Atom g4+start timing

[James c]

Hi.  I have a high compression zetec race engine (12.0:1 )  that  is very prone to starter kickback.  It has a high tourqe starter and new battery etc but the only way I can get it to crank freely is to drop the 0 to 500rpm and 500 to 1000 rpm timing to 1degree.  It then spins nicely and will start but the trouble is when it's cold idling it runs in this rpm range and with +1 degree timing it obviously won't idle very well.  Once warmed up its above this rpm range (1000 to 1500rpm is 16degree timing) and it runs nicely.  Is there a separate setting for start timing that doesn't use the main map?  This way I could have start timing at 1 degree and the main map at around 10degree in the 500 to 1000 window.  

[Scott]

You could use a virtual auxiliary channel to trigger a 4D overlay ignition table. You could set the conditions of activation of the virtual aux channel based upon engine speed and ECT. Here is an example of how it might look, you would adjust the actual values to suit your setup.

[James c]

Thanks for your reply.  Could you elaborate on how to do this as I don't quiet understand ?  Keeping in mind that I still need it to start under 1degree timing when the engine is at operating temp as well. Cheers

[Scott]

If you need the under 1 degree timing at starting regardless of operating temp, then you are probably best to add extra engine speed columns to the main ignition table. I'm unsure what values you have in your ign table, but this is how it could look:

You can adjust the x axis values by right clicking on the table and selecting axis setup, or by pressing x when the table is the active view.

Edited April 29, 2016 by Scott

[James c]

Hi Scott. Thanks for your help. I have added a few more columns ( 400 and 700rpm) and it will crank freely and eventually start but it's a bit of a loose loose situation because it won't fire quickly due to the fact the timing is so retarded.  As soon as I go over 2degrees it starts to stall the starter.  Not to sure how to overcome this considering the starter is already a high tourqe unit.  

[Scott]

What values do you have entered in the 0, 400, and 700 rpm cells? If you look at the ignition angle parameter during cranking is the ECU applying the angle specified in the 0-400 rpm cells?

 

[Scott]

Hi James,

I've just re-read your last post and understand the problem better now. One possible solution is to have the ECU receiving the start switch signal, and then control the starter motor and ignition angle. Here is how it could work:

1. Digital Input 1 is activated by start switch. This is momentary (a quick flick of the key or press of a button).
2. Digital Input  1 activates starter control, this will keep the starter cranking for 4.0 seconds/until the engine speed is over 400 RPM, which ever comes first. Digital Input 1 also activates a timer that runs for 5.0s and then stops. The timer will restart if the key is momentarily flicked to start position again.
3. The timer activates Virtual Aux 1. Virtual Aux 1 will stay active while the timer is less than 3.0s. Virtual Aux 1 will not activate if engine speed is over 800 rpm.
4. Virtual Aux 1 causes Ignition table 2 to become active, and Ignition 1 to deactivate.
5. Ignition table 2 progressively increases ignition angle from 0 to whatever the normal idle ignition angle is (I've used 15 degrees). It does this based upon the number of engine cycles that have occurred.
6. Timer 1 goes over 3.0s, Virtual Aux 1 turns off, this causes Ignition table 2 to deactivate, and Ignition table 1 becomes active.

You would need to tweak the settings to values that actually work well for your engine.

 

[Adamw]

James,

Have you ever actually put a timing light on the engine during cranking to check the timing is steady and as commanded?  I wouldnt expect any kick back issues in a modern 4 valve pentroof engine like this.  Does yours have the 36-1 trigger?  Are the missing teeth away from TDC?  Is cranking RPM normal?  I recently tuned a high comp, low inertia,  NA zetec in a sports car and that didn't have any problem cranking or need any special tricks to aid starting - it only had a little motorbike battery too...

[James c]

Yes it's a 36-1 trigger and running waste spark.  They have no timing marks from factory so all I have is a TDC mark on the pulley to work off. Cranking rpm seems normal.  Not sure if the missing teeth are away from TDC but the teeth are part of the flywheel which can only go on one way so they should be as per factory.   It will start but it doesn't  like to do so with any amount of timing in it and is worse when hot which I put down to the fact that there's even more compression when hot plus possibly some starter heat soak.   If I crank it then give it some throttle which obviously adds timing it will stall the starter. 

