'91 ecu troubles

[greenfield54]

I have 2-wire ckp with the same part no. as yours. Haynes says the gap should be 1.5 mm. Hmmm.
that makes me wonder. I need to check that gap. It was my son who installed it. Brand new and with brand new o-ring. Does that ring a bell to you?
Maybe. Just maybe my son did not tighten the bolt enough for fear of breaking it as I often warn him not to do. That o-ring might not have been compressed enough such that the gap is more than 1.5 mm.

Just a thought. What is the advantage of using magnetic ckp sensors over hall effect sensors which have stronger/cleaner signals?

The timing cannot be wrong because the tone wheel is fixed rigid to the crank and there is no adjustment to move the sensor around. I will check it next,next weekend. Wish me luck!

Thanks.

[fixkick]

1996 Beta 1 engine. 1.6 to 1.8L.
they are cheaper, for the maker (hyundai) (no electronics, no amp, no driver inside, just a passive coil) dirt cheap..

some cars like this, i saw that they have this coupling device on the cam. that allows the cam sensor to work at all and fails or is put on wrong.
the ckp has tone wheel, that on some cars can fall off (yours looks bolted on big time)
some times the person puts in a used engine, and it has the wrong tone ring.
the non hall sensors the cap is important, as is a good ton ring and , the right ring, with the the missing tooth.
the non Hall sensors have very very weak signals, this is why the HALL was invented, to make the hall sensor output 4v signals at 1 rpm or 6000 rpm
see my video above going at 1 RPM?
the hall sensor is electronic, it has 3 pins, does yours, i keep asking, .?? are both your CKP and CMP both 2 pin sensors.????
2 pin bad.
3 good. see?
but you must deal with what Siemens gave you,

this all all backwards.
first we scan it to see if the ECU is throwing CKP or CMP errors. after all. its the ECU that needs to be made happy, so asking it , tops our list.
if the ECU shows the CKP bad, the car will never start.
if so, then we scope it (after cleaning its mag tip, they can bet dirty, with metal particles over long time)
then replace it, if that easy step fails. that leaves but 2 more steps
scoping it. does it make 1 to 2 volts cranking ,yes, no, if yes, then the ECU is bad.
if no? the wiring for the sensor is bad.

the hall sensor makes perfect digital wave forms at all times and speeds.
the non hall coil sensor does not, it just very weak at low speeds, ok at idle on up.
its just a coil of wire and magnet. at low RPM Faradays laws, make the voltage very very low. (too slow)

---

http://www.fixkick.com/videos/FLV-all/showtime.html#CMP

this is a 1991 cam sensor, that works at 1 rpm, even slower, even 1/10 of a RPM. by hand.
see the LED flash.
that is a HALL sensor.
the NON HALL can never do that ,ever.

the Siemens company if they they used 2 pin Non hall sensors (did they) for both CKP and CMP?>?????
if they did,
then the ECU has the electronics in side IT to do all that.
it has a powerful amplifier with a zero crossing detected (the true names) and this makes the cheaper gutless non hall sensors work.
but the gap needs to be right.
the wires must be perfect to transport those weak single to the distant ECU, like what 4 feet distant?
have you checked each sensors wire, for damage and continuity end to end?


advantages of hall.
1: costs more, oops not advantage.
2: strong perfect signals at all RPM , even in ALASKA, with near dead or weak battery.
3: no fiddling with GAPS.
4: vastly easier to test, with even an LED, lamp.... see mine. flashing????

dis-advanges of HALL
1: takes 1 more wire to make it work (power) so wire costs are more, say 10cents? again is your a 2 wire or 3 wire sensor.
2: takes 12vdc to the sensor, or 5v or 8v , depends no maker of car. but takes power there, if power is lost to there, the sensor goes DEAD.

---

---

ok the 3rd pin is ground. on CKP
if the schmematic is accurate.
the CMP is HALL
and the CKP is non hall , a VR coil sensor its called. coil1, coil 2 and ground.

