Samsung SyncMaster 940T blows inverter fuse

[PCBONEZ]

If you have room to lay a cap on it's side in the tough spot then choose a cap with an extra low ESR and do so.

1/2" to 5/8" longer leads results in raising the ESR by roughly .01 Ohms which compensates for using a cap with extra low ESR.

.

[PCBONEZ]

Originally Posted by stevenps

I remember reading somewhere that someone (PlainBill?) specifically recommended against using these for some reason. Also, I'm in the US and don't really want to pay for Royal Mail to get stuff to me; I think Newark may have those, but I'll wait for someone to confirm the suitability of these caps before I try to order them.

Nothing wrong with FK.
Newark gets them from Farnell.
.

What series are the 820uF 25v that are in parallel?
Are they the brown ones more or less in a row?
.

[PCBONEZ]

Never mind. Found it..

Capxon GL 10x20mm have an ESR of 0.50 ohms.
Three in parallel result in a total ESR of 0.017 ohms.

I doubt the caps are on 25v.
If the actual voltage is 16v or less that throws you replacement options wide open.
-
Shouldn't be too hard to find a single 2200-2700uF cap with ESR at .017 or less at the lower voltages.

So, what's the actual voltage there?
.

[stevenps]

Originally Posted by PCBONEZ

If you have room to lay a cap on it's side in the tough spot then choose a cap with an extra low ESR and do so.

1/2" to 5/8" longer leads results in raising the ESR by roughly .01 Ohms which compensates for using a cap with extra low ESR.

.

I did investigate that option. If you look at the picture, you'll see that the caps are walled in by a transformer, some big diodes, and the edge of the board. The only way to lay one on its side is to get one that has enough capacitance to replace all 3 of the 820uF 25v caps. I couldn't find a suitable one even as large as 12.5mm x 30mm, which is the largest size cap that would fit in that space. I posted earlier about the largest size caps that will fit in the available space on the board.

Originally Posted by PCBONEZ

Nothing wrong with FK.
Newark gets them from Farnell.
.

What series are the 820uF 25v that are in parallel?
Are they the brown ones more or less in a row?
.

I ordered some of those 820uF 25v Panasonic FK caps that are 10mm x 20mm, but they are out of stock. I don't want to wait the quoted 40 days to figure out what's wrong with these monitors, so I went ahead and ordered two monitors' worth of the smaller ones from Digikey. I have 4 of these dead monitors, so I can try the FKs should they arrive sometime this year.

Originally Posted by PCBONEZ

Never mind. Found it..

Capxon GL 10x20mm have an ESR of 0.50 ohms.
Three in parallel result in a total ESR of 0.017 ohms.

I doubt the caps are on 25v.
If the actual voltage is 16v or less that throws you replacement options wide open.
-
Shouldn't be too hard to find a single 2200-2700uF cap with ESR at .017 or less at the lower voltages.

So, what's the actual voltage there?
.

Working voltage is around 13.4v, though it spikes up to around 18v for a few seconds while the inverter is starting up. I thought it was a really bad idea to ever exceed the voltage rating on the caps, so I didn't consider using 16v ones.

[PCBONEZ]

Yes but if you can replace the three in parallel with a single then you will gain some mounting space.
If the three caps are 10mm diameter you might be able to get a 30mm long cap in there on it's side.
.

[stevenps]

Originally Posted by PCBONEZ

Yes but if you can replace the three in parallel with a single then you will gain some mounting space.
If the three caps are 10mm diameter you might be able to get a 30mm long cap in there on it's side.
.

I ordered the 680uF caps last night, but I did look at Newark again and I see a 2200uF 25v 12.5x30 Panasonic FK cap (http://www.newark.com/panasonic/eeuf...25v/dp/32R9015) that might work if the extra 160uF is a really big deal. If I do that though, I cannot use two 1000uF 10v caps I ordered because they won't fit; I'll have to use the 1800uF 10v Chemi-Con LXY and separately order them. Is this pair (2200uF 25v Panasonic FK + 1800uF 10v Chemi-Con LXY) really that much better than what I already ordered (3x 680uF 25v Nichicon HE + 2x 1000uF 10v Nichicon HE)?

