Request for a wheel decoder feature + ...

[ekam99]

Greetings to everyone,

Having 3rd generation 3sgte engine and using standard 24-tooth +1 tooth in the distributor for my ms3-pro setup, I quite often find my engine not able to sync. It runs sequential with COPs. This happens most of the times when the engine is warm. I either get backfire or just cranking with no sync. Off-on cycle and engine starts right on. I blame compression in the cylinders and lack of teeth in the wheel for this.
OEM ecu starts the engine flawlessly every time because OEM's cranking advance is defined by the tooth. OEM 10 degrees advance is the position of 1st tooth. The pulse goes to the igniter once the g tooth passes its sensor (its location ~ 25-30 degrees BDTC). So, 15 degrees for 300 rpm gives 8.33(3) ms for dwell. When the 1st tooth passes it sparks on either 1st or 4th cyl (two g sensors on OEM setup with a distributor).

Could similar behavior be added to the wheel decoder? In my case the first spark would always go the 1st cylinder.

Ps. For tooth #1 I tried 10 degrees (OEM setting), 27, 90, 115. I also tried 0-6-10-15 BDTC for cranking advance Same result. Once started, ignition is spot on. A friend of mine has the same setup but with forged pistons. He has this trouble less frequently. (less compression)

------------
Second thing is the request of feature for programmable ON/OFF outputs. Right now we have up to two conditions that have to be met to activate an output. These conditions are independent.
Is that possible to make an option that makes them dependent? (if first=1 look at the second). For example, it's very useful if second one is 'time' So, the timer starts ticking not since ignition on, but only after the first on is met.


Thanks in advance, this forum and members have been a great help!

Ev.

[jsmcortina]

Megasquirt starting timing is also defined by the teeth.

Best to post datalog, msq and composite logs of failed and good startups.

James

[ekam99]

Megasquirt starting timing is also defined by the teeth.

Best to post datalog, msq and composite logs of failed and good startups.

James

In OEM it's not 'defined', it's just set to be certain Ne tooth (#1 in a squirt language) regardless of rpm, accel/deccel of a crank, etc. Could the limitation of a dwell contribute to this issue? I tried anything from 3 to 8 ms. It didn't matter.

As for the logs, I will post them as soon as backfire/no sync happens again.

Thanks again,
Ev.

[ekam99]

Here we go.

[ekam99]

What I also noticed looking at the scope of signal g ( green line) is that it picks up the tail of another g sensor (all vr sensors share negative g-). That could be a contributing factor as well.(max9926 keeps threshold of 80% only 80 ms, if memory serves me right, than it resets to zero. While cranking the time between to g's is greater than 80 ms). I will try to short this second sensor to g-
Sometimes, when engine starts log does not have any sync errors.

[jsmcortina]

In OEM it's not 'defined', it's just set to be certain Ne tooth (#1 in a squirt language) regardless of rpm, accel/deccel of a crank, etc.

Sounds about the same.

If you think it is syncing on the wrong phase, try "wasted-COP" temporarily.

James