Trigger Wheel Problem

[IanEpperson]

I'm converting my T4 2270 engine to FI using Mario's MaxSFI ECU, sexy cam sync sensor and hidden trigger wheel (among other toys). The engine is installed and everything is finally buttoned up and am finally to testing these sensor while cranking, but I'm getting bad data.

I turn the key to engine on, make sure I have comms in TunerStudio, click "start" on the "Composite Logger" page, crank the starter for a few seconds, then click "stop" on the "Composite Logger" page, and see this:

2017-03-28 06.18.39 pm.png

I tried a few more times and got nothing - no graph at all, then realized that the battery had low voltage and maybe it was cranking too slowly. I charged up the battery and tried again with the original result.

I disconnected the ECU and tested the Cam+ and Crank+ pins. Pulling that plug also pulls the grounds to those sensors. When I rotated the engine I saw 12v - 2v - 12v - 2v on Crank+ as the teeth moved past the sensor, and 2v - - - 12v - - - 2v as the engine rotated to every other TDC (12v when around the firing position for #1) on Cam+. I know it's supposed to drop to 0v, but I think it's because the ground wasn't connected for this test. I'll try again with it grounded, but I suspect it will be correct.

After talking with Mario for a bit, I guessed that my Cam+ and Crank+ pins might be swapped, so I triple-checked them to his diagrams (pinouts and connector)and they're correct. He suggested that I retry my tests with "Render Including non-interrupt Data" checked - I did and it didn't seem to change the result. Here's the log.

So, now what? Maybe the Cam+ and Crank+ pins actually swapped and the docs are wrong? Maybe I've got polarity reversed on the sensors and that's throwing something off? Maybe the Cam- and Crank- pins on the ECU have to be grounded? Something else?

Here's my wiring diagram, if that helps.

[IanEpperson]

Mario helped me out a bunch! There was a missing pull-up resistor and I now get a crank signal... now and then.

2017-03-30 06.26.59 pm.png

Now it looks like I'm not getting a decent cam signal. I've carefully checked it (this time with a proper jumpered ground) and the cam signal is 0v most of the time, then jumps to 12v before TDC during one of the revolutions (I assume #1) then goes back to 0v about 20 degrees after TDC.

I'm going to go play with the "second trigger wheel active on" rising/falling setting to see if that changes anything.

[nogoodwithusernames]

Try getting it going without the cam sensor at first and then once you know most of the stuff works add the cam sync? Go in a couple steps so you're not trying to troubleshoot everything all at once.

[IanEpperson]

Go in a couple steps so you're not trying to troubleshoot everything all at once.

Very good advice. I keep wanting to jump to "it's all done!" and hoping everything is OK.

Last night I adjusted the crank sensor distance to the wheel (which turned out to be a bit of a pain) and went from 3.0 mm to 1.25mm, charged up the battery fully and got a fantastic graph:

2017-03-30 09.09.39 pm.png

I'll pop out and get some new gas and try starting it up today. Woot!!

[Clonebug]

Go in a couple steps so you're not trying to troubleshoot everything all at once.

Very good advice. I keep wanting to jump to "it's all done!" and hoping everything is OK.

Last night I adjusted the crank sensor distance to the wheel (which turned out to be a bit of a pain) and went from 3.0 mm to 1.25mm, charged up the battery fully and got a fantastic graph:
2017-03-30 09.09.39 pm.png

I'll pop out and get some new gas and try starting it up today. Woot!!

Sensor gap is important.........

[Piledriver]

The specs on those Honeywell sensors seems to contain a great deal of wishful thinking.
3mm... More like 1mm.

3mm might work if the wheel is has flying magnets.

[IanEpperson]

So, still not getting it right. I put in a rebuilt auto-stick starter to spin the engine faster, got the 4.7k ohm to 5v pullup resistor as Mario indicated for the crank sensor, checked all my grounds again and got the hall sensor at about 0.7mm away from the (steel) trigger wheel. To rule out any EMI problems, I pulled the connection for my fuel pump and pulled the fuses for the LS2 coils and the injectors.

The best I ever get is two complete cycles (cam sensor rise and fall twice) of good data then I lose sync and it never recovers. It seems if I don't have the battery charged to 14.4 volts (fresh from the charger) it only gets a single cycle - but that might be random.

Promising:

2017-04-07 05.00.47 pm.png

Fail:

2017-04-07 05.01.02 pm.png

The reported error code is: error 2 - missing tooth at wrong time. Which makes sense as it's just missing data.

I tried disabling noise filtering - specifically the "Noise filter enabled" and "Tach period rejection" but the first showed no difference and the second made the sync problems radically worse.

Any suggestions to check or try? I'm at a loss.

[IanEpperson]

Oh, and here's a link to my CurrentTune.msq.

[Piledriver]

Disconnect the cam sensor temporarily, and set it for crank only batch fire.

Pull the plugs, the battery will last a lot longer, and it will spin faster.

Try moving the crank sensor out to where it worked before.
(Might want to do that first)

Also try flipping the trigger edge, may require resetting timing.

[IanEpperson]

Problem solved. The pull-up resistor had too high of a resistance and only actually working when the system was at about 15volts. Double checked that using a 1k ohm instead of a 4.7k ohm pull-up would be OK (that 1amp power supply should not care about going from 0.001 amp draw to 0.005 amp, and 25 milliwatts is easy to dissipate). Now it's triggering every time and have no sync loss!

I got it to sputter to a start with a crappy idle, then learned that A-B-C-D is firing order, not cylinder order... d'oh! More learning.