Suzuki Swift T3 conversion - engine no start

[t3tom]

im sure its just a photo setup error but why is the suppressor unplugged?.. this is never a good idea...
The photo of the noise suppressor unplugged was also to verify that on my identical but running T3 car, unplugged noise suppressor does NOT prevent the engine from starting and running normally. Disabling the noise suppressor for now will help reduce the number of possible causes of the no start condition on my friend's car.

it has to be present, or the coil primary, may not oscillate properly.
the problem with suz, schematics, some are not totally accurate
on most of these cars, there is capacitor (condenser) on the coil minus terminal to ground. (it resonates "ring like bell" electonically )

this causes the primary to ring, and make full high voltages as a resort, (as seen on old points cars, if missing this Condensor, spark will be there, but weak)
the problem with SUZ is they hide its location, (ignitor, suppressor, or just cap on the fire wall and then to coil minus. 0.1uF is typical)
it must be there, is the bottom line, some where.

In addition to the noise suppressor, the T3 cars have a separate condenser wrapped into the harness connected to the coil + terminal to ground.

in all tests the spark HV terminal fails the spark plug wire at the coil test? .028" gap or more narrow for turbo, not sure on turbo spec here.0
turbo cars run tighter gaps, or special more powerful spark setups.
do all test with the diag jumper in place, this jumper is harmless , and must be left there until engine runs good.
Yes, a spare diagnostic fuse was inserted into the diagnostic connector in the interior fuse box, and it was in place during all tests (same as jumping diagnostic pins in the Diagnostic plug located in engine compartment)

you have now substituted all part parts but ECU and wires., right?
yes, all parts from the non running car, including 2 ECUs, were tested on my running car, one at a time. Each tested part resulted in engine starting normally on my running car.. The non running car's wiring harness wires between coil, igniter, ECU were checked for continuity and voltage if applicable. Spark plug wires were not checked but they were fairly new

hope this helps, and the ignitor needs a 100% perfect ground, or spark dies.
Yes, clean igniter grounds were verified.

coil minus to ground, fire is dwell time, this is coil charge up time, the it fires and you can see it RING.


vert , 10v/div
horz, 5mS/div

thanks for this. I wish we had a scope instead of just a multi-meter to see what the signal wires are doing.

[fixkick]

but are you checking spark at the coil?

coil >>>> center >>>spark plug wire>>> spark plug , ground.


like that.
known good plug and wire, to ground,
if this fails, then the ignitor is not firing,
if the ECU to ignitor trigger is dead (orange wire), then ECU inputs must be checked.

ok your suppressor only runs the tach,
lots and lots of spark systems , huge variance by year.

[t3tom]

but are you checking spark at the coil?

coil >>>> center >>>spark plug wire>>> spark plug , ground.

We will need to recheck that. During earlier tests a confirmed spark was observed from coil secondary to spark plug, but we have neglected to reconfirm this in the most recent series of checks. Thanks for reminding us not to skip a key test.


like that.
known good plug and wire, to ground,
if this fails, then the ignitor is not firing,
if the ECU to ignitor trigger is dead (orange wire), then ECU inputs must be checked.

We did have voltage measured between the orange trigger wire and ground during cranking test. We will also need to remember to re-check for presence of a code 41 after each test.

ok your suppressor only runs the tach,
lots and lots of spark systems , huge variance by year.

As we are both novices to electronics but trying to learn.. can you comment on whether there is way to use a digital multimeter set to volts measurement to confirm a normal signal pattern from the igniter/ECU wires? For example, if the igniter trigger signal wire output is a square wave 5V, should a true-RMS multimeter be set to measure AC Volts or DC Volts in order to capture a meaningful voltage value? If the meter were set to measure DC Volts, is the voltage reading a simple average of the wave form, or does the duty cycle (dwell?) impact the voltage value? Thanks again.

[fixkick]

the coil will spark if the voltage goes to 0v, holds there, for the 5mS seen on screen above
then if it releases, to 12v, the coil will induct, induction mode.
if this signal is good the coil is bad
if this signal is dead, the Ignitor is dead.
if the signal to the ignitor is good the ignitor is dead or has no power or ground.
or if the ignitor trigger is dead the ECU is bad, but if the inputs cmp or ckp are dead the the ecu will not fire its ignitor. (some of ours run with only CMP working (all G16b's do that) but most cars with CKP it must work, or no spark.

if direct coil spark is ok, then the rotor is timed wrong, or has junk china rotor, that is not indexed correctly.
or the distrib is timed wrong or 180 out, or wired wrong order to plugs, or wired CW when it should be CCW, (not sure what in this car)

the meter is only going to show averages.
what matters are clean signals,
some signals that are slow, can be checked with a meter, the needle wags on a CMP hall sensor, for example. 0v, the 5v, then 0v.
but if car has CKP and is just a 2 wire coil, not at all, the meter is useless here.
hall easy, coils only sensors impossible, lacking a scope.

the coil sensors, non hall, are just coils of wire and mag, and very very tiny signals there.

maybe this goofy system does not report cmp/ckp errors.
unplug them 1 by 1 and crank and see.? em? DTC fail?
if they throw errors good, if not, the ecu needs to be kicked thru the goal post of life, LOL.

