'91 ecu troubles

[greenfield54]

Ok so this is what I did this week. I could not get to the car so I tried bench testing the ecu. I applied all power to pins 24,32 and 52. Ground to pins 2 and 3. I then monitored pins 38 and 47. Both has a steady output of 12volts. No wagging whatsoever. I tried switching on and off power to pin 24 but no go. Of course all sensors are disconnected at this time. Will try this again with the ecu installed in the car. FYI.

[fixkick]

the cel pin must flash.
on the bench you need a lamp or led on the CEL pin,
or the meter, in volts mode
the meter must be connected to the +12vdc for it to wag, just as the lamps does, the Ecu can only ground the cel pin,. is why.

I think this Germane ECU has special way to get the CEL to work.
one car a 1998 has knob slot on side of ecu, a switch there that turns with a screw driver, odd but real, some hondas did same deal.
there iS NOT stardards for OBD1, each maker was free to do , yes really odd things.
your car it shows the CEL lamp is wired, but missing, that means they are so cheap , to save 10cents on a bulb then, so the tech has to connect his.... how lame is that?
some ECU have leds on side port of ECU, (again seen on some hondas (old EFI)
if i had this car, id not touch it lacking a matching FSM book, id buy one, it be first day thing,
10x that outside USA< (our cars are documented well here)

here is a quote proving my points above.
"Some vehicles such as pre 1998 Mitsubishi's and Hyundai's, plus some Kia and Mazda vehicles do not display codes on a flashing dash light.
These vehicles give their code signals, directly TO the diagnostic connector.
You will need to use an LED tester or a Voltmeter to read the codes from these vehicles. "

some posts show jumping pin S(ground) to pin Q (DIAG)
PIN T (LAST) IS MY GUESS ON LED PIN,see drawing below.
this allows the LED to flash codes. (a 12v 0.050 ampere (50mA) lamp works with no resistor, too)

i read all over the world to find this. this is my best guess, not having a real FSM. BOOK.
if the left pin B ? is 12vdc key on, (Power, connect the LED ? to there) then great ! we have some power to use.
if the T port is dead, jumper-ed. like this? drawing.....

then move the LED minus lead, to the engine fail pin P. see if that flashes, ?
codes:
http://www.troublecodes.net/Kia/
omg, what an odd ecu with long PULSE rules. the long = 10 short. what silly complexity that. and no good codes (im good)
watch out , ive seen some ecu's, send out infinite short pulses, and means IM OK> (like a snare dummer, didiididididididididi)

best to do this with 1 sensor unplugged. in order, if easy IAT, ECT, map or MAF (if have one) or last TPS. so you have hard failure key on,
no need to crank it, to get LED light working..
once we have flash working ,then we can fix this. problem.

---

sorry missed your older post (fuses)?
yes, ive added fuse to injectors and spark drivers.
10amps slow blow.
injectors are like 1 amp pulses. 10 ohm injectors I = E/R
same with spark. or 5amps. for 5 milisconds. so dont take to big a fuse.
if the coil shifts the fuse blows, before driver transistor blows to hell.

---

im sure we can get this to flash, if it is NON OBD2.
pretty sure..
but i can not read Korean...... LOL

i bet B pin is 12vdc (battery Key on) the Korean shows, B+ (battery plus) making the led more easy to fit up.
the ECU will only sink current (an open collector driver here is "traditional" or Open Drain MOS FET)
it can only sink current never source it,.
think of the CEL pin as a toggling switch to ground. at a 50mA max current spec.
if S pin is not really ground try R.
OBD1 is fun, no?????? LOL !

[greenfield54]

Yup this ecu is weird to the max. Big headache for me but I want to learn so this saturday,I will use a 10 dollar scan tool and see what happens. I will also recheck the reference voltages (5 and 12vdc) including continuity of wires to the ckp/cmp.

[fixkick]

really , the 1st question, is
is car , obd2 or not, EOBD, or JBOD, 3 different standards. (simular )
or is the old OBD1 as common in most countries before 2001. (my guess is this , but could be wrong, ive never seen any Siemens ecus.)


since you dont own a real FSM on the car.
you get to jump though vast hoops, and guessing. sure its a headache.
but we dont even know (if obd1) the specail rules for starting of DIAG mode, in truth each company uses its own rules. on that) can be a big crazy the steps.
one car, i saw was in diag mode full time, you just connect a light and look. (see? how unique that is:?)

