I'm pretty sure he's not even running closed loop fueling at idle. :)

Lowered idle to 950 (from 1000)... lowered that open RPM value and made a HUGE difference. I also increased the open valve PWM back to 65 or whatever I had before... Going any higher than what I used for P and I values resulted in pretty good oscillations so I think this is about as good as I'm gonna get it. I'm gonna mess around with idle advance next...if I accomplish anything I'll post that up as well. I had to actually lower my max decel load if anything. I tried 33 and it would never enter back into CL.

http://lh4.ggpht.com/_9Kr_cGb9rtI/TI...ddtU/idle3.png

I'm not, never have actually, do you?

I'm using the 4x4 VE idle table. I idle around 14.5:1 now. I can idle it smooth at 16.0:1 if I want now with seq. injection.

I'll go out at lunch and get a good log of my idle with and without a/c.

Yup, I am, one of those great moments after upgrading from my MS1 to MS2. IAT heatsoaking be-gone!

hmm... maybe I should give it a shot, just never thought of it really. Are you just letting EGO run above 600RPM or so?

I've played close attention to how the AIT heat corrections work. IIRC, at exactly 82*F it subtracts 1% of fuel, so I've catered my AIT corrections table to simply counteract what the MS is trying to pull. If I have a warm start, it idles lean, but it doesn't have a problem idling so I just deal with it until it starts moving again.

I'm letting EGO control above 800rpm with 8 ignition events per step. That makes it quick to react without oscillating too much. My idle is pretty much 1000rpm most of the time as I use the A/C 8 months per year.

look at that...works much better on a hot start. What's interesting is now my a/c idle up is overshooting and I had to lower the % increase, I guess since the fueling is reacting as well and helping things out.

Ye I run CL fueling at idle too.

I have to hit stoich to pass emmissions, (I could probably get away with it due to a friendly tester, but I'd rather not abuse the favours).

Running CL afr for idle as well @ 14.7

I just kept the standard 16 events per step, gonna give 8 a shot

I recommend strongly against running CL EGO on idle.

You should be able to get very close to the AFR you want in all conditions if you adjust the baro and MAT correction tables.

Using CL EGO on idle along with CL control for the idle air can cause the two algorithms to start "fighting" each other, getting you into some oscillation. You'll end up having to reduce the PID numbers on one or the other to compensate (usually the EGO) defeating the purpose of running it at idle.

I usually don't start close loop EGO until I'm in a cruise situation.

Ken

I hear what your saying Ken, but I've managed to make it work along with two other members on the board.

If it's good enough for OEM's ;)

Ken,

it actually works like a charm if you have your fuel table flat at the idle cells and dialed in well. Believe it or not, I would never switch back to open loop at idle.

Dimitris

this is exactly what I noticed on my drive home, when coming off a/c the PWM% and EGO were fighting. My idle was much better without it.

I didnt notice before but today my idle did seem to be fighting with the EGO correction... I instead did what Ken suggested, made idle Open loop and fine tuned the MAT correction and it worked like a charm. I needed a pretty big correction.. I am guessing the stock IAT in the stock airbox is not quite the ideal location for accurate air temps..

stock AIT gets hot and stays hot.


Here's how my idle looks like when cyling the a/c on/off. I'm using Idle advance to keep the timing at 10* without the a/c and 22* with the a/c. I found that's where my car likes to run in each condition, I might try to wean it off so much advance in a/c, just haven't had time to play.



https://www.miataturbo.net/attachmen...1&d=1284036413

this is open loop, fwiw. I'll log a hot start at lunch, although today was just 65*F out, unlike yesterday it was +80*F at lunch. It'll idle lean around 15.5:1, but it's still stable. I might simply try some insulation around the sensor itself and see what that will do.

It looks like the AFR goes a little unstable when you turn on the AC. You might want to try tuning that out too as it looks like it's causing slight (but not terrible) oscillations in idle speed.

Ken

I just tuned my closed loop yesterday. I am getting a pretty solid 900-950rpm idle. I tuned it with the IAC unplugged initially.

Here is my concern. Without the IAC, i.e. unplugged, the car idles at about 790 +/- 30rpm. It is stable with lights on, ac on, etc. It didn't matter what I did, the idle remained stable.

What is the point of running the IAC if the idle is worse than without it??? I think I am going to run without it unless there is something very negative about this.

By the way. Thanks for all the info in this thread. My values are very close to the ones posted here. I think these are a very good starting point. I still recommend following the msextra.com MS3 closed loop tuning instructions.

Changes in temp withouth an IAC means a change in idle speed, that's all the IAC is supposed to regulate.

i.e. you'll *always* have an idle of 850rpm, whether it's 0DegC or 40DegC outside

FYI

I had my closed loop idle working flawlessly this morning. Then all of a sudden my smooth 780rpm idle drops to 500 then 450, and the car stalls. Fortunately I was datalogging in case this happened. What I found when reviewing the log was that my smooth idle was at 17 degrees advance, and when it died the timing had changed to 15 degrees. I checked my timing table and it was set at 17 degrees in this rpm range. However, when I checked the "idle advance timing" it was set to 15 degrees. This can't be coincidence. I changed the idle advance to 17 degrees and i will test it later. I thought I would post this in case anyone had similar issue and wanted to test this theory.

the advance dropping to 15 shouldn't be an issue. i idle at 15* normally, and 22* with a/c on.

what was the idle valve doing when the rpms dropped?