Ignition failures

[spyderdrifter]

I was having issues with my ignition coils misfiring on cyl 1 & 4 a few months ago. So I swapped the coils around to see if that coil would fail on cyl 2 &3. Sure enough it did so I assumed that coil was bad. Since then, it has worked randomly. When it doesn;t work, the car will start and run rough for a few seconds and shut itself off. Each time I have checked for spark, and cyl 2 & 3 have no spark. A friend shipped me a spare set of coils which I installed on Monday, and it worked, and then I had the same issue come up, but only cyl 2 had no spark. By this time, I've concluded that the coils are not the problem. I decided that I need to find out the ressitence and Ohms on the coils and the transistor pack, but I can find any info for specs or how to do the test. I have a multimeter to use. I looked on ShopKey5 and did find some info (see pics) for the coils, but not the transistor pack. Can anyone point me to the right info? Also, I did the test for the primary coil resistence and got 1.1 ohms for each according to the connector. Am I right to think that's too high according to the chart? Also I couldn't really understand how to do the secondary and tested like I have the leads positioned in the last pic (the pic is of my old coils, just showing how I did the new coils). Is that right?

 

[luv2rallye]

Power transistor test (PT): http://www.dsmtuners.com/forums/articles-electrical-wiring/229419-how-test-power-transistors.html

Coil test: http://www.dsmtuners.com/forums/newbie-forum/268167-ignition-problem.html#post151257229

1.1 ohm is a little high but that would just cause a weaker spark, not stop the spark altogether. Heating coil and PT with a hair dryer helps bring out intermittent failures. [Heating coil also raises resistance slightly]. But the better coil test is to clamp plugs against metal (with cables attached), and then crank and watch spark intensity while heating coil and PT. Also make sure the PT is in it's bracket and bolted against metal (head) as that conducts more of the current than the negative wire does - it also disappates heat out of the PT.

Your coil secondary test pic has them correct for 1 cyl pair (& the right 2 for the other cyl pair) but make sure the probes touch the metal inside there - they usually can't reach. Clip probes to screwdrivers and hold them in place in there to touch the bottom metal part.

Another possibility is an intermittent CAS or CPS as they tell the ECU when to fire the plugs, but rule out the coil and PT first.
http://www.dsmtuners.com/forums/new...art-no-fuel-prime-no-spark.html#post151265626

 

[spyderdrifter]

Thanks. I did accidentally find the test for the transistor in my manual when I was looking for something else, but have never seen an analog multimeter. So I'm glad you gave me a link for another test. Will try that tomorrow after some sleep, been reworking things on my car all day today, so I've called it a night.

 

[spyderdrifter]

So I got my car running good for a week, then the ignition crap happened again. Ended up getting a new power transistor and replaced the battery cable linking the battery to the firewall. It started up and ran til today. Having the exact same problem. Transistor is new and the coils, plugs, and plug wires are all good. I don't know what else to test. Anyone have any suggestions?

I'm so tired of getting no where with this.

 

[luv2rallye]

As I stated in post 2 another possibility is an intermittent CAS or CPS as they tell the ECU when to fire the plugs. CAS selects which cylinder pair where CPS tells when. The CAS would be the likely culpret since there are some cylinder pairs that do fire. An intermittent CAS will be near impossible to prove (a hairdryer sometimes brings it out) without just swapping with a known good one.

 

[spyderdrifter]

I did do a test on the CAS, bit it was the one that involves removing it and turning the rotor to check for injector firing.

I just read through those posts again and they're for a standard 2G CAS. I have a '94 hall-effect CAS for my 6 bolt swap. Do you have info for testing it besides the test I mentioned?

 

[luv2rallye]

So you have the 4 wire black top CAS then?
If it's intermittent there's really no way to conclusively test it other than swapping it out. For a non-intermittent functioning/failing one you can measure voltage on it's wires (while connected):
The black top style CAS car harness uses White (1 - TDC), Black (2 - Crank Angle), Red (3 - 12v), Black (4 - Ground).
Both signal lines are pulled up by the ECU to 5v and the CAS pulls them low as it turns, 4 pulses per rev on the Crank Sensor signals and 2 per rev on the TDC line.

 

[spyderdrifter]

That is the correct one I have. But I have a yellow wire on the harness, not 2 black ones. I'll look around for another CAS. I do have a '90 CAS that has the wires dirrectly attached. That one I don't know for sure if it works properly or not. When I first got it, the wires were terrible so I took it apart and replaced the wires and it's never been used sonce. Would it be fesable to rig it to the wire harness to see if the car will start with that one? If so, it'd be enough proof to condem the '94 CAS. Then I can look for a new one.

 

[luv2rallye]

Here's a thread about another guy who has a yellow wire and about the 90 CAS: http://www.dsmtuners.com/forums/newbie-forum/271669-injectors-just-wont-fire-all.html#post151290464.

 

[spyderdrifter]

Maybe it's because its so late and I'm tired from dealing with this and school, but I'm not following the significance of the yellow vs black wire. I have my CAS wired exactly like the attached diagram from Magnus. I still have the original '97 ECU and didn't swap injector or plug wires (as writen in the procedure for my model). I know I'm getting fuel to the rail, as wel as through each injector. I checked that last weekend. The only thing not working is the cyl 2/3 coil. I'll test out the CAS according to post #12 from your link after I get some sleep, and then report back. Also, even though I'm not getting spark from that coil, I did test the resistance (primary and secondary) and it tested good. I even swapped with another coil set that likewise tested good, and still nothing.