[greenfield54]

OK this is what I did today:
1.Rechecked all voltages and grounding (floating and chassis). All OK.
2.Checked continuity of all HV cables. One has very hi resistance (almost open).
3.Earlier my son installed a new ckp sensor. I tightened the bolts some more thinking the air gap might not be 1.1 to 1.5 mm. This was the last thing I did.

Result. The car started to everyone's delight. There was a lot of smoke. My son in law apparently overfilled the crankcase by a full 3 liters. But take note that before I started troubleshooting, I already drained some of the oil. It did not start right away after I got the sparkies back.

[fixkick]

hv cables that open, are usually carbon junk cables.
get mag wires, from NGK or Bosch, use them. open HV wires, 5000 ohms/ per foot is ok, that is carbon spec . (mags read zero ohms but have INDUCTANCE)
ok amazing the gap was wrong, those ckp coil type sensors are very critical to gap (and may also fail when cold and cranking is slower..)
it needs to be gapped correctly on all cars like this, coil VR sensors.

ok now we know the scan tool would have been crying DEAD CKP all along... and that the secret STI transistors are good.
if one 1HV wire opens you loose 2 cylinders. and it can flash over add do damage to ECU.
at least put good wires there. and win.

congrats, you got it running, all that matters is that....

[greenfield54]

OK I got it running but something keeps popping in my mind. I saw a video on you tube by scannerdanner and we had the same situation (no start/no spark). He used his test light to touch the leads leading to the ckp sensor and the car suddenly sprung back to life. I did the same thing after tightening the bolt but in my case the ckp was not connected. But then I really donot know. I should have tried to start the car after tightening the bolt before I tried the bypass test. Check out the link below.

https://www.youtube.com/watch?v=EBABtCi_YQ4

[fixkick]

30 min vid to say one thing, unplugging CMP and putting it back fixed it, sure the pins were dirty or CMP is loose (gap issues) or is weak and marginal.
most are marginal , when cold and starter is slow, most are, if you scope it. (and will be worse if the CMP tip is dirty,(metal dust)
the HV spark test, is wrong tool and hand held is irresponsible to any viewer... ever get hit with 100,000 volts yet, ?

[greenfield54]

You said.
"He fails to know, that the VR sensors are just a coil of wire(pulse generator and is floating) and that his tests to bypass, wrong word to use
is really , him trying to simulate a sensors. simulate . (can link you to pages that do this with a tool, we had to make one from scratch)".

The video is clear that he knows what VR sensors are including the floating ground. I can use the link you can give me so I can make a pulse generator from scratch. I cannot invest in a good scanner that will work on most cars we have down here since I will only use it once and keep it for years. Bad investment but if I was in the business of fixing cars then that would be nice.

You are right. It is not wise to use a light tester to trigger ecu inputs. But then scannerdanner justifies what he did. Check out the link below and tell me what you think.

://www.youtube.com/watch?v=gEQnKNLlcFQ

[fixkick]

it's all moot now, car runs
its not wise to hot wire any ECU, (means putting 12vcd on a pin you have no idea what's on the other end, in most cases , no one does, there is no schematics on this) using and incandescent lamp as the current limit is crazy, it can be 250mA or more, for the lamp)
how ever tools can be made that can generate pulses, that are safe, a 0 to 4v pulse is safe, and current limited to say 10mA. (max TTL levels and currents)
the tool may work or not, that is because the ECU on most cars, (for sure newer) fail to accept signals out of time. and for sure weak levels, on all.
How the inputs are connected on the ECU side, for a CKP is unknown by anyone. (for sure on potted ECUs.... super hard to see)
it can be 1pin to ground and 2nd to the amp .
or the amp can be a differential amp and no grounds at all on the input. full floating inputs.

nor is that HV spark test, every step there he did is irresponsible. (wrong tool and holding it. like that)

there are better videos. by ASE masters.