Posts: 15
Threads: 4
Joined: Sep 2016
Reputation:
0
Does anyone know what the operating characteristics are for the build in IGBT in the 1996 16v tracker distributors? I'm trying to operate the IGBT using megasquirt to fire the internal coil but i do not know what the operating voltage/current requirements of the gate are.
Posts: 6,084
Threads: 12
Joined: Apr 2013
Reputation:
28
suzuki does not publish these details.
just this,
why not reverse engineer it>?'
the IGBT , (ignitor) in the early cars had A resistor from gate to source, 150ohm? (holds charging coil off by default)
this prevented burning the coil of if one opens the gate connection, (obvious)
use a dmm , meter, in ohms mode measure it.
http://www.fixkick.com/ECU/1996/EFI1.jpg
http://www.fixkick.com
Posts: 15
Threads: 4
Joined: Sep 2016
Reputation:
0
(02-28-2018, 10:18 PM)fixkick Wrote: suzuki does not publish these details.
just this,
why not reverse engineer it>?'
the IGBT , (ignitor) in the early cars had A resistor from gate to source, 150ohm? (holds charging coil off by default)
this prevented burning the coil of if one opens the gate connection, (obvious)
use a dmm , meter, in ohms mode measure it.
http://www.fixkick.com/ECU/1996/EFI1.jpg
I do not have a 1996 1.6 16v ecu to scope the igniter driver output (orange wire) unfortunately. I’m just not sure if the signal from the ecu is 5v or 12v to pulse the igniter. I will check the resistance between gate and source. Currently I have megasquirt set up to pulse a 5v signal to the igniter gate, but it’s not operating the gate. It may be the internal resistance you mentioned is too great for my signal to overcome.
Posts: 6,084
Threads: 12
Joined: Apr 2013
Reputation:
28
all that is in the distributor.
ID say its 5v signal with a series 400 resistor in the ECU stock (a transistor pull up in the ECU )
the ECU limits current to .005Amps, 5mA. so should you,
400 ohms in series to G pin to ECU. for 5v.
or 1800 ohms to 12v to the G pin,
but id measure the G pin first. gate pin on distr, to ground.
there are 3 issues here,
the induction coil spark.
the igniter device (mos FET it is, but is Gate termintated)
the cmp.
my guess is you will want to use the Megasquire built in, ignitor driver?, (which MS is it , model>? , many there are)
then that ends this question, as the coil wire. is brown-white not orange, and would go directly to the MS. ( stock ignitor bypassed)
The MS must programmed to have this pin at default 12vdc. or you will NOT burn the coil up if you ever get infinite dwell on that pin, best is to use say a 100 ohm dumm test resistor for the coil and measure the signal on this dummy coil first to be darn sure the coil is running a safe dwell time. (NOT n spec'd by suzuki btw)
dwell time is coil charge time, and can never land at full charge 100% dwell ever.
and the dwell expands with RPM on EFI. (due to short periods at 6000 rpm) a short period will not charge the coil enough., and spark dies. (goes super weak)
All EFI systems do that. DIS is even more strict here due to huge coil currents and HE , high energy ignition setup.
so dwell is first.
next is what to expand to at 6000 RPM
and next is inversion on the coil driver, (on or off) MS settings.
then test it with a scope and dummy coil there, to make sure it really works as you think.
the connect the spark coil and not burn it up.
keep in mind suzuki shells no 96 coils not, nor does, Autozoned.
ignitor stock:: issues.
here is the most old book by suzuki see him use a light bulb to test the gate of the ignitor, 650ohms inside so current is high.
8v
http://www.fixkick.com/IGN_timing/slides...ge_24.html
high means way more than a real Insulated gate MOS transistor.
so if you take your dizzy
measure dizzy pin 5 to 4 ground, the ohms you see there is the 650 ohm terminator (or other value) that is there.
if you see that the drive you send there must be current limited.
the 8v has this.
http://www.fixkick.com/IGN_timing/slides...ge_24.html
vast things are unknown on that distrib.
1: dwell spec. on coil static and dynamic.