[fixkick]

how ever we can test the cam sensor with any analog meter, (needle type)
key on, CMP pin 3 must be 12vdc, if not fix that first
key on, connnect meter black to ground (metal on engine or battery neg lug)
meter red lead to pin 2, above. pin 2 is the shielded signal line.
we do a thing called back probing, use a needle or pin and back probe the CKP
pin2
crank engine and it will go 0v, 4v , 0v, over and over, this is the CMP working., a true hall sensor.
this dont prove its timed right but proves it works and is spinning.

---

the ECU has a flash pin, fault lamp pin 38
connect thee red meter lead to pin 38 on ecu, using backprobing method (all tech can do that, google how to back probe car wiring)
key on, then needle may pulse does it? is there a pattern to it, yes, BINGO.
if you can get the needle to act as a lamp on pin 38 , you can crank the engine for 5 seconds.
release key and read the failure codes on pin 38.
like flash flash, pause flash, long pause repeats.
just like a toyota in 1995, same deal. but on the toyota the lamp is built in to the cluster, but no on your car, its MISSING , so says the schematc.

id do all i could, to learn how pin 38 works, sure would!!


this is the way to back probe.
http://www.fixkick.com/INJECTORS/NOID101.html#probe
we dont use that method here, banned in 1995 it was in uSA.

[greenfield54]

My CMP is 3-wire so that makes it a hall type. The CKP is 2-wire.
Yes I checked each wire including the shield. There is also 5 volts reference voltage on all sensors. My biggest concern is the shorted ignition coil which may have damaged the driver ic. Am still waiting for STI to feedback me on my queries.

Pin 38? I did not know that an OBD2 ecu has a flash pin. Now I can have fun. Thank you. I found the link below for the dtc. Is this ok?

http://www.totalcardiagnostics.com/suppo...dtc-manual

But then I need flash codes not just trouble codes since I do not have the proper scanner.

[fixkick]

the method to see if its HALL is the correct schematic.
the one you posted shows the ckp has 2 coax lines the means its COIL VR type,non hall. 3rd in ground.

the other sensor cmp has 3 wire but 1coax. and has power wire to it, no VAR sensors have power to them.

STI may never answer, my guess is they get 1000s of emails a day like that. and are bound by contract not to answer. (with truth, so why answer all them.?)

you only guessing, why the coils are both dead? this to me , means, the ECU is cutting spark. on purpose, not 2 bad drivers. (but i dont know history)
usually only 1 coil shorts, and blows 1 driver
if the owner tried swapping wires (ugh), on coils (ugly event) you then get 2 blown drivers(history matters)
but not normally, a user finds one driver dead and buys a new ECU./coil pack and end story.
those chips are contract chips (driver transistor buffers) and are trade secret..... ( some top ECU rebuilders pay Siemens for the facts, $$$) )



I know one guy that took his to a failure lab ($$$$) and they used nitric acid drip (a pro. machine) that stripped away the plastic case
and the lead frame was exposed, and then the chip is exposed, then with 100x microscope, he saw words on the DIE! (real name of chip is DIE)
they some times are clear., on types. others, are more secret codes. "Buf1234" or the like. but some use internal Makers codes, only the engineers know....
so... there you go.. we had such a lab at Phillips Semi. back in 1991. "customers asked us why it failed, and we answered in depth"
examples:
https://www.google.com/webhp?gws_rd=ssl#...ing+photos

what if both chips are ok. then what? say you found the real names and replaced them both, then what? what direction then?
what if the CPU sees errors on either sensor. and you don't know that>? called barking up wrong tree.

that is why SCANS are first. so the tech knows, what's UP. (ALL ECU' tell you when these inputs are bad)
you could replace all parts 10 times, and it's still dont spark and then find that one CKP coax is damaged in the center of the harness.... oops. you missed the scans telling you CKP is bad.(or intermittent)
the signal is bad.
or on my Jeep the ECU can tell me, the cam and crank are way out of sync. and is cutting spark. (a pure mechanical error inside engine.)
see?

all this begs to ask?
2 coils do not short the same day.
there are 2 coils there, and are not connected to each other. (save power or ground)
one cant short then magically short the other , how could it do that? unless it caught on fire....?

