Login Register

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Fuel pump problem-no three second prime or purge
the IG-coil fuse in cab box blew?.

IC COIL fuse #1 cause of blowing up is never ECU, (ECU uses FI fuse power) but EFI needs both to run car, sure.
a shorted spark coil possible.
but but,, the 2 oxygen sensors both run off the IG-coil black/white power feed,
this is the #1 cause of all IG-Coil fues blowing.
if you look at the two oxy sensors, they have 4 wires, and one is black/white (harness side on not 02 pig tail side)
these wires love to hit exhaust, with clamps for wires missing after 22 years of bad service.
and short to ground and blow IG coil fuse super easy.
see bottom left schematic, see O2 sensor pin 4? bingo shorted. super super common failure on all cars USA 1996 to now. (EPA mandated OBD2 here)

hot wiring the FP relay invited blowing up ECU
better (not) is to hot wire IG-coil fuse then bad wires short catch fire.

the ECU runs of FI fuse ,in fender box.
or the sensors heaters inside both 02 sensors shorted out, inside. oops.
cel lamp glows with key on, like it should

question Q106 is tiny transistor? Or looks like black FET style?

I did not see it clearly marked on your image. i will inspect again today (checked yesterday no bad caps no obvious damage no burnt crispy traces, no smell )

Next steps.
will connect scan tool.
check and correct or remove any bad hacks to wiring.
check TPS voltage.
will check grounds and voltages at ECM pins
any other steps or items to test?

stereo wiring-previous owner hacked. Plug still there but some wires missing, some cut.
Brake and signal wires hacked. relay under dash added to control signal and brakes from rv tow vehicle resistors added behind tail lights in back

thanks as always your experience and knowledge means a lot. you put a lot of time into your site. its very appreciated by many (me too!)
Ok i think i have found the problem. After finding 0v at TPS and MAP, and .45mv at ECM pin A5. I disconnected the grey/red wire from the harness at the ECM, and reference (pin A5) voltage popped up to a, steady clean 5.00v.

I have reference voltage back!!!

I have a couple questions for you, if you dont' mind.

Would a, short or disconnected pin A5 to TPS ( no supply voltage) cause the ECM kill the fuel pump relay ? Or is that going to be a separate problem?

Knowing that the grey/red wire to the TPS or MAP sensor is shorted, should I
A) run a new wire to the TPS and MAP,
B) pull the harness from the ECM to the sensors, and inspect/repair?

I'm crossing my fingers that the grey/red shorted caused the ECM to fail to energize the FP relay.

Thank you again for all you're documentation.
I've read near every page i found on the obdii 16v and it has been a blessing. understanding how the 16v circuits work and the ecm logic makes it much less frustrating
this is super bad, the ECU will not tolerate a short there.

Unplug EGR map first.
the the fuel tank pressure sensor (evap)
TPS never shorts out, REF, it just cant, with 5000 ohms of carbon inside, cant short.
so is the 2 sensors or simple wire short on 5v refr, shorted to body somewhere,
yes, having facts on this 16v , is great way to reduce stress , you you are on track, fix REFR.

a scan tool connect would be crying bloody murder on sensor errors, with loss of 3 sensors at once/
good luck. let me know how you progress.
Ok, it was a very long process, but i have reference voltage fixed, a working CEL and no codes . Codes properly show when a sensor if disconnected. (ie. Remove IAT connector, CEL light comes in and reports the bad sensor.

Fuel relay still does not turn on, ECM(pink) shows 12v when fuel relay should be working. ECM pink does not go to ground as it should to trigger fuel pump relay. (instead goes to 12v)

Also of note, the REV limiter is no longer working. ECM used to cut injection around 6.5k RPM. ECM no longer cuts engine at rpm limit. RPM can go above 6.5k(bad for engine i know). This rev limiter was verified working before fuel pump relay problem.

5v reference solution below**
The previous owner drilled through the lower drivers kick. panel and used a standard 3/8" bolt to secure the bottom door hinge. Great idea!!

The down side was that they drilled directly through the harness. They actually took the time to repair the damage to the harness however in there efforts they crossed the grey/red fuel pressure sensor wire with the seat belt grey/red in the body harness .

The problem was easy to find, but difficult to diagnose. I found continuity between grey red at ecm rear harness connector and the ground loop behind right signal light, but wires did not touch our cross due to damage.

This did not make sense because on inspection wires were not damaged. Further inspection showed the old owners mistake. Cross wiring body harness grey/red to ECM harness grey/red while repairing his drill damage.
Good news +5vdcRef is now good. ! (OMG harness hacks, )
yes the DTC , have 2 levels now in 1996 first year of OBD2,
level one is, dead sensors, open or shorted bam , you get a DTC error for sure, just like OBD1 did. just new DTC number codes P0xxx
level 2 codes (tests) are OBD2 new advanced monitors. EGR tests ( A and B two test) EVAP1 ,2, 3 advanced evaporative (gas tank fumes controls) and tests.
still dead ECU controls for fuel pump relay, if you had JEEP the JEEP shows, DTC for dead relay, every time, but not on suzuki's they are dumb , ok? still dumb.
how could you ever reach 6500 RPM impossible to do that, my guess is your believe that lamp analog cheezey dash cluster gauge, why do that,?
RPM is only accurate with OBD2 scans, page on scans show RPM, MPH, all sensor data, if the CKP is dead, then the RPM may lie. is the CKP sensor missing? front crank shaft bottom front?

