Julian's '91 200 20vt

[Marc]

mega trigger errors in your running log (less p.trig). what that means is the ecu is seeing not enough primary triggers inbetween secondary triggers (chipped flywheel tooth?) the engine will not run correctly until that is sorted.

[Marc]

it will also throw the same error if the wiring is not correctly reversed.

[themagellan]

marc, could a bad g28 or g4 also return this information? Or the fact that it is reading errors shows something else? way to go above and beyond for him we are watching lol

[88a5tq]

I agree he's been an extremely big help on here and on the looks like the project might be on hold for a little bit tho. Need to check some mechanical things out. I'm a little concerned about the coil though because I'm getting spark during testing from every coil. I'm exploring the idea that the COP harness's ground strap could have#too much interference at the stock ground bolt on IM. Do you guys think the ground could be not allowing coil 5 to fire properlywhen running but work fine when in testing mode? Definitely want to thank everyone whose stick around in the thread!

[88a5tq]

I tried to take a break today from the car after work (yeah right lol) but went to remove the FW sensors bracket and stripped the one damn alan bolt. So a break turned into 40 minutes of dremel work cutting the head off that little turd. I didn't get a look yet at the ring gear but has anyone had luck throwing some welds on chipped teeth through the starter hole and filing them down to the correct profile? Otherwise, im looking at a replacement gear from 034 and the trans being dropped again. Anyway, more tomorrow.

When I look back, that violent hydrolock happened I think before I got the car to even idle.
Just found this snippet: "Generic check engine light this function will light up CEL before cranking, if rpm goes above cranking threshold CEL will turn of. Additionally this function informs the user of FLOOD_CLEAR condition (tps above 70% during cranking shuts of injectors to allow to clear flooded engine) and MAP signal unexpectedly low value (below 80kpa) by flashing warning light fast (5Hz). " concerns the CEL functions of VEMS for V3. It seems the car was only starting because the flood clear thing was reducing fuel or something since I had to floor it to start. So anyone think one of my rods is bent from all this?

[88a5tq]

I'll let my stellar compression numbers speak for themselves:

#5 (suspect cylinder)

IMG_20140809_135559.jpg (1.17 MiB) Viewed 47581 times

#4

IMG_20140809_135822.jpg (1.08 MiB) Viewed 47581 times

#3

IMG_20140809_140218.jpg (1.02 MiB) Viewed 47581 times

#2

IMG_20140809_140336.jpg (1.08 MiB) Viewed 47581 times

#1

IMG_20140809_140512.jpg (1015.22 KiB) Viewed 47581 times

And now for the meat and potatoes of the matter. I decided to check the fueling side of things first. Here is my outputs page with testing mode enabled and the fuel pump checked and turned ON.

[88a5tq]

IMG_20140809_163802.jpg (2.48 MiB) Viewed 47579 times

now id like everyone to look at the line to the right of the fuel pump and look at what it says. "ignition output 5 (7A) grounding, injgroup" Does everybody have the same thing in that box? Anyway when I hit ON for the fuel pump output, I noticed that injector 5 was receiving the ON signal before I even step mouse on the injector outputs visual page. I'll repeat that, I don't need to test injector 5 in order for it to stay on... In fact when I click TEST on #5, nothing happens at all. I did notice that when I clicked one injector's test button, another one went off instead. This last one is a shot of my injector output visual page
With nothing turned on or changed and still have injector 5 stick open. Is this normal

IMG_20140809_153502.jpg (1.99 MiB) Viewed 47579 times

[Marc]

First thing first, the channel numbers are zero based. So, the cylinders get channels 0-4, fuel pump is 5. Are you saying you get no pulsing from cylinder 5 at all, it just stays stuck on? Test with a test light with engine running and #5 injector clip unplugged. If it stays lit constant, that sounds like a blown injector driver.

[88a5tq]

IMG_20140809_203053.jpg (449.38 KiB) Viewed 47564 times

I have a problem and the answer is not more cowbell... My flywheel needs to go to the dentist. That's $700 or more in damage in a split second and I didn't even hit a parked car

[chaloux]

Lol at your last line but no laughing at your situation. That sucks man. I know how it feels to have some wacky-ass issues.

I feel like I should help, I'm just so busy right now I can't read through everything.

[pilihp2]

Shit.

[88a5tq]

Thanks guys, Phil lol. Uh ya the culprit is the fuel pump signal also happens to turn injector #5 to stay full open... who woulda think right. It was never going to run. Proof of flywheel being fine b4 this is in the trigger log

[Marc]

my money is on config corruption. The 3b ecus I cant test before I ship because I dont have any running 200's at my shop so we leave the default .hu firmware loaded on them and just adjust them for things like coil on plug and injector configs. I suspect what has happened here is because its 1.2.27 firmware the latest stable firmware isn't new enough to properly run things like test mode (you can see where his fuel pump channel shows as "none" on the test mode screen).

I've advised him to download the latest nightly vemstune, reset channel 5 as the fuel pump channel and repeat the test mode procedure to make sure #5 doesn't light up when turning on the fuel pump channel only in test mode.