2: ignitor gate threshold but mine fired from 2to 3.5vdc. nor the input impedance mine was 650 ohms. (its current fired, 0.005 amps)
3: the max gate voltage before damage to the stock ignitor.? (my be over 5v or 12 we dont know, and cant test this for 2 reasons , unplugging the distrib. to see unloaed gate voltage also kills dead CMP and spark ends, and I was never found of cutting my stock Gate pin to test this)
4: the coil spec, primary impedance tops my list.
you can do many thing here with MS.
1: USE THE STOCK IgNITOR OR
a" BUY YOUR OWN
b: Or USE THE ms BUILT In ignitor
2: use a generic spark coil ending that worry on stock coil spec,. mounted to firewall like 89-95s had. or that new coil and powerful new ignitor to drive hotter coils?
3: set the MS up for full DIS ignition. 4 spark coils and fired from CKP..?
the manual shows, 3v test battery , (a warning to me, 12v is no good) , see my slide show here. I horsed this up for coverage on all pins.
http://www.fixkick.com/IGN_timing/slides...age_8.html
http://www.fixkick.com
Posts: 6,084
Threads: 12
Joined: Apr 2013
Reputation:
28
03-01-2018, 03:45 AM
(This post was last modified: 03-01-2018, 03:48 AM by fixkick.)
dwell in degrees is not dwell time.
at high rpm the degrees is very long but time is short.
In the old days they ran points and the coil was designed to run huge long dwell times at idle (or worst failed start cranked and points landed closed, and not burn up (unless a 1968 VW , lol) a CHevy V8 (1965) had ballast restore to end key on coil burn ups. BUT BYPASSES CRANKED.
then at say 4000 RPm spark was still good.
then EFI happened.
and the dwell time is now dynamic. and the coils run why higher primary currents for 2 reason, it should and we can for more spark power.
and the coil can burn up easy, if you dont know the coils spec. modern. (Custom EFI)
http://www.fixkick.com
Posts: 15
Threads: 4
Joined: Sep 2016
Reputation:
0
(03-01-2018, 03:40 AM)fixkick Wrote: all that is in the distributor.
ID say its 5v signal with a series 400 resistor in the ECU stock (a transistor pull up in the ECU )
the ECU limits current to .005Amps, 5mA. so should you,
400 ohms in series to G pin to ECU. for 5v.
or 1800 ohms to 12v to the G pin,
but id measure the G pin first. gate pin on distr, to ground.
there are 3 issues here,
the induction coil spark.
the igniter device (mos FET it is, but is Gate termintated)
the cmp.
my guess is you will want to use the Megasquire built in, ignitor driver?, (which MS is it , model>? , many there are)
then that ends this question, as the coil wire. is brown-white not orange, and would go directly to the MS. ( stock ignitor bypassed)
The MS must programmed to have this pin at default 12vdc. or you will NOT burn the coil up if you ever get infinite dwell on that pin, best is to use say a 100 ohm dumm test resistor for the coil and measure the signal on this dummy coil first to be darn sure the coil is running a safe dwell time. (NOT n spec'd by suzuki btw)
dwell time is coil charge time, and can never land at full charge 100% dwell ever.
and the dwell expands with RPM on EFI. (due to short periods at 6000 rpm) a short period will not charge the coil enough., and spark dies. (goes super weak)
All EFI systems do that. DIS is even more strict here due to huge coil currents and HE , high energy ignition setup.
so dwell is first.
next is what to expand to at 6000 RPM
and next is inversion on the coil driver, (on or off) MS settings.
then test it with a scope and dummy coil there, to make sure it really works as you think.
the connect the spark coil and not burn it up.
keep in mind suzuki shells no 96 coils not, nor does, Autozoned.
ignitor stock:: issues.
here is the most old book by suzuki see him use a light bulb to test the gate of the ignitor, 650ohms inside so current is high.
8v
http://www.fixkick.com/IGN_timing/slides...ge_24.html
high means way more than a real Insulated gate MOS transistor.
so if you take your dizzy
measure dizzy pin 5 to 4 ground, the ohms you see there is the 650 ohm terminator (or other value) that is there.
if you see that the drive you send there must be current limited.
the 8v has this.
http://www.fixkick.com/IGN_timing/slides...ge_24.html
vast things are unknown on that distrib.
