3.0L Megasquirt w/ EDIS: Headers running at different temps?

[jangell]

Not having much luck today. I did a few things in preparation for ignition:

- Replaced that coolant hose that was leaking with the correct sized hose. Of course, now there’s a pinhole leak in the hose to the hot water valve on the other side of the T that I have to fix...

- Lengthened the bank 1 injector wires so that I could run the harness around the throttle body instead of in front of it. The original routing was limited by wire length, and got very close to one of the coils, and wasn’t in the best location with regards to engine heat. In theory this will fix the sensor noise (or at least I can’t see what else would be the problem).

I cranked the car over, but it won’t catch. It just keeps cranking, occasionally firing some cylinders, occasionally backfiring out the throttle intake. TunerStudio shows it sometimes gets out of the cranking range, but almost immediately drops back down. It’s cranking fast enough to try to run (300-600 RPM range). The oil pressure is good as well. Ironically m, the AFR is sitting in the 14s after these tests, too.

I used a test light between each coil and spark plug, and it lit as expected for all cylinders. I also used TunerStudio test mode to make the injectors click on both banks, so I know they’re all working and I didn’t mess up the wiring. I have a fuel pressure gauge hooked up, and it shows proper pressures.

I did forget to re-connect the MAP hose briefly, but plugging that back in had no effect.

I’m stumped. Any idea what I should check next? I can record a video or grab TunerStudio logs of that would help.

Thanks!

— Joe

[FABombjoy]

occasionally backfiring out the throttle intake

Are you 1000% positive the wires are on in the right order?
Timing light shows expected results?


On an unrelated note, I peeked at your wiring diagram again. You have your TPS plug diagrammed like this:

Gnd
Vref
Signal

Seems like these are usually physically layed out as

Gnd
Signal
Vref

Or flip flop gnd/vref if the signal to MS is inverted. Wasn't sure if this was the issue or not.

[opethmike]

Backfire out the intake means that spark plugs are firing when the intake valves are open; that is, 180 degrees of crank rotation opposite from when they should be firing.

[jangell]

I’ll go through the wire again, but I’ve labeled each so I know which one goes to which coil — I’m sure the coils are on the right plugs.

When I lengthened the injector wires, I cut and soldered one at a time so I wouldn’t mix them up (that, and they’re color coded). The passenger side bank all run on the same wires (the orange and the blue ones), so I only had to lengthen two.

I think I found something — while I was pushing on the harness while MS was powered, I could hear the passenger injectors click on, and releasing it clicked them off again. Still narrowing it down. Weirdly, using test mode to repeatedly click the injectors mostly works fine, but when I push on the harness the they stop and won’t start again until I test the other bank and then go back to the first bank.

<sigh> I’ll have to go over this part of the harness again tomorrow; out of time today, but I wouldn’t be surprised if half the injectors not firing is the problem. Oddly, I was once able to get the car to run on two cylinders (very, very badly), but that was one cylinder on each side of the engine (1 and 4 or 2 and 5; it was one of the EDIS banks), which maybe helped a bit.


I finally added a TPS gauge to TunerStudio and noticed that it seems to set at 42.6% and not move as I adjust the throttle. If I disconnect the plug, it changes, presumably since it’s a floating input now.

A multimeter on the connector pins shows 5v, which I think matches what I have wired. The middle pin is at about 1.5v. I think I just swapped the pins in the wiring diagram, but the actual wiring is correct. Also, I checked the colors of the wires in the connector pigtail vs the diagram — I definitely just messed up the diagram. The wire colors, however, are correct.

I pulled the TPS and confirmed that nothing is mechanically broken on it or the throttle, so that’s good.

I the tried using the TPS calibrator, and got only 10 steps from closed to open, so something is definitely wrong there. It changes each time I turn the sensor, too. I’ll see if I have another sensor at home.

Thanks for help!

— Joe

[jangell]

Backfire out the intake means that spark plugs are firing when the intake valves are open; that is, 180 degrees of crank rotation opposite from when they should be firing.

Ok, that seems unlikely then. I mean, it ran last weekend, save for those two cylinders that I’s swapped the coils on. I was getting afterfire from that.

