FAO: Isaac and any other electronics people....

[Kieran]

Peeps...

As some of you may have gathered, GLS Power has once again returned to KieranCorp Towers.:afro:

However, this car has the common 'leaky capacitor' fault on the ECU board. Now, on this car it appears, judging by other photos I've seen, to be in its early stages. The board hasn't melted through and the car does still run, it just idles a bit funny and the CEL is on - Oh, and it's a complete pig to start from freezing cold.

Problem is I don't really know enough to judge if this amount of damage would be remanufacture territory, or whether or not I could just get a fresh set of caps soldered on and off I go? What do you lot think?

(Click to enlarge the picture - I've circled the damage).

uploaded/326/1207777869.jpg

This is the worst of the four. Two are undamaged, one has leaked slightly... It has oozed out underneath the cap, but not got to any other components.

Personally, I think it looks okay, but I am concerned about that resistor (R63). That's surrounded by capacitor ooze - I don't know what damage that may have done?

[Kenneth]

So long as the tracks are all ok, it shouldn't be a problem

This is how I would do it.

I would go down to your local electronics store and get some spray for cleaning up the circuit board. Forget what it was called when I got it... Also get some spray on stuff for sealing the circuit board afterwards. (as well as getting the capacitors. It may pay to replace that resistor while you are at it, just in case. If you can't get hold of a replacement, then don't worry tooo much about it)

Remove the damaged components, clean up the affected area thoroughly, solder on the new components, clean the area thoroughly, then spray the sealant over it all.

[AncientOfMu]

K dont do anything till you have spoke to me on msn or similar. I can fix that for the price of postage. I can also return it to you the same day.

That cap is a standard 1206 100nf de coupling capacitor. Im fitting and removing those plus other stuff all day every day.

Give me a shout.

Don

[Davezj]

the the radial caps (big blue sticky up ones) can be replaced like for like if that is going to help, they are through hole caps so a basic soldering iron, solder sucker or braid is all you need to desolder these. but that surface mount resistor (R63) looks like it is burnt rather than cap leakage.

the radial caps are electrolitic, (pos and neg side to them) and these do have life expectancy. so they do burn out after a certain time.

If you want to replace them then take the ECU to maplin and say you want replcement for these caps and they should be able to supply them for a few quid and get someone to fit them or get someone to get them from RS, Farnell, or the like.

as an aside When someone talks about cap leakage as being a problem it doesn't generally mean there is actually oozing stuff coming out of them. It generally means they are not holding the change they should and it is leaking away quicker than it is desgined to, the rough dicharge should be 2/3 RC value. the R being the input resistance of the meter being used to test the cap. the RC value being measured is in time. this would mean the cap would need to be removed from the board to test it properly, and a scope would be handy.
Hope this makes sense.

[I-S]

Give me a shout K - I'll clean up, replace caps that need replacing, ultrasonic clean (flux is corrosive over time), etc. I have everything in stock or available.

[MarkSanne]

No comments as everything has said/offered before. Only thing is a picture of one of my 6G engine ECU's that was a lot worse, just for fun (was repaired at a local shop for 35 euro). Btw: I also had a faulty stepper-motor what caused strange idling behaviour.

Before:
uploaded/1887/1207814846.jpg

After:
uploaded/1887/1207814860.jpg

[miller]

K,

You've been lucky, like the guys said before nothing a circuit cleaner and caps replaced wont solve! It will run even better than it already does!




Mike

[CANDEE]

6G's are known for doin capacitors... Ive even seen EVO 4's have the same problem :(

[Davezj]

No comments as everything has said/offered before. Only thing is a picture of one of my 6G engine ECU's that was a lot worse, just for fun (was repaired at a local shop for 35 euro). Btw: I also had a faulty stepper-motor what caused strange idling behaviour.

Before:
uploaded/1887/1207814846.jpg

After:
uploaded/1887/1207814860.jpg


wow that looks nasty,

i take it, G8's don't suffer the same fate over time. look like it might be worth a check in side as i have my ECU out on the passenger footwell at the moment.

[Robotnik123]

as an aside When someone talks about cap leakage as being a problem it doesn't generally mean there is actually oozing stuff coming out of them.

Well I have an old AthlonXP motherboard where several of those large round type caps actually have burst tops and there really is stuff coming out of them. Interestingly that mobo can still boot and run WinXP (last time I tried it which was a while ago), but other times it will hang at POST.

[I-S]

Davezj was meaning that "leakage" is an electronics term that describes an electrical characteristic of capacitors, distinct from physical leakage of the electrolyte in the above example.

Kieran - send me a list of values (uF) and voltages. I'm considering what the best parts to use in it are.

[MarkSanne]

I haven't heard of bad 8G ECU/Capacitors, and I think Mitsi would have learned to use better capacitors when they found out their 1987 - 199x ECU's needed to be replaced a lot...

[Davezj]

That's good.

but it is always nice to have look around and take some pics. does anyone know if you can open them up with out damaging them. screw holding the lid on rather than folded metal seams.

yes i did mean cap characteristic "leakage" in the bottom part of my post.