the ecu is just a black box of logic.
with logic boxes (CPU, brains) we first check inputs. not replace outputs until we know the inputs are correct.
after all you can do more damage attempting fixes not bad, then good.

like the FSM states on all cars on the ECU page.
"never condemn any ECU until all inputs are checked and all outputs for shorts."

i think a 10 $ scan tool works on this car.. ever seen an ELM327 based scan tool, for 10 bucks.
connect it up, see if the ECU will scan. keyon, see TPS work, just KEY ON... all do.
if not the ECU is dead or the ECU is not using OBD2 signaling protocols..... its 10 $ check. to find out.... not too expensive

What's in the box, i dont know, only Hyundai knows this. or the FSM book sometimes flat tells you. (on the SENSOR pages) (using OBD2, or?)
I'm told that there are scan tools that work for this box.
you could remove one chip
and then scope the input pin, and crank engine, and see if the spark inputs to the chip are good, if not good, its barking up wrong tree time.

if the inputs to the chip are bad, (spark driver chip) id bet the ECU is cutting spark on purpose...
but that invites wrecking the removed chip that nobody sells...
for that fact alone, id do all i could do to scan it first.

scan it.

[greenfield54]

As I said before this is what happened:
1.Car suddenly ran roughly.
2.Found only 2 cylinders firing. No spark on 2 plugs.
3.Then no more spark on remaining 2 cylinders. I did not exchange connections.

What I did:
1.Pulled all spark plugs and checked for color. Two plugs normal. Other two plugs no sign of ignition. Very clean.
2.Pulled out ignition coil and bench tested using external bench power supply. No spark and drawing very heavy current. In fact I tested it again using a battery charger. Battery charger over scaled.

So my conclusion is shorted coils.

[fixkick]

so why would the ecu, blow up the second coil?
first actions for spark for me are:
are there 4 new spark wires, and 4 new plugs. (are those wires crappy carboln junk wires, or nice NGK mag core wires, or bosch mag core.>???)
if 1 wire opens then 2 plugs dies. see?
and do you know how to test wasted spark?????

1: unplug the 2 new HV wires, connect them to 2 new gapped plugs, ground both plug shells to the engine. (1/4 pack)
crank it, got spark now? this is the only true way to test wasted spark, with known good loads, a load is the 2 wires new and 2 new spark plugs...
do not use screwdriver checks, this risks overloading the coils and for sure those 2 very rare driver chips... why stress, parts that cant be bought?

2: then repeat on pack 2, ( 2/3 pack)
new wires and 2 new plugs. gapped, to spec. cranking.... (it really is 2 systems here, 2 spark systems.) Only the ECU can kill both, in most cases.

are both really dead now?

again, if 1 HV wire is bad, you will lose 2 cylinders, that is how it works, same if any plug will not fire. same results. 2 dye.

if both SIDES (x4) those fail, then scanning is next.

most ECU will also report, besides CMP,CKP dead, it will say the IGNITOR IS DEAD! most do. 1 or both.

is not scanning that makes all this so hard, id find a way to scan my car.
or buy an ECU and pray it fixes this... it may not. (working blind is like that, causing guessing and failures)

spark plugs can fail many ways, ever use a spark testing machine, we had one in school. and did so under pressure, (simulated combustion pressures)
1: open , circuit , fail (kills to cylinders dead)
2: SHORTED, MORE COMMON. (only 1 cylinder dies, the shorted one)
3: Misfires. (can cause both to fail if one plug fails to spark.) they are wired in SERIES after all...
4: misfires only under pressure.

that is why we do spark tests with a box of new plugs in a box, called test spark plugs, they never are used in the engine, ever. we keep them fresh in pairs.
see why? (to prevent dog chasing tail )
or we use our cute , spark simulator tool that has an adjustable gap, in pairs. ever seen that? you set it to a gap that matches you cars spark energy levels.
1: lawn mower, 10kv
2: 1960 ford, 20kv.
3: first electronic 50kv (my guess is yours)
4: today 100kv. high energy. the gap is clearly marked there... new cars run hug currents and coil turns ratios in the coil. )
we can not only check for spark but see just how good it really is.... quality checks.