[PCBONEZ]

What exactly is it that the tops of the caps hit?

[stevenps]

Originally Posted by PCBONEZ

What exactly is it that the tops of the caps hit?

In what orientation? We've discussed a number of different ways I could mount the capacitors. If I just put 25mm tall ones in vertically to match the ones I removed, they hit the back of the metal mounting bracket that the IP and display boards are screwed in to. I have read suggestions from others about drilling a hole where the caps go, but I would really prefer not to go that route. I also mentioned in an earlier post that I spaced the board up with washers on one monitor I fixed, but I didn't like that because the screws just barely held the board in. Fortunately these IP boards have a plastic back cover that will protect the back of the board from shorting against the back of the LCD panel, but there is airflow space around everything the way the monitor is designed, and I'll be squeezing off that space somewhere if I try to adjust the spacing on the IP board.

[PCBONEZ]

Originally Posted by stevenps

I also mentioned in an earlier post that I spaced the board up with washers on one monitor I fixed, but I didn't like that because the screws just barely held the board in.

When that happens I get longer screws.

I did the clearance hole thing on one but there was nothing close to the other side of the plate.

I can't think of any new ideas at the moment.
.

[momaka]

I think the 25v, 680uF Nichicon HEs would be fine. The 820uF rating of the CapXon GL is probably bogus anyways. Reason I say this is because on the SyncMaster 931b I recapped, there was a 470uF CapXon GL of the same exact voltage rating and size as the two 25v, 820uF CapXon GLs. So how is that possible? Was the 470uF cap just a re-labeled 820uF cap that did not meet specs during production?
Well either way, it doesn't sound too good to me. Moreover, I've seen this inconsistency with other CapXon series, too, such as KM.

[stevenps]

Originally Posted by PCBONEZ

When that happens I get longer screws.

I did the clearance hole thing on one but there was nothing close to the other side of the plate.

I can't think of any new ideas at the moment.
.

It sounds like you think that using slightly different caps from what the manufacturer used is not a good idea. Just curious, but why is that? I believe I've read that the manufacturer uses whatever parts they can get cheap, and staying within 20% of the values of the original parts is usually safe. I've got good replacements for the 1000uF 10v caps, and 680uF is within 20% of 820uF (984 - 656). Is the single 2200uF cap a better choice than the 3 680uF caps? I have 4 (possibly 5) of these monitors that have the same problem, so I could try different things to see what works best. Newark says that my 820uF 25v Panasonic FK caps should ship by the end of the month, so I can try some of them if nothing else seems to be working.

Originally Posted by momaka

I think the 25v, 680uF Nichicon HEs would be fine. The 820uF rating of the CapXon GL is probably bogus anyways. Reason I say this is because on the SyncMaster 931b I recapped, there was a 470uF CapXon GL of the same exact voltage rating and size as the two 25v, 820uF CapXon GLs. So how is that possible? Was the 470uF cap just a re-labeled 820uF cap that did not meet specs during production?
Well either way, it doesn't sound too good to me. Moreover, I've seen this inconsistency with other CapXon series, too, such as KM.

I've seen others say similar things about many of the cheaper brands. For example I've heard that a 3300uF cap is a fine replacement for the cheap 4700uF caps as their rating isn't accurate. I'll use the 680uF caps on the one board I haven't touched yet and see what happens.

[PCBONEZ]

Originally Posted by stevenps

It sounds like you think that using slightly different caps from what the manufacturer used is not a good idea. Just curious, but why is that?

