they can cause oscillation problems but not as often as ceramic capacitors. Capacitors with an ESR lower than 30 milliOhm are generally speaking a bit dangerous as far as causing oscillation in some cases.

I have personally ran into this issue when I have used a low-ESR capacitor in situations where the original cap was not low-ESR.

[PeteS in CA]

Re: Double-check my replacement guesses?

Huh? HE has a longer endurance lifetime rating than PW does. (2000-8000 hrs for PW vs 4000-10000 hrs for HE), and the ripple ratings are slightly higher. I would have thought that HE would last much longer.

My past observation is that water-based electrolytes handle heat less well than conventional low-resistance electrolytes (water-based electrolytes expand more under heat). PW (like LXZ and FC, which have similar impedance and ripple current ratings) of the latter type, while HE (like KZE, which has similar impedance and ripple current ratings) uses a water-based electrolyte. Thus, because of the poor cooling in the application, I think PW caps might last longer than HE caps. Chances are, either series might last until the monitor failed for some other wear-out mechanism or became laughably obsolete. But I think PW would be slightly more likely to last that long. I tried to convey in my post that the difference in life I think likely is not large. Maybe I should have added that if another 5 or 6 years' life for the monitor is what is wanted/expected, PW and HE should both do fine.

Sorry, wester and c_h, I hadn't read further in the thread, and didn't see that you had already covered most of what I posted above. Nichicon's 5000-hour PA series has impedances and ripple current rating similar to HE, but its life rating is lower, and I'm not sure whether it's available from disties like Digi-Key, Newark, Mouser or Allied. Also, PA's highest voltage rating is 35V, last I checked (a month or two ago).

[theOracle]

Re: Double-check my replacement guesses?

PeteS - where can I find a definitive list on which caps are aqueous and which are organic solvent based electrolyte?

I am particularly interested in know which these are:

Panasonic FJ, FK, FM, FR
Chemi-con KZE, KZM
Rubycon ZLJ, ZLH, ZLG
Nichicon HW, EP

also would you classify all of the above as great capacitors?

[Wester547]

Re: Double-check my replacement guesses?

I believe Nichicon PA is a direct cross to Panasonic FK (also non-aqueous) - that puts it in between Chemi-con KY and KZE in terms of ripple and ESR ratings.

PeteS - where can I find a definitive list on which caps are aqueous and which are organic solvent based electrolyte?

I am particularly interested in know which these are:

Panasonic FJ, FK, FM, FR
Chemi-con KZE, KZM
Rubycon ZLJ, ZLH, ZLG
Nichicon HW, EP

also would you classify all of the above as great capacitors?

It's actually really simple, you can tell by the temperature ratings. The lower the lowest temperature rating, the less water, the higher the lowest temperature rating, the more water. Capacitors that are rated for -55C to +105C are generally non-aqueous, capacitors that are rated for -40C to +105C (or +85C and +125C) are aqueous, and the ones rated for -25C to +105C (or +85C and +125C?) even more so. But the amount of water in electrolyte is also determined by how low ESR a capacitor is. The lower ESR a capacitor is, the more water content it has.

[PeteS in CA]

Re: Double-check my replacement guesses?

Nichicon and Chemicon make it easy, the tell-tale is in the series name. With Nichicon, H_ series names are aqueous; low Z non-aqueous Nichicon series are P_. With Chemicon, KZ_, KY_ and KY are aqueous. Other than the YXF series, I don't know if Rubycon makes a non-aqueous low Z series - they bet the farm on aqueous. The only non-aqueous Panasonic series I know is FC, but I'm not very up to date on Panasonic.

[theOracle]

Re: Double-check my replacement guesses?

I have a famous friend who is a star engineer for Sony in Japan, now retired, and a holder of over 100 U.S. patents - and he says Rubycon had the best reputation of Nichicon, Rubycon, and Chemi-con inside Sony Corporation (Sony would never use a Panasonic cap for obvious competitive reasons, so he has no real world info on them)

I asked him why he felt Rubycon had an edge, and he said "like with beer and bourbon, it's the water!" - apparently Rubycon's main plant in Japan is on the site of a natural spring of incredible purity, and he really was being serious about the pure water they had exclusive access to, being a real advantage! -

very interesting!

[momaka]

Re: Double-check my replacement guesses?

The only non-aqueous Panasonic series I know is FC, but I'm not very up to date on Panasonic.

I wonder where HFQ stands. I've seen them used a lot in older PSUs. I think they are discontinued now.