there are lots of parts in all ECU, any can go bad. Any !
what matters is what is common..... (in most cases)
caps are the all time #1 failure, 1989 to 1995,, never seen them fail after that.,(they changed brand makers.... is why)
the diodes only fail when folks hot wire the ECU, this is number 3 fail. IMO
number 2 fail is injector driver transistors. (open, shorted or weak or leaking....)
one other failure is that 1 of the 2 injector drivers is shorted, causing out of control injection.
the top and bottom traces are very easy to fix. the center (multi-layer PCB) traces are not.
and those pen trace makers, do not work for power bus. it's useless for this and the #1 failure is the corner cap and is power, (jumpers do work, done vast numbers)
i worked in factory for years, and repaired like 1000s of boards... seen every fail there is, (we took in field returns, too) 'I worked the super dog shelf, no one there would touch"
i've done many boards damaged by lightning (hopeless). and many other causes.. flooding (water damage) (crash damage?)
here are the hot wire fails. (nincompoops that love to hot wire do this and is common,)
1: injector drivers 1 or more. (the cure is use a scope and make sure all signals are perfect !)
2: +5vdc refr diode blown (blew the protector) and then blew up the trace, to this line.... (have photos of that in my slide show)
3: battery reversal, and load dump events blowing up those diodes.
4: the guy that loves to hot wire the fuel pump. their fingers slip 1/8" of an inch and bam, he just blew , this diode to H3LL. D102 here
http://www.fixkick.com/ECU/8v-tbi-schematic91to95.jpg
if the ECU is in Europe (guess loc) and has an Immobilizer. option, fail 4 above blows the chip for that. feature or a transistor (never touched one but its there , im told.)
5: any output the badguys hot wired, bam, blown up. (lots there are , an most chips that drive these or not made or sold now.... lots of outputs there are....
inputs almost never blow. (resistor protected (series) and cap filtered they all are)
one other failure im keen on is the corner car
it has hidden center of board traces on that cap.
the cap blows. and the acid, eats up the pad ,the repair guy attempts to fix this and ends of with the center traces dead. (it's not his fault,its acids fault....)
remove the cap , with bad pads, invites, damage to this critical feed through (VIA's there are called) and that ends the JOY.
keep and open mind to ALL parts on the PCB can fail... ALL.
it's electronics, and if you do enough boards, you learn that. fast. or never fix most of them.
assumptions are the killers to success, fixing PCB or cars..... never assume. check it first.
all great techs are skeptics (max)!
keep in mind my page is just for the easy , gedunk fails... (easy peasy fails.... only....) or easter egg fails.... look and guess...
gee why is this diode blown in half, or something that looks like BBQ hot dogs. or carbon....
or smoke soot on the lid...
no lie, careful visual inspection are #1
keep in mind, if you see smoked parts, that can mean , something else is shorted. caused that part to smoke (a victim) to blow.
we check that carefully before doing the next smoke test.
letting out the smoke again...
the lack of 12's means, the ECU is in the non documented "BACKUP mode" (book has like one line of text for this mode)
backup means, the main processor is dead , and is running under a backup chip or backup processor or some simple, hardware sequencer (super dumb logic) that gives. static spark and static fuel. or fuel only based off TPS outputs. crude and can be. but gee it runs kinda........ floods and all that..
Backup mode is very serious fail. the worst.
in the old days, the ECU would end spark and fuel. (OBD1 changed all this) circa 1987+
in a factory ,we had the bed of nails tester , interface and Major ATE test system.. (and software for board, test vectors)
we run the machine and it stops on a vector (failed step) and is live or the results of State machine failure. (complex logic pre steps)
the tech can loop on this failure.. and find the failing NODE. easy.
and then find the failing part.
lots of fun this,,,, all this...
testing on a bench is basically pure luck,,,, using a meter and scope. and some skill.... but mostly just luck or "common fail points checks"
what matters is what is common..... (in most cases)
caps are the all time #1 failure, 1989 to 1995,, never seen them fail after that.,(they changed brand makers.... is why)
the diodes only fail when folks hot wire the ECU, this is number 3 fail. IMO
number 2 fail is injector driver transistors. (open, shorted or weak or leaking....)
one other failure is that 1 of the 2 injector drivers is shorted, causing out of control injection.
the top and bottom traces are very easy to fix. the center (multi-layer PCB) traces are not.
and those pen trace makers, do not work for power bus. it's useless for this and the #1 failure is the corner cap and is power, (jumpers do work, done vast numbers)
i worked in factory for years, and repaired like 1000s of boards... seen every fail there is, (we took in field returns, too) 'I worked the super dog shelf, no one there would touch"
i've done many boards damaged by lightning (hopeless). and many other causes.. flooding (water damage) (crash damage?)
here are the hot wire fails. (nincompoops that love to hot wire do this and is common,)
1: injector drivers 1 or more. (the cure is use a scope and make sure all signals are perfect !)
2: +5vdc refr diode blown (blew the protector) and then blew up the trace, to this line.... (have photos of that in my slide show)
3: battery reversal, and load dump events blowing up those diodes.
4: the guy that loves to hot wire the fuel pump. their fingers slip 1/8" of an inch and bam, he just blew , this diode to H3LL. D102 here
http://www.fixkick.com/ECU/8v-tbi-schematic91to95.jpg
if the ECU is in Europe (guess loc) and has an Immobilizer. option, fail 4 above blows the chip for that. feature or a transistor (never touched one but its there , im told.)
5: any output the badguys hot wired, bam, blown up. (lots there are , an most chips that drive these or not made or sold now.... lots of outputs there are....
inputs almost never blow. (resistor protected (series) and cap filtered they all are)
one other failure im keen on is the corner car
it has hidden center of board traces on that cap.
the cap blows. and the acid, eats up the pad ,the repair guy attempts to fix this and ends of with the center traces dead. (it's not his fault,its acids fault....)
remove the cap , with bad pads, invites, damage to this critical feed through (VIA's there are called) and that ends the JOY.
keep and open mind to ALL parts on the PCB can fail... ALL.
it's electronics, and if you do enough boards, you learn that. fast. or never fix most of them.
assumptions are the killers to success, fixing PCB or cars..... never assume. check it first.
all great techs are skeptics (max)!
keep in mind my page is just for the easy , gedunk fails... (easy peasy fails.... only....) or easter egg fails.... look and guess...
gee why is this diode blown in half, or something that looks like BBQ hot dogs. or carbon....
or smoke soot on the lid...
no lie, careful visual inspection are #1
keep in mind, if you see smoked parts, that can mean , something else is shorted. caused that part to smoke (a victim) to blow.
we check that carefully before doing the next smoke test.
letting out the smoke again...
the lack of 12's means, the ECU is in the non documented "BACKUP mode" (book has like one line of text for this mode)
backup means, the main processor is dead , and is running under a backup chip or backup processor or some simple, hardware sequencer (super dumb logic) that gives. static spark and static fuel. or fuel only based off TPS outputs. crude and can be. but gee it runs kinda........ floods and all that..
Backup mode is very serious fail. the worst.
in the old days, the ECU would end spark and fuel. (OBD1 changed all this) circa 1987+
in a factory ,we had the bed of nails tester , interface and Major ATE test system.. (and software for board, test vectors)
we run the machine and it stops on a vector (failed step) and is live or the results of State machine failure. (complex logic pre steps)
the tech can loop on this failure.. and find the failing NODE. easy.
and then find the failing part.
lots of fun this,,,, all this...
testing on a bench is basically pure luck,,,, using a meter and scope. and some skill.... but mostly just luck or "common fail points checks"
http://www.fixkick.com