[fixkick]

i have more rules for working HV circuits, (i worked on huge military systems with this, so...)
the HV circuits.
should not be allowed to go open circuit, (or short it to ground directly ) 2 rules, keep the outputs loaded at all times the spark plugs have what is called plasma resistance. let it do that.
shorting a coil just asks for trouble. never short it to ground.... (with out the right gap)

this OPEN circuit state, invites, break down or flash over. (induction coils do that, they can spike to real high voltages , unloaded)
imagine if the coil was dirty on the outside.
then you unplugged the spark plug wire in any way, this causes a total open circuit state.

the Coil will and can flash over, what if that flash over, followed all that dirt or worse carbon there and ended up on the 4 pin primary pins.
this will blow the ECU to hell. if you are holding said coil you blow up.(SHOCKED badly)

can be many failures inside ECU now., many transistor blown up. ever seen lightning damage, same.

ive worked complex ATE systems that test circuit boards. that were lightning damaged, and replaced 10 chips to get it going again.
in some cases it fails again. due to the wounded effect, (we call that the walking wounded in the lab)

use real spark plugs (new) and wires to test coils, never short them or run them open and never have problems.

also running old spark plugs risks , the coils to breakdown, in fact it can flash over the the primary side and blow that up.
this is why turn ups should never be skipped.

here is good site that warns of these issues

http://www.aa1car.com/library/ignition_coils.htm

and never run carbon spark wire, ever.
kick those to the curb now,. they all have short short lives. and if open they can kill a good coil.

there are 3 wire types.
carbon
solid wire (never use this) its for old magneto cars. (race)
mag wire, this is still wire, but is Inductive, it uses Inductance to create a resistance, NGK sells it and all top Bosch wires us this, and is BEST in all cases.

both carbon wire and mag are EMI interference rated. (radio noise suppression rated) the latter is like $1 wire, and is seen on new cars due to its dirt cheap cost.
the later, is more expensive in short terms thinking, but not long term.
The carbon just hates vibration, guess who vibrates on cars.>????? everything/one.


see carbon junk here, at "fold back"
https://www.oldbritts.com/51_150102.html


flash over
https://youtu.be/xXAFQa0R9hA?t=9


id not use water, that dumb idea invites the HV energy JUMP and go to the primary pins and burn the ECU up.
but it is old skool. that is what it is.

if the spark parts are old, why not nut just do the tuneup, all parts , in this case, wires ,coil and spark plugs.
with distributors, cap,rotor,wires. plugs.

if the coil looks ok, no carbon tracking they are mostly ok.
keep it clean, on the outside and no corrosion in the lugs.

these are comments on how to keep the ECU running, after its fixed,

[greenfield54]

Actually the mechanic (not me but someone else) unplugged the HV wires one by one and flash over did occur as described by my son in law who was driving the car when it broke down. I found one HV wire open.

[fixkick]

never run the HV spark system unloaded, (open circuit)
on this car , it can hit 50,000 volts. and will jump where ever it wants, "least path of resistance" that can be you , not careful. or the ECU.
loaded the voltages are way less.
the spark gap , creates a resistive plasma and a proper load.

the problem with 50,000 volts present, is it will follow dirt , found on all used engines.. there will be dirt/grease films and that can take the Lightning bolt to bad places.
there are no igniters made that can handle that damaging hit...

check out the Tesla coil here.
see those ladder feet ?
a brave guy. here.. im sure OSHA rubber short pants...


graphic demo of electrons, seeking out ground..... and does every time,,,,, (and is un predictable in free space (open circuits)) air ionizes etc.
the plastic parts can usually take this hit, but it's dirty, it tracks the dirt and this can carbonize the plastic, wrecking it, as all distributor caps always did in the end....
there are no parts in any ECU that can take this hit,. not one. not 50kv. and it too can be dirty, and take this damage to where ever.....