I flashed the car with the coding in place from 2000cc back to ID1000, with the IPW minimum set to 1.280. Car would not start like that... I took the coding and scaling out and reflashed the car so that the multiplier was back at 32 as stock, and the car fired up first try...

 

oh HOLD ON.

you need to modify the code to make use of the 32x change

96530006 right?

 

yes 9653... I have no problems with idling on the ID1000s, I figured I would have problems with the 2000s though... When I had the 2000s in with the IPW minimum patch, it was set to 1.1 or there abouts, then when I swaped back to the ID1000 map, I bumped it back up to 1.28 so it would be back to stock values... Did I just rock the whole process?

 

This patch modifies the default step from 0.256ms to 0.008ms

This is ONLY for the Minimum IPW

96530006:
After doing the code changes (top 3) you will then need to use the "new" Minimum IPW SHLL0 table to set your Minimum IPW - you will need to replace what the stock value is with ~1.1 or whatever you want. Feel free to use +/- keys to increment/decrement

 

There are two things going on here

1) Is the minimum IPW - by default this is set to 1.28ms (unless your in MrFreds list in which case its 1.024ms)

2) The step/jump - normally this is 0.256ms, my patch in the above post reduces this from 0.256ms to 0.008ms

 

Gotcha, I was using the scaling and code from page 3 to try and adjust that...

 

Looking at the IDC2000 charts, it seems like 1.1ms minimum puts you right at the point where opening becomes reliable, but it is still very non-linear.

This is the graph relating the difference between actual IPW and what the injectors would deliver in fuel if they were perfectly linear once the injector latency was overcome.



Here, I've eliminated all but the 14V trace and zoomed in to the area of interest.

IMO, the light blue line shows likely where the injector will start to reliably open. Earlier then this, the spike and low spot to me look like they may be unstable flow areas. The yellow area highlights the difference between the theoretical fuel flow and reality. We could easily use the 66 element table to linearize the injectors through this region, making fuel delivery below 2ms MUCH more linear and predictable, hopefully matching the straight red line. I think this may lead to an improvement in light cruise and idle control.





One last thought, but shouldn't we be looking at Injector Pulsewidth without latency for any kind of fuel calcs we do? In 96530006, it would be FFFF8A6A. Although that address might not update in a manner that lends well to datalogging and a patch may be needed to provide a loggable channel?

 

Tephra the x15 multiplier is fairly common, they used it on the DSMs from 95-99.
You get more resolution, but can max the table out at low voltages with high deadtime injectors.


Are you finding the pulsewidth limits in the big sub that first checks RevLimit, then has the Load/LoadClipped decision about halfway down?

 

yep

 

Throw some ID2000's in mine

 

Aight, put the big boys back in today after work, for some reason they fired right up today??? Anyway with all stock settings they were idling at 11.9... a bit rich, but liveable. I applied the IPW minumum patch and started messing with it... it gets squirly under 1.1 until you get to where the STFT is basically always adding fuel to keep idle... at 1.048, it idles at 14.5-15.3 with the STFT adding 8% of fuel steady, and kinda settles down a bit... I unhooked the battery so I can see what the trims are gonna do... will report back later.

 

03whitegsr - yeh at 1.28ms resulted in an idle AFR about 25% too rich.

that means we would need to run an IPW of about 1.124 to get stoich.

however @ 1.124ms the deviation is about -30%...

i feel like the mid point between 1.124ms and 1.28ms will probably result in the right amount of fuel being injected reliably.

 

fast freddie - are these the FIC2150's?

from what your saying it seems like they have a lot better sub 1.5ms control... I wonder if Jens has a fact sheet like ID does?

 

No, these are the SlowBoy variants... got them for a good deal, so I went for it... Also try your injectors at 1.040... mine were pulling fuel like mad as well, until I lowered the IPW minimum to the point were it started adding fuel instead, it idles steady at 14.7ish, with the occasional dip etc... I was running pretty lean at WOT though, I eyeballed the fuel map, so we will see...

 

yeah anything under 1.120ms the injectors don't like

STFT is adding fuel, and apparently a rough idle..

So we will keep going up until we get close to 0% STFT then tune from there