Since the white to blue wire connection was the only one I had to connect myself, I'll recheck it to see if my solder job held up or not. It should have, but anything is possible right now.

 

[luv2rallye]

Senerio for bad CAS: If the CAS TDC line only gives 1 pulse per rev (instead of 2), it will just fire the same 2 cyl pair over and over (and never fire the other pair).

 

[spyderdrifter]

How can I find out if it gives 1 or 2 pulses per rev?

 

[luv2rallye]

Remove it, leave connected, connect voltmeter to the TDC line, key on, slowly rotate it's shaft by hand.

 

[spyderdrifter]

I get all of that except this

connect voltmeter to the TDC line

how do I do that?

 

[luv2rallye]

Either stick probe through white wire end of connector far enough to reach wire terminal, or pierce wire insulation with pin to reach wire, or connect extra wires between unplugged connectors so you can get at them. It's not easy fumbling with it trying to hold everything. It's easier to just swap with a known good one. Also using a needle type voltmeter is easier to detect the pulses than a digital one.
And it's only going to tell you if it's failing solidly - not intermittently.

Another way is to not remove the CAS and just measure the TDC voltage (white wire) while running or cranking. On a 2g with engine idling should be 0.5-2.0V [cranking engine - should be 0.4-3.0V] - a 1g is probably similiar. If you have 0V (or constant 5V) your CAS is dead since you should read something (even though they are pulses the voltmeter will read the average which will be more than 0). Of course this method really doesn't tell you if you have 1 or 2 pulses/rev easily (unless you already know what 2 pulses/rev average voltage is). But it would tell you if it's dead.

Keep in mind missing spark can be bad ECU, CAS, CPS, PT, coil, plugs, plug cables, or grounds on any of these.

 

[spyderdrifter]

Well I went to the salvage yard and found one CAS. It not in the greatest shape. The loops for the adjustment bolts are broken off, but I took it anyway just to see if there was any change. My car started up with it so I've turned it to where I can still have the lip of the bolts touching the CAS casing enough to hold it in place and still idle properly. This is probably the best I can do without having to waste the $500 on a new one, until I can find a better one in the salvage yards. This one came out of a '91 Laser and I took the cap off to verify it was an Optical CAS, and it is. Is it possible to swap the guts out with my '94 hall-effect CAS? Are the cases the same? I know they look the same outside, but more curious about the inside setup.

 

[luv2rallye]

I doubt they're swapable insides but you can pm Steve to be sure. Can you hold it in place with a large washer and lock washer? Or you might try the classifieds on here or eBay.

 

[spyderdrifter]

I don't know who Steve is. I can use a large washer (and lock washer) as a temp fix, and I'll run through the classifieds. I won't touch eBay for anything, ever.

 

[Diamond Star Werks]

My friend has a black top cas. cheap. Pm if interested

 

[spyderdrifter]

So for anyone interested in this, I did some more research on the CAS testing and dug through ShopKey5, since I use it for school, and found this test. Here's a screenshot (entire test)

 

[spyderdrifter]

If the previous screen shot isn't readable (it sucked after I uploaded it), then here's the procedures followed by the ajoining Fig listed in it:

1. Using engine analyzer with oscilloscope capability, connect special patterns probe to connector terminal No. 4 (1.8L) or terminal No. 1 (2.0L).

2. Start engine. Compare oscilloscope wave pattern with known-good wave pattern. See Fig. 28. Verify that the wave length (time) decreases as engine RPM increases. If a wave pattern is output and it fluctuates to left or right, check for loose timing belt or an abnormality in sensor pick-up disc. If a rectangular wave pattern is output even when engine is not started, substitute known-good CMP sensor. Repeat test. If a wave pattern is still abnormal, go to next step.

Fig. 28: Identifying Known-Good CMP Sensor Wave Pattern
Courtesy of MITSUBISHI MOTOR SALES OF AMERICA

3. Disconnect CKP/CMP sensor connector. Turn ignition switch to ON position. Using DVOM, check voltage between chassis ground and CKP/CMP sensor connector terminal No. 2 (1.8L) or terminal No. 3 (2.0L). if battery voltage does not exist, repair ignition circuit between CKP/CMP sensor connector and IG switch. If battery voltage exists, go to next step.

4. With CKP/CMP sensor connector disconnected, check for continuity between chassis ground and CKP/CMP sensor connector terminal No. 1 (1.8L) or terminal No. 4 (2.0L). if continuity does not exist, repair wiring harness as necessary. If continuity exists, go to next step.

5. With ignition switch in ON position, check for voltage between chassis ground and CKP/CMP sensor connector terminal No. 4. If 4.8-5.2 volts do not exist, replace ECM. If voltage is as specified, condition required to set trouble code is not present at this time. Test is complete. Intermittent problem may exist. See TESTS W/O CODES article in the ENGINE PERFORMANCE Section.

 

[luv2rallye]

I doubt anyone on here has an oscilloscope which is why I didn't list that test.

 

[spyderdrifter]

Yeah, I just posted it for reverence. I'm gonna bring the old one to school next week so I can learn how to test it since we have the equipment. I appreciate your help. Been beneficial again