Diagnostic Codes 14 and 43 - Intermittent failure to start, cranks ok

[suprarx7nut]

Update: I have trouble starting again. It's been trouble free for over a year with consistent driving. This time, no codes are present at all.

I tried to check spark by grounding the plugs one by one outside the cylinder and with my wife's assistance as the visual eye while I cranked. She could easily see spark on 4 5 6, but could see only a very dim or no spark on 1, 2 3. I'm checking again tonight with an actual spark tester.

Idle has been a little low (550-600 or so) and "unpure" for a few months, but it still ran and drove just great. No stalls, no drop in power. Timing is correct at 10-12 degrees BTDC.

Compression check was excellent at 145-170 on all cylinders. The higher cylinders had some oil drop into the chamber so all dry I think the number would be even tighter.

Ideas are welcome.

 

[suprarx7nut]

Used an actual spark tester tonight. 1, 2 and 3 are barely firing once or twice then go completely dim. 4, 5, and 6 are much better - perhaps even normal.

Coils show good resistance and no shorts. Coil wires are good. Spark plugs are good.

Igniter is failing the test in the TSRM, but it's a little tricky with my homemade 3V setup. I'll try again tonight to make sure I'm performing it correctly.

Are there any other alternative ways to test the igniter? Would a failure of all three outputs be strange or do they rely on a common component that could be the culprit?

Also, can anyone clear up with terminals are which on the igniter/ECU? The TEWD seems to have an error. It mislabeled them on the page below:

 

[Nick M]

It has been a while but the ignitors that I saw fail didn't go out like you are describing. And that doesn't make it so either way. The car would not restart hot. Not to be confused with the FPU system. Corollas don't have one. The car would restart after cooling. Spark came back. And there was no test in the manual. At the end of the flow chart was "check with known good unit" or something close to that.

 

[suprarx7nut]

Update: Ran through the ECU testing portion (most of it) in the TEWD. Everything I tested passed except I'm getting no voltage on IGT and 1.1V on IGF with the ignition sw on. This seems to be a failure per the TEWD specs.

From an old Jetjock post on MKIII tech it sounds like IGt voltage comes from the ECU. I've verified that the 12v +B going to the igniter and ignition coil is present and healthy. All other signals that require the Ig sw on seem to be normal so the ECU knows ig sw is on.

I'll open up the ECU tomorrow. Perhaps it's a failed component in there? Are there any other pre-qualifying signals required for the ECU to put out IGt or IGf?

The ECU is a JDM GTE and the metal casing is pretty beat up. I've never opened it up so I'm unsure on any internal damage. Since I was having trouble earlier with the igniter testing maybe the igniter is bad. That would only make sense though, if the IGt voltage came from within the igniter. My understanding at this moment is that IGt is produced in the ECU.

More to follow...

 

[jetjock]

Yep, IGt is the main timing signal that originates from the ECU. The others are coil pack selection and spark confirmation. Only confirmation (IGf) is fed back to the ECU. No IGt = no spark. That said a scope is needed to check it properly.

Since the ECU deals in low level logic signals the 12 vdc from the ignition switch is required to supply the necessary current to drive the packs. It also powers internal igniter control circuits. Ground is via the case and mounting bracket and your added wire. Make sure you have a solid connection there. Btw that wire really ought to be chained to the intake grounds after being run directly to battery negative.

Wiring from the USDM TEWD. It looks goofed up too:

http://www.cygnusx1.net/Supra/Library/TEWD/MK3/manual.aspx?S=Main&P=040

The TSRM igniter test is flawed. I forget the exact problem with the procedure but 3p should be able to shed some light on it.

Did you ever get the STA signal working?

 

[suprarx7nut]

Thanks for that confirmation JJ.

I have not tested STA since I had trouble last time. Mostly due to the corresponding trouble code never coming back. I went to test it out today and low and behold the car fired right up. Let it run for a while, shut it off, tried again and it started fine, again. I didn't watch for the STA signal.

While it ran for a few minutes I checked through some voltage testing at the ECU. Everything looks normal except IGt/IGf when the car is not running. When running, IGt and IGf are within spec. When shutting down the TSRM says IGt should go to 9-12V a few secs after shutdown. It doesn't for me. IGt goes straight to 0V. It also remains at 0V when ignition switch is "on" with the car not running.

This isn't making a lot of sense to me. The 5V and 9-12V being absent on IGt seems to be a fault of the ECU, correct? It makes me leery that a new igniter won't solve the issue. At the same time, a bad ECU seems incredibly rare (not to mention hard to test).

I opened up the ECU and didn't see anything odd. No burnt traces, no burned components, nothing. I could trace the resistance of IGt's pin back to a transister on the circuit within the ECU without problem. I couldn't get resistance beyond that, but I'm not sure if that's just how a transister works. (I'm a Mech E, not an EE so sue me, )

In other news, the gauge water temp sender just failed. While letting it idle for 10 minutes or so, the temp gauge spiked above red. That seemed odd, so I checked with the multimeter and the sender had fluctuating resistance between 18 and 35 ohms. It was indicating ~240F or higher, even after the car had sat for 20 minutes. I checked the ECU temp sender and it was indicating a more reasonable ~170F. Looks like a bad gauge sender. Odd stuff. I double checked it with the engine cold and it still shows a crazy high temp (low resistance).

The igniter ground currently runs to a ground point on the fender hump right behind the fuse box in the bay. It shares a bolt with a ground coming straight off the battery and another ground wire going into the harness for all the ECU grounds. That ground line gets spliced into the intake manifold ground so I'm assuming that's sufficient. When i redid the harness I wanted to make sure grounds were well covered. I have the OEM ground points plus a couple extra. Ground should be real solid.

 

[suprarx7nut]

Got an 89 JDM Auto ECU. Plugged it in, no other changes. Car runs beautifully. Idles high when cold, settles just below 1000 when slightly warm.

Plugged my old ECU (90 JDM Manual) back in and idle was low (around 500) and rough.

Something's going on inside that damn ECU it appears.

Anyone know if the Auto ECU behaves differently based on any particular input signal? If not then it looks like my JDM ECU has some sort of fault regarding at least its ability to control idle, perhaps my difficult start issue as well.

Strange stuff.

**Edit: Oh, and I checked the igniter voltage with IG sw on and after shutdown. Still no voltage. I'm wondering if that test is just plain wrong in the TSRM and TEWD. Maybe different for JDM ECUs?