Car Starts! Dies after 3 sec

[T3Bunny]

Another vote for the fuel pump not running. I have MSed a bunch of VW's, and repaired a bunch more installs, and seen this EXACT problem 4-5 times. Don't jumper the relay to the battery, but remove the relay compleatly and jumper the power delivery portion for the pump.

If its dieing after three seconds, this is whats going on. It is always three seconds and no fiddling with settings will effect it. I have tried a bunch of ways to get around this too including creating an rpm signal on EDIS setups. Never managed to get the stock relays to work. Its a better idea to bypass the VW fusebox compleatly on the Fuel Pump control anyways.

[Sodaguard]

Ok, I understand. But that cant be it cuz I jumped the fuel pump strait to the battery. I will try it again, I will also give the fuel pump a new ground.

thanks,

Phil

[whittlebeast]

Just a guess. Your reqFuel is too small. At 80 degrees the the 12% ASE is holding for 4 sec and then decaying off to 0. I bet you are rich enough wile ASE is active but when if fades off you are running too lean.

AW

[jhickey]

Attached is a log taken showing several starts. Note: 0 RPM/0 Duty cycle occuring during, or directly after start for one .2sec event. I don't think this is a fueling problem because you can see RPM return in the next event & then it may or may not recover. Somtimes these 0's do not ocurr. I have a VW type 1 with EDIS & MSII. When it runs it gets 60mpg @ 70mph

[whittlebeast]

We need a compleate description of how the ignition is working. This can be fixed

AW

[jhickey]

I believe the spark system is working. Installed by the book & works well in all conditions (except if it is involved with this starting issue). Note: that RPM signal dies yet engine is turning & processor working as witnessed in MAP & Spark signals. Attached is another datalog with comments added. My sense is there a strategy issue in the handoff between crank & afterstart. How does this next pulse work? Should I try a larger tolerance?
Spark settings are:
Offset: 0
Skip Pulse: 3
Algorithim: 1st High, 2nd Low
Gain: 20%
Next Pulse
Crank: 50%
After Start: 50%
Run: 50%
Input capture: Falling Edge
Cranking Trigger: Trigger Return
Coil Charging: EDIS
Spark Output: Going High(Inverted)
Max Duration: 2.0

[grippo]

Why is everyone using Trigger Return. The default is trigger rise and in the case of EDIS it should be "calculated", since in reality EDIS implements trigger rise. Trigger return is only applicable when you have a special trigger wheel with a wide tooth.

Also, the prediction option should be Last Interval, but I don't think using the derivative predictions causes a problem, just unnecessary.

[jhickey]

Thanks Mr Grippo.
I used settings reccomended in the EDIS MSII manual. Copied below for convience:

In MegaTune, set:
Trigger offset = 10° (this will vary, depending on the wheel/pickup configuration),
Ignition Input Capture to 'Falling Edge',
Cranking Trigger to 'Trigger Return',
Coil Charging Scheme to 'EDIS',
Spark Output to 'Going High (Inverted)'.

I will try your recomendations.

Other things I intend to try include increasing the skip pulses from 3 to 6 and increasing the crank to run taper time from 2 to 4 seconds

[Bernard Fife]

Why is everyone using Trigger Return.

Al,

This is entirely my fault. It dates from when there were only two options, calculated and trigger return. Some people had reported problems with calculated for the cranking trigger, and trigger return worked better for some reason.

However, that was quite a while ago, and I never updated the documents when the trigger rise option became available. In any case, I have set the docs to have people use 'calculated' for the cranking trigger.

Lance.

[grippo]

That explains it. But if it works even for some people, then we don't want to take it out - maybe tell them to try each one. But it would be nice if we understood it or at least had a feel for which of the 3 options works best. Hopefully when jhickey tries again we will get some information.

From a code standpoint, based on a quick review, I think the following is going on:

In trigger rise it is firing spark (meaning holding the SAW line low) whenever it gets a tach input. This has nothing to do with the SAW advance signal, nor does EDIS care about SAW during cranking.

In trigger return, things are similar, but when it goes past cranking speed it will reverse polarity of the tach in that it gets from EDIS. This is not good unless the polarity was wrong at the start.

In calculated mode, it will start providing the SAW pulse to EDIS immediately and EDIS should ignore it until it determines it has finished cranking. If this is causing a problem, it would be very easy to inhibit sending the SAW until cranking speed is reached.