The ecu says, the ignitor signal is wrong. in some way. (dead or weak or intermittant)

[t3tom]

the coil will spark if the voltage goes to 0v, holds there, for the 5mS seen on screen above
then if it releases, to 12v, the coil will induct, induction mode.
if this signal is good the coil is bad
if this signal is dead, the Ignitor is dead.
if the signal to the ignitor is good the ignitor is dead or has no power or ground.
or if the ignitor trigger is dead the ECU is bad, but if the inputs cmp or ckp are dead the the ecu will not fire its ignitor. (some of ours run with only CMP working (all G16b's do that) but most cars with CKP it must work, or no spark.

if direct coil spark is ok, then the rotor is timed wrong, or has junk china rotor, that is not indexed correctly.
or the distrib is timed wrong or 180 out, or wired wrong order to plugs, or wired CW when it should be CCW, (not sure what in this car)

the meter is only going to show averages.
what matters are clean signals,
some signals that are slow, can be checked with a meter, the needle wags on a CMP hall sensor, for example. 0v, the 5v, then 0v.
but if car has CKP and is just a 2 wire coil, not at all, the meter is useless here.
hall easy, coils only sensors impossible, lacking a scope.

the coil sensors, non hall, are just coils of wire and mag, and very very tiny signals there.

maybe this goofy system does not report cmp/ckp errors.
unplug them 1 by 1 and crank and see.? em? DTC fail?
if they throw errors good, if not, the ecu needs to be kicked thru the goal post of life, LOL.

The ecu says, the ignitor signal is wrong. in some way. (dead or weak or intermittant)

We will recheck the coil and dizzy spark plug order. The distributor rotor is identical to a Suzuki OEM part.

But since the igniter, ECU, and coil taken from my friend's non running car and installed into my running car, resulted in a perfect engine start, does this prove the parts are good? Is it more likely that we are dealing with a wiring fault that was not anticipated by the FSM procedures?

For example, the FSM does not include any trouble shooting scenarios that cover incorrect wiring that could occur during an engine harness swap, the FSM only covers open or shorted out wires.

[fixkick]

the coil will spark if the voltage goes to 0v, holds there, for the 5mS seen on screen above
then if it releases, to 12v, the coil will induct, induction mode.
if this signal is good the coil is bad
if this signal is dead, the Ignitor is dead.
if the signal to the ignitor is good the ignitor is dead or has no power or ground.
or if the ignitor trigger is dead the ECU is bad, but if the inputs cmp or ckp are dead the the ecu will not fire its ignitor. (some of ours run with only CMP working (all G16b's do that) but most cars with CKP it must work, or no spark.

if direct coil spark is ok, then the rotor is timed wrong, or has junk china rotor, that is not indexed correctly.
or the distrib is timed wrong or 180 out, or wired wrong order to plugs, or wired CW when it should be CCW, (not sure what in this car)

the meter is only going to show averages.
what matters are clean signals,
some signals that are slow, can be checked with a meter, the needle wags on a CMP hall sensor, for example. 0v, the 5v, then 0v.
but if car has CKP and is just a 2 wire coil, not at all, the meter is useless here.
hall easy, coils only sensors impossible, lacking a scope.

the coil sensors, non hall, are just coils of wire and mag, and very very tiny signals there.

maybe this goofy system does not report cmp/ckp errors.
unplug them 1 by 1 and crank and see.? em? DTC fail?
if they throw errors good, if not, the ecu needs to be kicked thru the goal post of life, LOL.

The ecu says, the ignitor signal is wrong. in some way. (dead or weak or intermittant)

We will recheck the coil and dizzy spark plug order. The distributor rotor is identical to a Suzuki OEM part.

But since the igniter, ECU, and coil taken from my friend's non running car and installed into my running car, resulted in a perfect engine start, does this prove the parts are good? Is it more likely that we are dealing with a wiring fault that was not anticipated by the FSM procedures?

For example, the FSM does not include any trouble shooting scenarios that cover incorrect wiring that could occur during an engine harness swap, the FSM only covers open or shorted out wires.

there wires are all just as important as the sensor and actuators.
if you had a scan tool i bet it shows some dead, sensors.
some older cars with turbo even have this MAF, or VAF sensor that has flapper switch in it and if it sticks, no spark. no fuel.
we are working blind here, like there is no spark direct at the coil top, then The ECu can not fire the coil,
sure the part removed all work, but there are parts you missed, or timed wrong.
or other parts not discovered yet. for sure all input to the ecu matter. many are to get spark cranked.

the ECU fires the ignitor, this must be known first.
the ECU tells you it cant fire it., it dont know why? only that the spare read back pin reads wrong,, but it could be bad wiring from ECU to ignitor
if say you scanned at and saw ckp or cmp or rpm dead or intermittent, cranked, then you'd know those are the cause.

