IPW minimum?
 

0xDEAD and I are trying to tune a e8 96531706 SD v7 car. The specs of the car are a 2.3L running double pumper walbro's with ID2000's.

We have the everything running OK, the injectors are scaled and the car is running good once we start out. However at idle the car is crazy rich (in the high 10's!) We believe our issue is that the IPW at idle needs to be around 0.6 ms using this formula posted by mrfred to mathematically figure out the new IPW:

(old startup IPW value)*513/(current injector scaling)

However at idle I am only seeing a minimum of 1.28ms is this the lowest the ECU can spit out? If so, is there a way we can lower that limit?

PLEASE HELP!

Basically I can't get it to lean out more than 1.28ms at idle. I reduced the load in the kpa vs load table and reduced the VE at idle RPM's and am not seeing a reduction in IPW. I realize that a large change in the SD settings will have a relatively small effect on a 1.28 ms pulsewidth so I'd like to know if other people have seen sub 1.28ms injector pulsewidth at idle or if I am fighting some sort of "minimum injector pulsewidth" setting in the ecu.

I see a VERY occasional dip to 1.024ms, but only after a slight tap on the throttle from idle.

Normal idle for me is 1.28ms.

Are you running the ID2000's? If so what are your settings? Also what is your AFR at idle like?

No, sorry, I'm running the PTE880s, just thought I'd throw the info out there incase it helps :)

AFR at idle is pretty good, 14.5-14.6, trims are good.

Please note, I used the LM Programmer to trim the analogue output rate to 1/12s.
If left on instant, things got a bit 'funky'.

http://i250.photobucket.com/albums/g...w/afr_idle.jpg

Here's an example of the dip:

http://i250.photobucket.com/albums/g...j_w/pw_dip.jpg

If you like, when I get time, I'll disable closed loop and lean it out a bit, see what I can get the PW down to...

Ok, good info. I didn't have much time with the car yesterday but it looks like we need to put a little work into it. I just wanted to make sure I wasn't fighting a hard limit.

I have mine running 1.024 pretty consistent going back through my logs. The IPW isnt the one that you want to rescale that way, its the IDC. Your IPW is alot higher than 1.28 if I read your formula right. Doing the math that way-

1.536 * 513 = 787.968 / 1462 (guessing based on mine)= .5389

I just leave that alone on my cars (actually I dont play with the IDC stuff either) and let it idle at 1.024 which is a pretty consistent -7% on STFT and 14.0 or so in open loop. What are your idle RPMVE settings and low vacuum MAPVE (i.e. 31/41/61)?

What's the dead time?

Seems like 0.6ms dead time and 0.6ms desired IPW would put you up around the 1.2ms you are seeing?

Also the IPW MUT channel is just the 1-byte, correct? You may be seeing a true IPW between the 1.024 and the 1.28 and it's just rounding up.

Here is what we have for SD settings, no matter what we do or change the car wants to idle at 10.9/11!!! Can't figure this out.....

http://thefrost.net/randomfiles/tuni...sdsettings.jpg

It looks like you may have too much VE at idle. What map does this car idle at...40kpa? If so, then you're at 100% VE map and 84% VE RPM, for a total of 84% VE. My guess is the true VE is probably more around 60-70% or so. That may be why you are running so rich.

IPW MUT being logged is 1-byte. My recollection is that unlike most other 1-byte values, IPW MUT is the high byte rather than the low byte. So, this means that the logged resolution is less than what actually occurs. I believe that 03whitegsr is correct about IPW MUT including latency. I'll double check on Monday night when I have access to my PC laptop again.

My DSM with evo ecu and evo 550's is doing the same thing, 1.28ms min. Trims are 0 at idle and STFT is hunting around 0 like it should. You would think that stock injectors would idle at less than 1.28ms. I think mrfred may be right, the value that we see in the logger is just the minimum that can be displayed with 1 byte, which is fine by me.

MUT 29 and MUT 2B are the IPW 2-byte channels. I don't have evoscan open, but I would imagine it's only logging MUT29.

I would think the IPW would be near the final IPW with all correction and latency taken into account?

Yes, logging MUT29. Here is a screenshot.

Might be interesting to log 2-Byte just to make sure it is varying and you are not hitting some minimum pulsewidth.

Probably not the case though and it is varying, just not enough to hit the next lower step.

Are you finding the RPM and MAP VE table you are using provided good off-idle driveability and AFRs but then idle is extremely rich?

Or is the car rich in general at low loads and engine speeds?

My car on the IDC1000s has AFRs slightly on the lean side just of idle, but it stills wants to idle insanely rich. I used the MAFComp table to pull out a bunch of fuel in the idle airspeed range.