That isn't it exactly.
.
Total ESR of the paralleled caps should be kept the same or lower.
That is far more important than uF and I'm not sure you are considering it.
[Total ESR of caps in parallel is calculated like they are parallel resisters.]
.
In general I don't like to go down in uF because uF goes down as caps age.
You might start out within 20% but it won't stay within 20%.
- A cap 20% lower age-drifts out of the original's range.
- A cap 20% higher age-drifts in to the original's range.
For a temp fix that won't matter, but what if it doesn't end up temp..
.
And you say the circuit has voltage spikes.
The uF is what dampens voltage spikes.
.

[stevenps]

Originally Posted by PCBONEZ

That isn't it exactly.
.
Total ESR of the paralleled caps should be kept the same or lower.
That is far more important than uF and I'm not sure you are considering it.
[Total ESR of caps in parallel is calculated like they are parallel resisters.]
.
In general I don't like to go down in uF because uF goes down as caps age.
You might start out within 20% but it won't stay within 20%.
- A cap 20% lower age-drifts out of the original's range.
- A cap 20% higher age-drifts in to the original's range.
For a temp fix that won't matter, but what if it doesn't end up temp..
.
And you say the circuit has voltage spikes.
The uF is what dampens voltage spikes.
.

I did consider this. momaka recommended some brands and series based on ESR and ripple, and I think the parts I chose are fine in this regard. The CapXon GL caps (http://www.capxongroup.com/files/GL%...1288681976.pdf) are rated at 1280 mA ripple and 0.052 ohm impedance at 10x20 25v. The Nichicon HE caps I ordered (http://nichicon-us.com/english/products/pdf/e-he.pdf) are rated at 1400 mA ripple and 0.046 ohm impedance at 10x20 25v. Both are measured at 20 degrees C, 100Khz. I think that's 0.017 ohm for the 3 original 820uF 25v 10x20 CapXon GL caps in parallel, and 0.015 ohm for the 3 replacement 680uF 25v 10x20 Nichicon HE caps in parallel.

Given what momaka says:

Originally Posted by momaka

I think the 25v, 680uF Nichicon HEs would be fine. The 820uF rating of the CapXon GL is probably bogus anyways. Reason I say this is because on the SyncMaster 931b I recapped, there was a 470uF CapXon GL of the same exact voltage rating and size as the two 25v, 820uF CapXon GLs. So how is that possible? Was the 470uF cap just a re-labeled 820uF cap that did not meet specs during production?
Well either way, it doesn't sound too good to me. Moreover, I've seen this inconsistency with other CapXon series, too, such as KM.

and what the data sheet says about the physical sizes of the 820uF 25v caps in the GL series (there isn't one, but there is a 680uF cap at that size and voltage), I'm not hugely concerned about the rating on the original caps. Either CapXon custom-manufactured these caps for Samsung, their data sheet is incorrect, or these caps aren't really 820uF 25v caps. Given that the Nichicon caps I bought are better caps than the original CapXon ones, I would believe the rating on them more than I would on the original caps. Do you disagree with my reasoning, or did I do the math incorrectly?

[PCBONEZ]

How to find a mystery cap's ESR & Ripple:
~~~~~
*Within a Series* the ESR and ripple are based on THE CAN SIZE.
It changes with the *physical attributes* of the cap [including length, width, lead properties, type foil and electrolyte]. Volts and uF don't affect it.
*Within a Series* all of those are the same for all the caps - except the can size.

ESR/Ripple of good brands don't vary at all if you look at can size.
Crap brands jump around a bit because their caps quality does.

All you have to do is look in the GL table for the nearest cap [by volts and uf] that is 10x20mm and you will know the ESR and Ripple. [Spot on for good brands, nearly spot on for crap brands.
The numbers for the 680uF 25v don't follow the others and are probably a typo.
The 10x20mm 1000uF 16v GL is 1180/0.050
The 10x20mm 470uF 35v GL is 1180/0.050
Using 1180/0.050 makes sense for the 820uF.

That's how I got to:
Three 0.050 ohm caps in parallel result in an ESR of 0.017 ohms.

