For those with Car ABS and ECU problems;!..

[geoff]

Yes

[geoff]

In message , geoff writes

Its an ideal example of what happens when all these conditions that, who was it, dave ? stipulated, just don't exist

It also ignores the fact that plenty of electronics is designed to fail, just like other parts in various commercial goods where there is money to be made from spare parts

[clonet]

Join the UK Volvo Forum and ask questions there.

don't have an 850. I have two 240's. I have been a member for many years. The Volvo UK Forum is excellent.

[Dave Plowman (News)]

Anything that 'vibrates' or transmits some movement in any way which is soldered direct to the PCB is a likely failure area with time. Hence things like large transformers not usually being directly mounted. Where they are - like TV LOPT - is a common place for failed solder joints.

[tony sayer]

In article , clonet scribeth thus

Yes seen that, but its only since the unit foxed the local mech's that it became an issue. Never needed any other questions or things doing:)..

[Man at B&Q]

Not at all.

Which followed on from the OP. I'll let you lok up the title.

It was a good generalization anyway, not restricted to car ECUs in any way.

MBQ

[Man at B&Q]

Man

Care to provide proof of that?

I'll accept that things are not designed to last but you seem to imply malice.

MBQ

[The Natural Philosopher]

Never a hard thing to do.

Especially with electrical.

These are 'replace and see' merchants. They know little and understand less.

[The Natural Philosopher]

usually a different packaging as well. Ceramic packages and special glues take more extreme temp cycling than the trad epoxy style packages.

Mil spec is less about chip design, and more about selecting more tightly. using more conservatively, and packaging better.

A lot of early car electronics failures were down to the fact that designers simply did not appreciate how hostile and environment a car engine compartment was. Ambient temps way up. Or way down overnight in frost.. high G vibration, corrosive water and salt spray..voltage spikes well over a hundred if any part of the ignition fed back into the supply, and potential for high induced voltages on all inputs of any impedance due to ignition, and starter motor EMF radiations..

It's a miracle it works at all.

[The Natural Philosopher]

The other good one is a can of freeze spray.

Any area of the board that creates or clear a fault with application of freezer, is almost certainly suffering from a dry joint..

I resoldered ALL the eyelets - over a hundred - that were the substitute for plated through holes - on one telly.

Fixed all the problems that banging it cured. That was after resoldering half a dozen ID'd by freeze spray fixed SOME of the problems.

Totally my experience as well.

However the ones that I used to get handed to me were a skewed sample: they were the ones that could NOT be fixed by such techniques.

And were far and away the most interesting.

I recall one, which was a shorted polystyrene capacitor. Solder bath had actually melted it internally.

Showed up by prodding it.

Another power transistor failure was due to getting hot and the actual lead out wire on the emitter expanding enough to short to the base..I sawed it open and could JUST see where a microspark had welded the base to the emitter. Completely off target spot weld on the wire, that JUST passed testing, but failed the first time it was put in a unit.

Other nasties are digital timings and propagation delays..Yuk. stuff that statistically mostly works until you get a hot (or cold) day, and a bunch of all out of spec one way.

Conservative design and soak testing is the way you get round this sort of stuff.

You HOPE that you have put the unit through worse than it will ever get in service.

>

[The Natural Philosopher]

I think its less designed to fail, than its designed down to a price, not up to a standard.

If uyou have a design that cots 90[p., that in general fails less than one in a thousand on the line, less than one in a hundred in the home under guarantee, the fact that 30% fail in ten years is of little interest to you.

The time alone needed to get an MTBF of 100,000 hours, and the cost, would probably make your product as uncompetitive as er - a Miele.

Its only when pople actually find the Miele is cheaper in the long run, that the pressure is on to do better.

It was teh case in cars.

German cars didn't blow up at 80 mph continuous. British cars did. Japanese cars didn't have dodgy electrics. British cars did.

So peoples expectations changed. They decided they were happy paying more for something that could in fact be used for more than a drive to the nearest garage for fixing.

[geoff]

Well - for starters

sticking to boiler pcbs

the real killer of pcbs is fans - the back emf destroying the contacts of the relay driving it. A snubber network, costing less than pence would mitigate the problem

also, e.g. Potterton use a smaller relay to drive fans than they used to do

there are two possibilities, mad or malicious

I could go on, but I have to go to a residents meeting in a few minutes

laters

[Steve Walker]

Tin whisker growth in germanium transistors? ;)

Actually chips can and do fail after a period of working perfectly, as the stresses of thermal cycling can break the welds on the fine wires connecting the actual IC to its external pins.

SteveW

[The Natural Philosopher]

No, that's unusual, especially in 'proper' packaging, because they are not 'tight'

Normal CHIP (as opposed to package) failure mode is dopant migration, causing the semiconductor to become less of a semiconductor: that's happens in a few seconds at 200C, a few hoursr at 199C..a few days at

198C..you get the picture. That and crackled seals leading to chemical poisoning are the two main causes of ageing.

I've never had a bondwire actually come off.

[Man at B&Q]

Regrowth of fuse link in fusible link PROMs and PLAs.

Tin whiskers are causing concern again in some spheres with unleaded solder.

Commercial grade FLASH memory in MP3 and the like is not so good, especially the multi-level cell versions.

MBQ

[Grimly Curmudgeon]

We were somewhere around Barstow, on the edge of the desert, when the drugs began to take hold. I remember The Natural Philosopher saying something like:

Rubbish.

[The Natural Philosopher]

Oh they did.

I was there.

I fixed em.

And modified them so they would not.

[Tim Watts]

The Natural Philosopher wibbled on Thursday 17 June 2010 10:41

Come on - "Lucas, Prince of Darkness" !!

We all did...

Some of that too :)

[Dave Plowman (News)]

A name coined in the US. Personally the very worst electrics I had on any vehicle was Delco - the brand owned by GM, a US company. On a Bedford van. Every major component failed.

Lucas made stuff to what the market would pay. Hence plenty of rubbish. But could make very good stuff too - like for Rolls Royce.

[Grimly Curmudgeon]

We were somewhere around Barstow, on the edge of the desert, when the drugs began to take hold. I remember The Natural Philosopher saying something like:

I was there too, you know.

In general, the quality of most British vehicle electrics was superior to Italian and Japanese, especially in the quality of the main wiring loom. The real problem lay with clueless owners, as always.

The Italian faults didn't generally tend to be the wiring (although I've seen a few FIATs on fire), but componentry and it cost an arm and leg to replace Magnetti Marelli parts, they also had pretty crap light and switch connectors /switchgear.

The Japanese were the first to introduce re-cycled copper in their mass-market looms, which led to all sorts of mysterious faults in the wiring in later years. It took a while for Japanese reliability to become established - early on, in the main period we're referring to, they were nothing special for reliability. If the cost of Italian parts made an owner blanch, the pricetag associated with Japanese electrical bits would lead to him passing out.

Fortunately, there were/are usually workarounds for a 'genuine' part.