1: dwell spec. on coil static and dynamic.
2: ignitor gate threshold but mine fired from 2to 3.5vdc. nor the input impedance mine was 650 ohms. (its current fired, 0.005 amps)
3: the max gate voltage before damage to the stock ignitor.? (my be over 5v or 12 we dont know, and cant test this for 2 reasons , unplugging the distrib. to see unloaed gate voltage also kills dead CMP and spark ends, and I was never found of cutting my stock Gate pin to test this)
4: the coil spec, primary impedance tops my list.
you can do many thing here with MS.
1: USE THE STOCK IgNITOR OR
a" BUY YOUR OWN
b: Or USE THE ms BUILT In ignitor
2: use a generic spark coil ending that worry on stock coil spec,. mounted to firewall like 89-95s had. or that new coil and powerful new ignitor to drive hotter coils?
3: set the MS up for full DIS ignition. 4 spark coils and fired from CKP..?
the manual shows, 3v test battery , (a warning to me, 12v is no good) , see my slide show here. I horsed this up for coverage on all pins.
http://www.fixkick.com/IGN_timing/slides...age_8.html
I have megasquirt 2 with the version 3.0 board. I have the built in IGBT circuit but i'm using it to drive the stock tracker igniter (or at least trying to) According to the manual using the IGBT circuit in megasquirt helps prevent damage to the processor since it adds a layer of isolation instead of driving the coil directly. Ok, so i found the gate to ground to read 1,440 Ohms! i did the igniter test like you showed and the gate is working fine with a 3 volt battery. When i apply the 3 volts to the gate i have continuity between pin 6 and 4 of the distributor. I measured the high voltage coil and it reads about 15,300 ohms and the secondary reads 1.1 ohms. I'm not sure why then megasquirt is not firing the gate when i'm outputting 5 volts to the gate.
Posts: 6,084
Threads: 12
Joined: Apr 2013
Reputation:
28
ok, is the distributor grounded"? pin 4
pin 7 power pin 12v
then the ignitor can fire.
if it charges with 3 volt on the gate (battery pack 2 AA cells) then the ignitor works and the coil,
then you release the 3v and the coil fires with 15,000 volts use a spark plug there, not and open circuit or you invite 15,000 volts back to the MS.
the if that works try the MS,
working with no scope makes this very hard.
but if you can , tell the MS to fire the coil 1 time (spin the CMP or CKP , not sure what you use to fire the MS,but make one turn of the CMP or CKP) then you see the coil charge then release.
one can see that even with a DMM on pin5 red, and black meter to ground.
that is 100% correct the extra ignitor is on more buffer protecting the MS. for sure. but needs the MS inverted, (you have 2 inverting buffers now) so invert the MS end or burn the coil up you will.
that is why I use a dummy load for a coil to make sure all is perfect, then blow up the coil, LOL!!!!!
make sure the distributor body is well grounded, or bad things can happen with 15,000 volts with no place to go but to a very bad place.
10 years ago i scoped this pin , i forgot
see the hand drawn wave at the bottom of my photo.
its 12v pulse i this car (I had 2 , 97/97 cars) and a nice scope, so drew that long ago.
note the inductive back kick (EMF) the suppressor may help the coil RING, IDK. sorry.
see bottom center.
http://www.fixkick.com/ECU/1996/EFI1.jpg
http://www.fixkick.com
Posts: 15
Threads: 4
Joined: Sep 2016
Reputation:
0
(03-01-2018, 09:28 AM)fixkick Wrote: ok, is the distributor grounded"? pin 4
pin 7 power pin 12v
then the ignitor can fire.
if it charges with 3 volt on the gate (battery pack 2 AA cells) then the ignitor works and the coil,
then you release the 3v and the coil fires with 15,000 volts use a spark plug there, not and open circuit or you invite 15,000 volts back to the MS.
the if that works try the MS,
working with no scope makes this very hard.
but if you can , tell the MS to fire the coil 1 time (spin the CMP or CKP , not sure what you use to fire the MS,but make one turn of the CMP or CKP) then you see the coil charge then release.
one can see that even with a DMM on pin5 red, and black meter to ground.
that is 100% correct the extra ignitor is on more buffer protecting the MS. for sure. but needs the MS inverted, (you have 2 inverting buffers now) so invert the MS end or burn the coil up you will.
that is why I use a dummy load for a coil to make sure all is perfect, then blow up the coil, LOL!!!!!
make sure the distributor body is well grounded, or bad things can happen with 15,000 volts with no place to go but to a very bad place.