[I-S]

Mark - it's not a matter of "better" capacitors. It's simply a fact of electrolytic capacitors that this happens to them after a while, particularly in an application subject to vibration and heat. They have lasted more than 15 years which is a reasonable expectation for the lifetime of a car. It's not really practical to use alternatives to electrolytic capacitors for values over 10uF.

It will happen to 8G units soon - the first 8gs are over 10 years old now.

[Davezj]

mine is 11 and a bit years old, so i will try to have a look at mine while it is out. i would imagine you would get that cap burning smell long before they actually blew, and i don't have that. but they are sealed units so who know maybe not.

[MarkSanne]

Isaac, is this really so? I mean, for example there are tons of mainboards that have been working for decades without any capacitor leaking or bursting, while some specific model/brands mainboards have it guaranteed. I thought I had read about this, where it was concluded that inferiour quality capacitors were used...

EDIT: btw, I'm not questioning your knowledge on these matters which undoubtly far outrank my skills, just popped up into my memory and made some (bad?) conclusions/suggestions?

[I-S]

Isaac, is this really so? I mean, for example there are tons of mainboards that have been working for decades without any capacitor leaking or bursting, while some specific model/brands mainboards have it guaranteed. I thought I had read about this, where it was concluded that inferiour quality capacitors were used...

EDIT: btw, I'm not questioning your knowledge on these matters which undoubtly far outrank my skills, just popped up into my memory and made some (bad?) conclusions/suggestions?

There's several different issues involved here...

1) Capacitor temperature rating. Commonly there are 85C and 105C rated electrolytics - in both computer and automotive applications 105C parts should be used.
2) Usage. What causes problems with capacitors is repeated temperature cycling and voltage cycling. In a computer server which is kept powered 24/7/365 and reaches a steady temperature state as well as keeping a constant voltage then there's little stress on the capacitor and it will last a very long time. In a car, it is cycled from extreme cold (left overnight or for a week or whatever in frozen canada or norway etc) to extreme heat (engine working, climbing a hill in death valley), and from 0V to 12V to 14.4V to 12V to 0V and back again. This stresses them much more and causes failure as observed. It's also the reason behind a much higher rate of failure in desktop PCs for dead capacitors than in servers (desktops are switched off, servers aren't).
3) Capacitor quality. As you suggest, cheap/bad caps will fail faster. For many years the term "computer grade" was used to signify the better quality capacitors. Computers have driven the development of higher temperature, lower ESR, higher density and generally better electrolytic capacitors - Sanyo's OSCON series being a prime example.

[miller]

Capacitor leakage is best described as a very bad 'fishy smell' when it gets real bad.
If your car gets this far to produce the fishy smell it could be bad news as the ECU may go beyond repair.

Little gremlins like erratic idle, stalling, missfiring and hard starting can sometimes all lead to the ECU especially if they are 10years old plus.

Its something that can be very very cheap to sort now (preventative maintenance) rather than a new ECU worst case scenario. Middle of the road scenario its a post off job to the ECU doctor for the realms of £80 to £180 repair bill.

On my Mitsi GTO which was 13 at the time the capacitors had just started to leak down the stalks, taking the motherboard to a local TV repair shop and not telling them what it was, just asking for like for like capcitor replacement across the board (excuse the pun)
It ran like a dream after that, after replacing the TPS twice to no avail on hunting idle the ECU job solved this instantly


Mike

[Turbo_Steve]

Agreeing with Isaac, except to add that you get a percentage of 'poorer' quality caps with every yield. We've found that when using high wattage portable stage amplifiers, which is probably one of the harshest environments for a high quality, large electrolytic.

Out of 30 amps over 2 years, you WILL lose at least one power cap. It just happens with pieces of kit that get banged about, that live next to massive bass arrays (sometimes) and that are deep cycled everytime someone uses a kick drum.

In reference to the 10 years comment, though, it is worth adding that the capacitors that last tend to be the ones that are mounted properly: the newer ECUs I've seen use "3-legged" caps, which hold in place much better, and are frequently arrayed in a line as far away from the evil heat generating switching components as possible, I guess with the intention of smoothing temperature changes.
Assuming the 8Gs have some (and other) design revisions, I'd suspect them to last closer to 20 years than 10. Most military equipment manages to avoid cap failure despite masses of abuse, so there must be SOMEthing they design in, even if it's little foam pads or something :)

[Davezj]

a lot of good info here.

but much as i expected. i will take a look at the ecu this weekend. it what it is like if i can actually open it up.

[Kieran]

Very worthwhile thread this - It's nice to do a bit of forward planning and risk assesment because we can, rather than because there's a need! :scholar:

[Davezj]

well just thought i would update this thread.

My ECU looks fine no leaking caps and it is conformally coated.

I had a look at my engine ECU and you will be please to know it does not have any of those leaking style of cap on the board, there 2 electrolitics and they look to be a better quality and are fixed with a pad between the board and cap to stop vibration.

i will post up some pic and part numbers later.

[I-S]

Kieran - Is it working now that Don has waved his magic iron?

Davezj - I'd be interested to see...