Engine wont start, sparking correct and injection occuring..

[zolar]

I've uploaded the primary trigger settings and secondary trigger settings
http://www.webster-group.co.nz/robert/Primary%20Trigger%20Settings.JPG
http://www.webster-group.co.nz/robert/Secondary%20Trigger%20Settings.JPG

the config is also available, and I've taken the falling, rising (didnt work however) trigger log also available at
http://www.webster-group.co.nz/robert/

The sensor I'm using is in the front/middle position, not the rear one. Same as used in Fero's documentation.

Should I be trying the rear one?
I understand its only 360' out, so surely it'd be the same result?

Thanks

[mattias]

"Falling, rising" looks way better, this should work.



What VemsTune version are you using?

I can't load your triggerlogs. I'm using 2011-02-10.

Under ECU calibrations, the high and low bytes for ECU calibrations should be 4 and 215.
If they're not - you won't get any spark, fuel or nothing as no outputs will activate until a proper config has been uploaded. An upgrade to 1.1.87 with an old VemsTune unaware of how to properly set this up will not work.  It's described here :
http://www.vems.hu/wiki/index.php?page=GenBoard%2FUnderDevelopment%2FFirmwareChanges

[zolar]

I've done a few things now, upgraded Vemstune to 2011-02-10, Saved and reuploaded the config again, as per the Safety mode instructions
I checked the high, low byte settings, mine read: High:132 Low:215.
Tried to change it and it came up with safe mode when I attempted to run it, so reverted back to High:132 Low:215, also tried the High:190 Low:239 just in case...

Sorry I've been uploading the CSV file, after clicking the view results button when doing a triggerlog.
I've uploaded the latest triggerlog however.

Falling, Rising did appear to work better with the plugs out.. no trigger errors, set up the timing for TDC cyl#1, all ok. Replaced the plugs and tried it again, but started getting trigger errors, so i turned on filtering for the primary trigger and it went away (I assumed this might be due to the low rev's ~220 RPM)
this is the latest config I'm using http://www.webster-group.co.nz/robert/v3.3_n002286-2011.02.22-23.33.18 (Falling, Rising).vemscfg

I am unable to get the engine to actually fire up, injector pulse width gets to about 0.4ms maximum, so I quickly used the injector test function to dump a bit of fuel in before it ran (Not a good idea usually I know..)
It then turned over and fired up for a few seconds, but while the RPM was higher it only gave trigger errors again..
http://www.webster-group.co.nz/robert/Falling, Rising - Firing over but stopping, trigger errors.png
http://www.webster-group.co.nz/robert/v3.3_n002286-2011-02-22-23.19.25 Falling, Rising, Firing but stopping.triggerlog
http://www.webster-group.co.nz/robert/v3.3_n002286-2011-02-22-23.19.25%20Falling,%20Rising,%20Firing%20but%20stopping.csv

I can clearly see that about the same time as "MISSING TOOTH: 1, cycle=25" changed to "MISSING TOOTH: 2, cycle=25" would've been when the engine started firing over, the same time as the trigger errors started coming up.


Also have I got something wrong in my config for injector starting pulse width to be so low or is 0.6ms normal?
Thanks

[rob@vems.co.uk]

Yeah 0.6 is too low, but while there are trigger errors you should not be thinking of adding fuel yet.

Why do you have missing a missing tooth value at all?

Terrible news about Christchurch.

[zolar]

I assumed the missing tooth value that comes up in the triggerlog is simply a description thats been used for a secondary trigger.. is it not?

If not, what would be wrong with the configuration, I'm sure I've selected Coil-Type and everything else correctly to suit a 12+1 trigger..

I wasnt able to get a trigger error until the RPM's were much higher than I could achieve with the starter and the plugs out, which was why I put some fuel in..

I did read this on http://www.vems.hu/wiki/index.php?page=InputTrigger%2FTriggerLog

new triggerlog implementation (since 1.1.78) can also log secondary trigger timestamps (and use CRC checksum)

    * so the margin between secondary and primary trigger pulses can be measured without scope.
    * This must be minimum 10 crankdegrees from the missing tooth (if a missing tooth wheel)
    * or min 10 crankdegrees from the last tooth for other triggers, eg. coil type or InputTrigger/MultiToothNoneMissing.
    * Except InputTrigger/AudiTrigger where it's quite close to the previous tooth (often around 1 crankdegree), that is OK.

Could I be coming up because the secondary trigger is too close (less than 10 degrees) from the primary trigger?


It is terrible about Christchurch, I'm in Auckland however but I might end up going down there yet, to assist with the supply and connection of generators..

[mattias]

Your triggerlog looks healthy. Trigger errors don't just appear from nowhere. I suspect a grounding/shielding issue, either the coils or the sensor wiring.

So you get no trigger errors if you :  Remove the plugs, unplug the coils, and turn the engine ?

If you have power to the coils, they must be properly grounded - most have a shield ground that attaches directly to the head. When the coils are powered and the plugs are out of the engine but attached to the coils, then the plugs must (should be) properly grounded to the head using naked copper wire wound around each plug, or the spark can arc and jump to bad places - potentially causing RFI or killing the coil.

I'm not all clear on this, what type of sensors are you using? Are both pri/sec of Hall type?

[zolar]

Sorry for a slow reply, had a weekend away - went for a drive in a Ferrari Enzo at a track day!
It really is an adreneline rush 

To answer your questions, I've converted both sensors to hall effect (GT101) type, using the existing stock VR shielded wiring, the negative connects to the head of the car, the 5V+ through one wire of the VR wiring, the signal through the other wire.

The coils are the stock Coil On Plug, and are well earthed to the head..