 

[Adamw]

Im still pretty confident it is just something with your setup that has been over looked, an engine like this shouldn't be difficult to start.  A few suggestions or thoughts from me...

They have no timing marks from factory so all I have is a TDC mark on the pulley to work off.

So how did you calibrate the base timing?  Does it run ok locked to zero degrees or did you use a dial back light?

I would still be getting a timing light on it during cranking to ensure you don't have major scatter.  Even if you have to paint a temporary mark on the pulley for reference - it doesn't need to be accurate just something to prove timing is not jumping around.

Have you scoped the trigger to ensure the polarity is correct?  36-1 inductive will often still start/run either way but with errors and drift if connected wrong.

Could you do a compression test? Warm engine, plugs out, wide open throttle.  I know you have said 12:1 static but perhaps dynamic compression might be extraordinary for some reason - cam timing etc.

Finally could you post a PC log (started before cranking and with all parameters included) of a good start and one of a bad start?

 

 

[James c]

Hi Adam

I have done a hot compression test and it resulted in 130psi across all 4 which I put down to valve overlap from the cams so this has eliminated to much cranking compression.  I just put a timing light on it and when cranking the timing is all over the place.  What causes this scatter?  I think your on to something.  Cheers

[jonnyt80]

Run the trigger scope to check the crank signal. It could the the polarity is wrong 

[Adamw]

Run the trigger scope to check the crank signal. It could the the polarity is wrong 

Unfortunately this is an Atom so he doesnt have the scope function.

 

I just put a timing light on it and when cranking the timing is all over the place.  What causes this scatter?  I think your on to something.  Cheers

That indicates a trigger problem.  Since most of your trigger system is relatively "fixed" and OEM quality I think the most likely problem is the polarity of your crank sensor is wrong.  Ideally you would get a standalone scope on to it to check but if one is not easily accessible then I would probably just try swapping the wires over and see if it improves the scatter. You will need to do your trigger offset calibration again after swapping polarity.

 

I have done a hot compression test and it resulted in 130psi across all 4 which I put down to valve overlap from the cams so this has eliminated to much cranking compression. 

Something seems wrong here to me too.  Even with very big cams I would expect ~200psi.  It may indicate you have a cam timing issue also. 

Edited May 21, 2016 by Adamw
formating

[James c]

Would the car run fine with the crank sensor wired the wrong way around because apart from the starting issue it runs really well and makes good power.  

[Adamw]

Would the car run fine with the crank sensor wired the wrong way around because apart from the starting issue it runs really well and makes good power.  

Oddly enough they will still often run reasonable with the polarity reversed, although usually with more obvious symptoms than yours is seemingly exhibiting.  In theory you would think if you trigger off the wrong edge of a missing tooth waveform the ecu should see extra teeth and not run at all but I assume there must be some error logic built in somewhere that resets the tooth count every cycle and allows it to continue to run.  I have had numerous cars "driven" to me for tuning and have found them running with wrong sensor polarity or trigger edge.

Im still not saying this is definitely your problem, just one thing you can easily try to eliminate it from the troubleshooting. 

[James c]

I am going to swap them and see what happens.  The car was wired and setup by a well known tuning shop so I would have expected it to be right but I guess there's always the element of human error.  I unplugged the sensor and the black trigger 1 wire is to earth according to information I have found so it is looking that way.  Im just puzzled as to how it can run great in all other areas.  The other thought I had was weather it's getting electrical interference from the starter motor which is mounted right next to the cps.  The cable does appear to be screened though.

[James c]

Just an update, I have switched the crank sensor polarity and recalibrated the base timing and what a difference.  I have moved the 500 to 700 rpm back up to 10degrees and it starts beautifully. No stalling at all.  Thanks Adam you were right on the money and thanks Scott for your help to. Much appreciated 

[Adamw]

Good to hear.  Now that raises the question if the tuner tuned the ignition map to MBT with the trigger connected back to front with lots of scatter and drift then your ignition map is probably not great.  Depending on what fuel you are using I suspect your engine shouldn't be knock limited so its probably not high risk but I would certainly load that engine with caution until you are happy there is no knock.

[James c]

You think I should get it tuned again?  I don't have equipment to check for knock apart from my ears.  It runs on 98octane.