[88a5tq]

I updated to the one month newer nightly release from this. It fixed the "none" label

IMG_20140810_115633.jpg (1.64 MiB) Viewed 47542 times

but as you can see I didn't need to change those settings but burned them anyway then went to this screen to see "none" was now fuel pump.

IMG_20140810_130250.jpg (848.78 KiB) Viewed 47542 times

Here's the low down:
Ok so I updated to the nightly build. Now it says fuel pump channel for injgroup 5. It doesn't stop there though. When I click ON for injgroup 0, the fuel pump freakin runs like it used to when I tested the fuel pump. I was literally about to start the car again when I found that out. So now the fuel pump channel test just clicks and turns injector #5 on full but only blips the fuel pump for a fraction of a second. both injgroup 0 and 5 turn the 5th injector on still except testing the fuel pump test only blips injector 5. Its all reversed now. I tried my older fuel pump relay again (been using new one for about 3 months) but the results stay the same. Testing the fuel pump from outputs menu still turns both fuel pump and injector 5 full on.

[88a5tq]

Redacted

[Marc]

I think at this stage its best for you to ship me the box so I can do some bench diagnostics. I have a new theory on what is going on (there are "clamp plates" that hold down the FETs for injectors and coils. I'm thinking these clamp plates might be touching things they should not in the case. this will cause one channel to also fire an adjacent channel). anyways, ship it back to me, I'll credit you any shipping costs and turn it around as fast as I can.

EFI Express
61 Thompson Mill Road
Lee, NH 03861

[pilihp2]

Had a busy weekend. Wish I could have helped more. Will brainstorm as I sit and do nothing at work for the first couple hours tomorrow.

I won't be of much help. But dammit it's the thought, right!?

[88a5tq]

Hah, Phil are they hiring lol. ECU is going back. All tests with stock motronic are normal/correct.

[88a5tq]

Wanted to make it clear that Marc truly knows how to take care of his customers! The other guys though, not at all so far. He was even correct on his remote diagnosis before he even saw it. Two FET's were touching and ruining stuff while operating. VEMS won't answer my calls (3 days) at this point but time will tell. Hopefully they realize their customers come first. Thanks again Marc.

[88a5tq]

Im curious, but pretty new to it also, is there any way of using the missing tooth to still maintain a correct primary trigger? Saw something called "missing_tooth" but it looks like the missing tooth has to be in a specific spot on the sensor's ring gear so its useless. My only other option would be to get a tone gear on the crank pulley I guess but that may prove just as pricey as fixing the flywheel. New wiring and a new config would be irritating also. Anyone have any suggestions for the p. trigger or basically just fix the FW's ring gear? TIA

[carl]

Fixing the wheel is doable, but through the starter hole might be a huge pita, maybe a modified file on a sawzall or jigsaw at slow speed to shape in between. As long as you have a cut piece of cardboard to get the outline and height just right it should work. I've done it on an mc engine setup, wheel was on a table though, mac10/11 ecu was ok with the new welded/grinded teeth.

[88a5tq]

Thanks Carl, I don't have a suitable welder unfortunately and the nearest shops won't weld it. I wish somebody would say hey I did it all through the starter hole lol. I can't goof with it I don't think. good thing the ring gear is removable!

[88a5tq]

im quite happy to have a working ECU back in my hands. Now the white smoke has drastically reduced to what appears to be normal for cold start exhaust and beyond. I've got a new issue now but it might be minor. When my wbo2 is warmed up and taking readings it makes the idle jump around as its lambda reading goes rich to lean about twice a second. I updated the firmware and reloaded the config today but forgot to check if the wideband needed recalibrated. I will check that. I didn't have a secondary trigger reading until I updated it so that was a relief. As usual, though, when I add minimal throttle the car dies promptly. I charged the battery again today, as well. I think the stalling is to do with the missing flywheel tooth maybe?? When I calibrated the wbo2 originally it seemed to read fine so I think its wiring is good. However when I pulled it yesterday there was alot of dark gook and residue on the top of the sensor that I wiped off. Should I clean it in fuel to fix the reading?

[88a5tq]

Making slow progress here. Turns out my courtesy O2 sensor was ruined by the poor running condition when I first received everything and the FET's were screwed up. I soaked it in fuel overnight and blew it out with compressed air this morning and let it sit for after work. Open-air calibration was pointless as I maxed out the calibration number to 255 and still couldn't break 18 on the gauge. Ran to AutoZone with the trusty CIS 5k beaster and picked up a new sensor which calibrated fine around 235 so that was successful. Took and running log and checked a few items including fully torqueing down the O2 sensor. I then unplugged the IAC to find a rock solid stock-like idle. SOOOOOooooo, anybody got the magic IAC settings?? Currently mine are the base-map settings. Phil, I perused your 3B config settings and found many differences in the IAC areas but that was before I knew to check those settings for the issue. If they served you well I may copy them. Thanks dude! My settings are below. I was finally able to get a 1.0 lambda by adjusting my Req_fuel setting down to 6.8 I believe it was. Any help is greatly appreciated. With the IAC unplugged I am also able to rev the engine slightly higher than not at all before it stalls. That's some sort of improvement.

[88a5tq]

VEMSlog