WRX9-10 - optimizing cranking for faster start

Covers current and earlier Subaru and Mitsubishi Plug-In systems

WRX9-10 - optimizing cranking for faster start

Postby mrsaturn7085 on Fri Nov 11, 2016 12:19 pm

Could anyone give me a brief primer on how you go about tuning the cranking enrichment to provide a faster start? As it stands, I am using the base map settings and the car (2006 USDM STI 2.5L) starts in about twice the time the OEM ECU can handle the task. MOTEC cranking time is probably 3 sec to start while OEM was 1.5 sec. It's obviously not bad... but I would like to learn how to optimize it.

I assume the best way to do this is just crank (no pun intended) the compensation value up to add more fuel...

Base map settings are attached as image.
Attachments
cranking.JPG
Cranking Settings
cranking.JPG (14 KiB) Viewed 24838 times
mrsaturn7085
 
Posts: 77
Joined: Tue Dec 22, 2015 11:35 am

Re: WRX9-10 - optimizing cranking for faster start

Postby Stephen Dean on Fri Nov 11, 2016 1:50 pm

It depends on the reason for the delay in starting. If it is needing that extra time to build up enough of a fuel film to provide sufficient fuel to start and run the engine, then yes, adding more fuel to the Cranking comp is going to help. If however, it is waiting to Sync up, before even starting to inject fuel of fire the ignition, then you need to work out why it is that it is slow in syncing.
Stephen Dean
MoTeC Research Centre Melbourne, Australia
User avatar
Stephen Dean
Site Admin
 
Posts: 1724
Joined: Tue Dec 02, 2014 10:29 am
Location: Melbourne

Re: WRX9-10 - optimizing cranking for faster start

Postby mrsaturn7085 on Sat Nov 12, 2016 9:26 am

Any immediate REF/SYNC errors upon IGN clear up within about 0.5 sec of cranking time (Check Engine Light turns-off) - this is VERY repeatable.

In the table I posted, is the F(ET) peak axis value is 200 (base setting) - is this valid? I do not entirely understand how the F(ET) value is calculated; can it reach values above 100%?

I have also noticed that the longest cranking time usually takes place somewhere between a dead-cold engine and fully warm (~90 deg C). Cranking at ET = 60-70 deg C this morning took approximately 5-6 sec to start.

Given this appears to be a lack of adequate fueling, what is the most logical way to reach the optimal value while keeping the F(ET) axis? Creep up on the top left value (50-100 points at a time) until I see diminishing returns on cranking time reduction?
mrsaturn7085
 
Posts: 77
Joined: Tue Dec 22, 2015 11:35 am

Re: WRX9-10 - optimizing cranking for faster start

Postby Scott@FP on Sun Nov 13, 2016 7:15 am

F (ET) is the value you set up in the fuel engine temp comp table.

These tables the value is in percent IJPU, so if your IJPU is high they will need to be much smaller, and if low much larger to achieve the same actual fuel inj comp time in mS.

First verify fuel rail gets pressure quick and you aren't wasting 3 seconds of crank time filling the rail. Most OEM fuel systems hold residual pressure (leaks) well most aftermarket don't. Increase pump prime time and operator delay for cranking if its slow to fill.

Are hot starts OK but cold starts iffy?

Use the first injection, cranking comp and post start tables to all work together.

Set up cranking comp with crank time and ET as axes. Crank time in say 0.5 sec intervals to 4 sec. Set values high at 0 time and taper off fairly rapidly. Set values high at low ET.

Try turning on fuel first injection start with a single value of say 50 see if it helps. If so set up with ET as an axis I find in these cases where manifold wetdown helps it needs quite high values when very cold to very small values warm to slightly higher hot.

Post start helps mainly at cold ET's but it can help at hot ET's too, with little post start needed warm.

Injection timing can also effect starting fuel requirements.

E85/alky can need a very large initial wetdown cold, very large cranking comp and very large post start comps.


WOT is quick and easy, this is the stuff that takes time.
Scott@FP
 
Posts: 421
Joined: Thu Jul 10, 2008 7:57 pm

Re: WRX9-10 - optimizing cranking for faster start

Postby mrsaturn7085 on Mon Nov 14, 2016 6:28 am

Scott@FP wrote:Are hot starts OK but cold starts iffy?