10 years ago i scoped this pin , i forgot
see the hand drawn wave at the bottom of my photo.
its 12v pulse i this car (I had 2 , 97/97 cars) and a nice scope, so drew that long ago.
note the inductive back kick (EMF) the suppressor may help the coil RING, IDK. sorry.
see bottom center.
http://www.fixkick.com/ECU/1996/EFI1.jpg
ok so update. You will never guess what i found wrong. so i had to build a custom fused power block to power some of the stock 16v components (since the samurai never had these parts), and i went through everything again and found that one of the crimps i made on the 12v wire that powers the coil was crimped on the wire insulation, and not actually contacting the wire itself, so the whole time i have been testing it there has not been 12vdc at the coil (black/white wire). I also checked with the extra IGBT in the circuit which one of the spark setting makes the orange wire read zero volts going into the distributor. I actually had to set spark to Normal (going low) for the 5v on the orange wire to not be on it constantly. I know if 5v is present all the time, that means the igniter is firing and burned coil or soon to be. so i tested it and i have spark now, i had the dwell time set to 3ms and simulated it at 600 RPM (megasquirt has a input/output test area that allows me to test all inputs/outputs without running the engine) and gives a nice spark. The spark was blue and not super bright so i may play around with the dwell a tiny bit, but 3ms seems to be ok for now. That was the last major piece to the puzzle. Now its time to program everything. You have no idea how useful your notes have been, i've been using your notes on your page to integrate megasquirt into the stock components for quite some time now. I don't think i could have done it nearly as quickly without all the info you had, so thanks!!!
Posts: 15
Threads: 4
Joined: Sep 2016
Reputation:
0
03-01-2018, 02:06 PM
(This post was last modified: 03-01-2018, 02:11 PM by Agk007.)
Here is a couple pictures of what i'm doing. I used the factory engine harnesses to the injectors and to the distributor and stripped it down and rebuilt it by removing what i don't need and then i wrapped it again. Makes hooking all the sensors up a breeze since it uses the factory connectors. Also i'm controlling the PWM idle air valve on this engine so i can control RPM during cold starts. Soon i should be able to get it running. I have the cam hall working with megasquirt, and the stock igniter/coil working. Also i tested the injector fire and they fire properly. Hopefully it all comes together. The challenging part is figuring out the spark map since this engine is relatively unknown in those regards.
https://imgur.com/a/CVQoL
Posts: 6,084
Threads: 12
Joined: Apr 2013
Reputation:
28
cool
get a spark map? from a 1995 toyota 1.6L corolla, 40million made!, vast. surely one of those is to be found.?
cold idle is not controlled until 150f temp is hit, the IAC therm device does that seen mounted below TB. (id bet they just keep the ISC closed or nearly until 150 hits)
The spark advance table is complex and 3d;
RPM and load.(and hot cold engine and even cranking rules) yes complex it is. unlike say the old 89 with vacuum+centrifugal system. (even in Acceleration mode, different)
One of the reasons the engine can get near 30mpg is that. (perfect advance under all conditions, it's really cylinder pressure that matters most, and rpm)
getting that pressure to max out at the sweet spot just after TDC happens (rod angle), (the leading of the spark sets that point) (and Rod to crank pin angle is the key there (a lever it is)
I starting map should be easy (not cranking that is at 0 advance of base) the starting of making a map. there are computer models that can do that easy.
it asks for displacement 1.6L
and VE (volumetric efficiency (head type 16v very high) (the CAT will lower this at high RPM if present )
AND MODE, flat cruise or hills and acceleration modes. (3 rules there)
many universities. have mechanical engineering dept for engines (MIT), and have this on their web sites. combustion engineering.
it's a lot of fun to run those modeling programs . and not hard to do, (users)
that is my spark of ideas to you...
http://www.fixkick.com
|