I can make spark work on the bench with the 8v ecu
cmp (simulator)
and coil , ignitor , and it makes spark, on a work bench.
but this 91 has no ckp at all, or other odd devices. to block spark.. that need to work ,that maybe a Turbo ecu wants working first.
that scope photos above is bench run, spark.

STIM word there means im using the Megasquirt STIM , (simulator ) it is a cmp /cpk generator, (oscillator )
this connected to the ECU CMP pin and it injects and sparks, instantly, after all car can gravity or push started, no starter motor in car.

so you have swapped ECU, that leaves wires or sensor you have yet to discover.
There are many turbo systems. some have extra fail safe sensors, and Im not sure on this car. never seen one.,

but here are some ideas, based on other cars
no oil pressure, (to turbo)
air flow vain VAF? stuck closed or failed in that mode, electrically like the early Mx5 Mazda turbos had... before electronic MAF. (same year too)_
starting in gear, or not in park, or park lock out failure? , if stick shift, clutch switch.
over heat sensor lying? wire fell off, any heat sensor, ect, iat, ac overheat, any heat sensor...
idle switch not closed?
that is all that is in my head this pm.
good luck to you.

[fixkick]

bad wiring
or sensors you've yet to discover.
its easy to damage wire ripping harnesses out of donor car.

[t3tom]

bad wiring
or sensors you've yet to discover.
its easy to damage wire ripping harnesses out of donor car.

Yep.. I can't imagine the kinds of problems techs in the future will face when troubleshooting 2016 model cars as they turn 25 years old, with miles of old wiring, and sensor networks.

Thanks for all the help so far.. Diagnosis will continue!

[fixkick]

bad wiring
or sensors you've yet to discover.
its easy to damage wire ripping harnesses out of donor car.

Yep.. I can't imagine the kinds of problems techs in the future will face when troubleshooting 2016 model cars as they turn 25 years old, with miles of old wiring, and sensor networks.

Thanks for all the help so far.. Diagnosis will continue!

I to think those same words every ,day.
what will be the future, of all this, when all those vast parts are not sold in the future,( ever notice how obsolescence, is accelerating?)
remember air pumps from the late 70, blowing air in to the exhaust ports, to burn the unburned fuel, (most are not sold now, spare pumps)
and cars with 6 air bags, and then you need one, to pass smog, they will fail you for any warning light.

id try to find a tech1 scan tool that can scan this ECU. (barrow one,?) needs correct plug in...
see all those bits. good bad or ugly

here is ours. our 32 bits..(words)

http://www.fixkick.com/ECU/Tech1/full_scan2.html


id bet there are surprise there.. for sure working on any car, no matter how old (or new) many problems can ONLY be found using the scanner.,
OMG, i never dreamed this part here failing would cause ignitor errors... (secret software it all is, inside)

I see the turbo version has a longer page on this.... and more DTCs. its very much unique and more complex.

cheers and good luck !

[t3tom]

bad wiring
or sensors you've yet to discover.
its easy to damage wire ripping harnesses out of donor car.

Yep.. I can't imagine the kinds of problems techs in the future will face when troubleshooting 2016 model cars as they turn 25 years old, with miles of old wiring, and sensor networks.

Thanks for all the help so far.. Diagnosis will continue!

I to think those same words every ,day.
what will be the future, of all this, when all those vast parts are not sold in the future,( ever notice how obsolescence, is accelerating?)
remember air pumps from the late 70, blowing air in to the exhaust ports, to burn the unburned fuel, (most are not sold now, spare pumps)
and cars with 6 air bags, and then you need one, to pass smog, they will fail you for any warning light.

id try to find a tech1 scan tool that can scan this ECU. (barrow one,?) needs correct plug in...
Hmm. it seems that a 1991 Turbo Sprint does not have ODB1 or even a 12 pin ALDL connector, only the CEL and a Diagnostic fuse socket. This must have been one of the reasons these cars were never sold in the USA, as Canadian ODB standards may have been less stringent. It looks like the wiring harness will need to be re-verified wire by wire. Later on, I may get an inexpensive handheld oscilloscope as it seems to be a small investment to improve our understanding of electrical faults for cars of this era.

see all those bits. good bad or ugly

here is ours. our 32 bits..(words)

http://www.fixkick.com/ECU/Tech1/full_scan2.html


id bet there are surprise there.. for sure working on any car, no matter how old (or new) many problems can ONLY be found using the scanner.,
OMG, i never dreamed this part here failing would cause ignitor errors... (secret software it all is, inside)

I see the turbo version has a longer page on this.... and more DTCs. its very much unique and more complex.

cheers and good luck !

Thanks for the encouragement!