~~~~

If you are going to use one or two in place of the three you still need an ESR of 0.017 ohms or less.
[I think you got it and I'm just repeating.... Wife is home today... Yak yak distractions while I'm reading/writing.]

You can use this to do the math. [It's just easy...]
http://www.1728.com/resistrs.htm
.

[stevenps]

Originally Posted by PCBONEZ

How to find a mystery cap's ESR & Ripple:
~~~~~
*Within a Series* the ESR and ripple are based on THE CAN SIZE.
It changes with the *physical attributes* of the cap [including length, width, lead properties, type foil and electrolyte]. Volts and uF don't affect it.
*Within a Series* all of those are the same for all the caps - except the can size.

ESR/Ripple of good brands don't vary at all if you look at can size.
Crap brands jump around a bit because their caps quality does.

All you have to do is look in the GL table for the nearest cap [by volts and uf] that is 10x20mm and you will know the ESR and Ripple. [Spot on for good brands, nearly spot on for crap brands.
The numbers for the 680uF 25v don't follow the others and are probably a typo.
The 10x20mm 1000uF 16v GL is 1180/0.050
The 10x20mm 470uF 35v GL is 1180/0.050
Using 1180/0.050 makes sense for the 820uF.

That's how I got to:
Three 0.050 ohm caps in parallel result in an ESR of 0.017 ohms.

~~~~

If you are going to use one or two in place of the three you still need an ESR of 0.017 ohms or less.
[I think you got it and I'm just repeating.... Wife is home today... Yak yak distractions while I'm reading/writing.]

You can use this to do the math. [It's just easy...]
http://www.1728.com/resistrs.htm
.

I think we're saying the same thing. That's why I was looking at a 10x20 cap and not the 820uF 10x25 to get the numbers I was using; apparently I should have checked some of the others of the same size to make sure the numbers were consistent. I think this means that the 2200uF 25v Panasonic FK 12.5x30 that I was considering as a replacement for all 3 of the 820uF caps might be okay, as Panasonic claims that a 12.5x30 cap has impedance of 0.018 ohms, which is really close to the 0.017 ohms for the 3 GLs in parallel.

[PCBONEZ]

0.018 sounds close enough to me considering CapXon probably isn't quite up to it's advertising claim anyway.
.

[stevenps]

Originally Posted by stevenps

I hope so too. Since nobody had any better suggestions given the physical constraints on this board, I went ahead and ordered two monitors' worth of Nichicon HE 1000uF 10v and 680uF 25v. I put in a backorder for the Panasonic FK 820uF 25v from Newark, which claimed a 40 day lead time so I may see them sometime this year. I'll happily use better/different caps on the other two monitors, once I figure out why they're misbehaving. I will report what happens when I get my replacement caps in.

I received the replacement caps today. I ordered Nichicon HE to replace the CapXon GL and KM caps that Samsung used. I plugged the monitor in and it still squeals. When I hook up my multimeter and circuit breaker in place of the 3A pico-fuse for the inverter, I see current start around 2A when I power the monitor on, and it slowly goes up to around 2.2A. Then it quickly goes up to around 2.6A and something starts hissing. I shut the monitor off at that point and it goes back to squealing. Any thoughts on what I could check to track down the problem?

[seanc]

What are:
C107
C123
C345

Have they been replaced?

[stevenps]

Originally Posted by seanc

What are:
C107
C123
C345

Have they been replaced?

C107 = CapXon KM 47uF 50v
C123 = CapXon KM 10uF 50v
C234 = CapXon ST 33uF 25v
C113 = CapXon KM 330uF 10v

None of those caps have been replaced. I have Rubycon ZL and Panasonic NHG 25v caps I could use to replace C113, but I'd have to order replacements for any of the others. If it matters, none of them are visibly damaged.

[seanc]

Can you use a toilet/paper towel roll, to see if the squeal is coming from any of those caps?