This time I was noticing an occasional pop as it tired to crank. Since I was playing with the throttle, I was able to feel hot air blowing out of the crankcase vent thing (where you put the oil in on the 2.8L). I thought there was something coming up through the intake, but I may have been mistaken. The engine also seemed to want to stop sometimes when it happened. I was wondering if wasted spark could be causing fuel left in the cylinder to burn if a coil didn’t fire for some reason. I’ll check for sure tomorrow, if I fail to fix whatever wiring I screwed up when on the injectors.

If I have the engine in run, can I hand crank the engine with the coils out and watch them fire one by one, or is that far too low RPMs for it to even try firing? It would be nice to be able to see them visually fire in the right order.

Thanks!

— Joe

[jangell]

I think that the wiring problem was a loose ground. I spent a couple of hours today tracking that down, eventually noticing that the MS was rebooting. I redid the ground wires on the engine side of the bulkhead connector to ensure that they are right and secure. Unfortunately, I still can’t start the car.

I also confirmed backfire. With the throttle open a little, I saw an orange flash in the throttle body. So definitely backfiring. I’m still leaning towards wasted spark igniting unburned fuel, rather than the coils being on the wrong plugs. Even so, I retraced the wires one more time to make sure that they are actually connected to the correct coils (they are).

I’m back to being stumped:
- Both injector banks fire fine in test mode.
- All EDIS coil banks are firing properly.
- RPM signal is good, and EDIS is increasing the timing once the engine exceeds 400 RPM, but the engine never actually catches and runs.

My battery is running low, so I’m going to wrap it up for the day. I’ll post a TS log and a video when I get home. I can’t imagine what I changed that broke ignition so badly since last week.

— Joe

[jangell]

And here is the video: https://youtu.be/kvTrS85jO20

And the log (this is not from the attempt in the video): 2018-08-12_13.26.07.msl

And the current tune: CurrentTune.msq. Mostly I reset the VE table from my idle tests last week (when I had two coils swapped), but I'm including it for completeness, since I might have messed with something else and forgotten. I did try the tune from last weekend as well, but it didn't work either.

The TPS is not connected here, so you can ignore its reading. I just set the TPS range so that the floating racing wouldn't trigger the flood clear feature.

Thanks again!

-- Joe

[dn010]

Looks to me like its firing at the wrong time. Kind of like trying to start a car with the distributor installed on the wrong stroke. Did your crank sensor get knocked out of position?

You’ve got a lot of wiring there. If a certain sensor (like temp sensor for instance) works then take those sensor wires and move them out of the way. Slowly tape up the stuff that is complete and get those wires out of the way so you can focus on the problems. How/where did you ground the MS unit, injectors, etc. ?

[jangell]

Crank sensor seemed to be in the right place at a glance, but I'll check it again next weekend to be sure. Easy to do. I'd be surprised if I knocked it, but it would be really, really nice if that was all that was wrong. I'll hook up a timing gun too, as FABombjoy suggested.

The wiring was fairly nicely taped up until today, when I cut all the self-fusing tape off to inspect the individual wires before I realized that MS was resetting and that it was a grounds issue. I need to buy some more self-fusing tape and wrap everything up again, although I was going to wait until it actually started again.

The MS, LC2 and EDIS grounds are connected to a bolt on the bulkhead where the coil used to be (you can see it on the right side of the video). On this same bolt is the end of my ground bus, so it has a very solid ground. There is also a short jumper to the engine block there, mostly because my bus wasn't quite long enough to reach all the way to the block, and two other small ground wires (transmission WOT switch and an starter relay I set up to work around problems where the car wouldn't start with a weak battery).

Injectors are run directly to the MS, and aren't separately grounded (they have two pin connectors), which I believe is how it's supposed to be done...

My latest wiring diagram is here, if that helps: https://static1.squarespace.com/stat...EFI+Wiring.pdf

Thanks!

-- Joe

[opethmike]

You are correct about the grounding for the injectors. Glad you commented on how you have everything grounded; when I read that the MS was resetting, it reminded me of when I first did my MS install, and I had the MS grounded at one point, and the O2 sensor at another - this caused a ground loop and the MS would constantly reset.