I agree, trigger errors dont just appear from nowhere, but during that log there was a trigger error coming up.. surely that would point to a configuration issue if the log appears ok..?

I will double check that there is no trigger errors when turning the engine over without the plugs in tommorrow, as I cant remember if I'd altered the configuration at all since those last logs I took.. will confirm that tommorrow.

Thanks for all the constant replies.

[zolar]

Just tried with plugs in and fuel pump on, but engine shows no signs of starting, trigger doesnt come up with any errors, ~200+ RPM, Pulse Width is 0.5ms, battery voltage 10.5v during cranking.

Trigger log for this here: http://www.webster-group.co.nz/robert/Falling,%20Rising%20-%20Plugs%20In,%20Fuel%20Pump%20on,%20no%20start.png
http://www.webster-group.co.nz/robert/v3.3_n002286-2011-03-01-21.31.33%20-%20Falling,%20Rising%20-%20Plugs%20In,%20Fuel%20Pump%20on,%20no%20start.triggerlog



Now with plugs out, fuel pump off, trigger IS showing errors: 'Sec Trig Pos Bad', 260+ RPM, Pulse Width is 0.9ms, battery voltage 11.0v during cranking.
This disappears after cranking for more than a few seconds, perhaps this is just the initial loss of cam signal, as it sometimes doesnt come up within the first revolution of the engine?

Trigger log for this is here:
http://www.webster-group.co.nz/robert/Falling,%20Rising%20-%20Plugs%20Out%203.png
http://www.webster-group.co.nz/robert/v3.3_n002286-2011-03-01-22.08.55%20Falling,%20Rising%20-%20Plugs%20Out%203.triggerlog

Sorry for any confusion about this..

I notice that in the trigger logs when viewing the result, it still says missing tooth when a cam signal is recieved, can someone confirm if this is normal or if I've got something set wrong?

[mattias]

The missing-tooth thing is normal. Or should I say, it will be fixed to make more sense in the future.
The trigger error at the beginning is also normal, it needs to sync up and the light stays on for a minimum amount of time so that you can have time to notice a potential problem.

Injector pulsewidth of 0.5 ms sounds like far too little fuel while cranking. Are you sure you're not confusing this with the prime pulse? (which you shouldn't have to use, set that to 0)

It might not be a problem just yet, but your trigger wiring sounds bad, the ground should be connected to sensor ground at the ECU, not to the engine at the sensor.

I like that you share triggerlogs and stuff, it's very important and I say to others  : Always datalog. Just press Alt-L whenever you connect or start cranking the engine. Having no log is useless and you can't really describe everything that a datalog can. You always have the live datalog, but that can't be shared with other people.

I would like to see a .vemslog (includes the config) from the failed start attempt if possible.

[zolar]

Sorry for the slow replies lately..

I hadnt thought much about datalogging anything at this stage, but seeing as it now appears to be good as far as triggers go.. I've carried out a vems log, and uploaded this:
http://www.webster-group.co.nz/robert/v3.3_n002286-2011.03.08-22.04.01.vemslog

Also set primepulse to zero, however I still see pulse width jump up as high as 0.9 when cranking with the plugs out..

As a side note, I get an error that has come up in the validation/config sanity checks:
Following symbols have invalid values
launch_out_conf_speed has INVALID value (0)
This symbol is not used by any dialog

*Also of note for the developers of vemstune: For some reason im unable to copy the text displayed within the validation/sanity check, either by right-clicking, or selecting and Ctrl+C
I typed the above out as I saw it - not a major, just a odd thing..

[mattias]

Check "view -> group -> calc model" to see the injector pulse width calculation, it's all correct.
Don't use prime pulse, it's for broken fuel systems.

You are using Alpha-N (TPS) as the main fuel control strategy, change that to Speed-Density (kPa) as I'm quite sure that's what you want.
Check the first dialogs in the Base Setup menu, and after that's done I would drag'n'drop a suitable config with VE table over your VE table dialog to get a proper scale of everything.

You might want to lower the crank VE% back down to 60-80% or something more reasonable.
Also in afterstart and cranking tables, a hot engine rarely needs any enrichment - use 100% at those temperatures.

[zolar]

Its running!! yay  

hard to believe its been over a year since I've heard it run.

the speed density did the trick..

thanks Mattias and Rob for all your replies and knowledge!

now to start on some tuning 

[zolar]

Sorry to drag an old posting up, I'll repost this elsewhere if it's prefered but thought the history might be worth keeping in one topic.

After I got the car running (As above), I had to disappear down the line for work for a month, so finally got back and fired it up. In the meantime I put the vems unit back in its case (It was out while I was carrying out the above testing, etc.)

Fired up the car but it wouldnt start, nor could I communicate with it, suspecting i'd shorted something I pulled it back out, had a look - nothing obvious, tried again.. noticed one of the IGBT's burning hot!!
Powered off the ecu, had an electronics friend test the IGBT's themselves, all ok..

so checked everything again and found a disconnected main earth/ground wire E36-Pin26!!! Arrrggg, can only blame myself however as I'd disconnected this to check the resistance at some point in time earlier..

Reconnected it all, tried again - runs, but sounds like its running on 3 or 4 cylinders..

Now comes the funny part, I notice every 10 seconds or so, all the coil packs momentarily spark about 3 times rapidly at the same time.. is this a feature and this is normal or is something wrong?
Although occasionally with a test light on even during this time I see the odd flicker, seemingly at random.
Using the test feature for the igniters seems to do some strange things, sometimes no response, sometimes they all operate.

Could someone let me know their thoughts, hardware fault or firmware corrupt? etc

[mattias]

Send it in for repair..