the ECU will only kill dead the PINK wire , only if :
1: ECU lost all power to it (power pins see on my ECU page 1996)
2: that includes ground pins, POWER means 2 pins , 12vdc and ground .
3: inputs wrong, the only input that can kill that pink wire is , (real inputs logic) are the TPS pin, TP pin is 1vdc, foot of throttle, at 3.5 to 4v or more this voltage at cranking has new
meaning, it has the logical meaning of "UNFLOOD MODE" this most cuts all fuel so the engine can start flooded with fuel (this steps are show in the users guide)
this may (ive never tested this) but may cut the Fuel pump relay too. (if I had written the code, that is what id have done , to cure injector leaks)
the spark is still active with Unflood mode active. cranking.
I guess your scan tool is too old or cheap or ? to have Pending functions> push a button called pendings and see ALL THOSE DTC error that do not SET until 3 driving cycles>?

hot wiring any ECU is going to kill them easy as eating Apple Pie.
lean this about 1990 ECU, the outputs (unlike new cars ) have not output protections, the output is a simple biPolar (not MOS) NPN transistor, if you , others or bad luck shorts this
pins to 12vdc that simple transistor burns to a crisp in about 1 second flat or less.

on new cars we have smart transistors now, with very very fast asking short detectors (thermal) and that turns the transistor oFF, savings its own bacon .
forget that, your car is a relic in this new world we have now.

let me speculate further (I do not have the firmware code from inside your ECU to speak more authoritatively, ever, I think nobody does, sill alive or not retired in Japan) or?

inputs, that must be good for FP relay to work. and not bad bad ECU.
1: power, not missing, told above.
2: CkP crank sensor, dead, but I have seen the car start and run with no CKP connected, but gee, again not seen every firmware revision on any 96_+ ECU, so.....
3: CMP cam sensor in dizzy, kills spark dead, so if spark is dead, the FPRelay will be dead, and by law, this is the no crash and burn code (firmware in all cars made EFI) check spark 1st ok?
cranking check spark
4: test the FP relay cranking, who knows maybe your car has some queer software running that acts like our 1990 sidekicks all do. you must CRANK it to test spark and fuel pump actions.
5: TPS inputs wrong, The MAGIC TP PIN is 1vdc no where near 3.5v or above, 5v is max or what ever 5vref is,4.75v is MIN for REF pin. Vref max is 5.25v, per the FSM, pages. and is true.
6: I can not think of any thing else that can kill the Pink wire dead, most times its hotwiring the FPrelay boom , blows out PIn wire transistor driver dead.

ill will study the page on your ECM more today, and noodle all possibilities.
if the OP hacked wires at point X , I then think gee where else did the mad man hack away, unseen in body panels off. etc.....

this Ecu has no FP relay failure tests like jeeps have,
nor has Fuel pump fuel pressure sensor to fail that via DTC or compensate for weak pumps or clogged filters as new cars have, fuel pressure is the mechanics job to cure. but we cant get there yet, FP Relay is dead. and is pink wire dead.
rpm 6500, in neutral is not good on any engine, parked hot or cold. use a real RPM gauge never trust any 20 year old suzuki cluster, best is read the scan tool , believe that.
driving down hill possible.
the OBD if in limphome mode , does on some sidekicks drop RPM limits.
(limphome means , spark advances is now disabled, 1 sensor is dead usually and not mimicked(crudely) closed loop dead, and rpm limits defeated, on some.
there is also , BLACK death mode called BACKUP by suzuki. (BLACK as in black fuel cloud every inch you drive)
BACKUP is when the CEL is dead, the DTC errors are all Dead, and ECU is dead for scans, that is right there is really 2 processors inside, 2 brains, not counting LIMP)_
1 is the full blown ECU processor, scan tool works.
2nd is a backup that is like a crudely simulated carburetor. (runs off throttle angle if TPS not dead or runs off vacuum and RPM only) ( I could make a ECu from scratch doing this way)

BACKUP mode the scan tool fails to communicate or the pre 96 cars, the CEL is DEAD as a door nail, and will never flash codes again.
Im sure yours is not in backup
but may be in Limphome. if any 1 sensor is dead.
Sorry for the long delay.
Delivering fuel and supplies to a friend in the mountains that's without a car or phone.
Still getting 12v on fuel pump relay(pink wire) when ECM if supposed to send to ground.

I understand if the CAS sensor at the crank was bad, I should at the very least get a code for misfire?
engine runs great otherwise.

I bought and used a elm-327 to scan for codes.
The only stored codes were for the MAP sensor.
Unfortunately I did not check freeze frame data. (Frustrated at myself)

Cleared codes. Drove 15mi on freeway.
13.5 miles off road, very rough secluded snow(delivered fuel and food to a friend in a cabin with no car)
15 mi freeway cruise home(total 65mi round trip)

On the way home after about 7 miles on the freeway at steady cruise(60mph) very flat slight downhill,
CEL light triggered a code P1408.

Engine ran excellent for the trip.
Warms up properly, RPM drops to steady 800
Enters closed loop,
O2 sensor voltage swings properly.
I AT and ECT temp sensors read as they should.
MAP looks proper (need to have wife drive while I watch MAF load)
Voltage good.
TPS looks good throughout range

Cleared codes no codes returned.
Runs good, tach works idles well true mileage good.

I took screenshots of the ELM software (torque pro sensor data)
However my phone met a untimely demise(mountains, snow, fire). I'll try to recover the screenshots and post them here.

Thanks for your time and experience.
I use the tracker to travel to stone very remote places. Your site and documents I've downloaded have saved me very long hikes more than once.
Forgot to ask
Do you have any recommendations for android software for reading ECM data from an elm device?

Forum Jump:

Users browsing this thread: 1 Guest(s)