Hot starts are okay, cold starts are... kinda okay, warm starts (ET = 50-60 deg C) are the worst.

Fuel rail and regulator are OEM.

Cold start ET fuel compensation is also the base map, which has worked well, so far. Cold start temps are approximately 20 deg C (25% comp), the rest of the ET comp table is as follows:

20 deg C = 25% ET comp -> 62.5% cranking comp
30 deg C = 13% ET comp -> 32.5% cranking comp
40 deg C = 10% ET comp -> 25% cranking comp
50 deg C = 8% ET comp -> 20% cranking comp
60 deg C = 5% ET comp -> 12.5% cranking comp
70 deg C = 0% ET comp -> 0% cranking comp

Knowing F(ET) is the direct comp. value from the ET comp. table, I've included my effective cranking comp values in the list above.

First injection is set to 60%.

Post-start should not need to be touched - the slow start is entirely during cranking. Once the engine reaches running speed, no rough running, stalling, or lean-out is happening.

I'm wondering if changing the X-axis values of the cranking comp from 0 & 15 s to 0 & 20 s will improve anything... stretching the high comp value into cranking a little more (via widening the interpolation window) seems reasonable.
mrsaturn7085
 
Posts: 77
Joined: Tue Dec 22, 2015 11:35 am

Re: WRX9-10 - optimizing cranking for faster start

Postby Scott@FP on Mon Nov 14, 2016 10:52 am

I find it better to shorten the cranking comp times, and add axis points,taper off comp quickly with high initial values. Gotta wet down the walls before fuel will get into the chamber in a burnable form. TGV may effect startup fuel needs.
Don't be afraid to try higher values. What is your IJPU? If its 2-3 first inj, post start and cranking comps will be quite high values.
Scott@FP
 
Posts: 421
Joined: Thu Jul 10, 2008 7:57 pm

Re: WRX9-10 - optimizing cranking for faster start

Postby mrsaturn7085 on Mon Nov 14, 2016 12:11 pm

IJPU = 11

Engine is a 2.5L EJ257 but using a JDM intake manifold w/o TGV system; all comparisons made to the OEM ECU cranking time was with the same mechanical setup (no TGV system in place).

Extending the time axis to 20 s seemed to start faster (a hair) from dead-cold. Warm restart was still pretty slow, however.

When you say shorten the time and add axis points, are you saying to do something like use (0, 5, 15) for the time axis, and make both the 0 and 5 s columns identical - essentially putting a value of '500' for both x = 0 and 5 when y = 200 (using my original image attachment for reference)?
mrsaturn7085
 
Posts: 77
Joined: Tue Dec 22, 2015 11:35 am

Re: WRX9-10 - optimizing cranking for faster start

Postby Rude Engineering on Tue Nov 15, 2016 4:19 am

It is worth looking at your End of Injection timing, even more so if the TGV have been removed. I would suggest trying the following Injection timing:
0 RPM - 450 degrees
1000 - 450 degrees
1500 - 425 degrees
3000 - 300 degrees
(Interpolate between these numbers)
This should help with idle quality at leaner mixtures as well.
Ross Buckingham
Rude Engineering Ltd. Racing Engine Technology. Hampshire, U.K.
User avatar
Rude Engineering
 
Posts: 27
Joined: Thu Jun 18, 2015 3:17 am
Location: Hampshire, U.K.

Re: WRX9-10 - optimizing cranking for faster start

Postby mrsaturn7085 on Tue Nov 15, 2016 5:56 am

Current injection timing in the relevant range:
Attachments
timing.JPG
inj_timing
timing.JPG (29.49 KiB) Viewed 24779 times
mrsaturn7085
 
Posts: 77
Joined: Tue Dec 22, 2015 11:35 am

Re: WRX9-10 - optimizing cranking for faster start

Postby Rude Engineering on Tue Nov 15, 2016 8:33 pm

I would try setting it at 450 in the starting and idle range. You will find that you can run the engine leaner at idle, it should run smoothly at Lambda = 1 and it may help with starting too.
Ross Buckingham
Rude Engineering Ltd. Racing Engine Technology. Hampshire, U.K.
User avatar
Rude Engineering
 
Posts: 27
Joined: Thu Jun 18, 2015 3:17 am
Location: Hampshire, U.K.


Return to M800 Plug-In ECUs

Who is online

Users browsing this forum: No registered users and 0 guests