Advice on tuning Close loop idle

[richyvrlimited]

Spec is in my sig, I've got the idle nice and stable for the most part, but adding a load sends it into oscillations (see datalog) and sometimes it'll just start oscillating with no extra load, I've yet to catch that on a datalog though!

I can't really add anymore 'I' as that just causes it to oscillate all the time...

Any suggestions on what too look at next?

Cheers!

P.S. Big kudos on the PID code for MSII it is far and away superior to MS1's closed loop idle. At least with the MSII code I could attempt to tune it myself and get somewhere! with MS1 I just went around in circles for months ;)

[muythaibxr]

I'm at work so can't look at the log/settings yet, but usually when I see that sort of thing happen, there is some external instability causing it, such as AFR fluctuations or timing is moving around too much.

Ken

[richyvrlimited]

Timing and fuelling is static in those areas. I struggle to hold a really steady idle AFR mostly due to 600cc injectors. I've found EGO correction can help a LOT at idle but I refuse to turn it on until I can get a steady idle

Actually whist this is mentioned.... To get a static AFR in the 500 - 1500 rpm range you need different figures in the VE table correct? BUT having different figures in the VE table at idle causes instability/oscillations in my experience.

So what do you do? have varying AFR's or have steady AFR's and varying VE numbers

Edited to add the bit in bold, pretty important bit to miss

[Black99rt]

Based on what I've read and Ken's post in this thread, stablizing AFR should be your goal, the numbers in the VE table are just a means to that goal. Don't be afraid to play around with your binning a bit if you need to.

[richyvrlimited]

I'm really surprised at that, I've experienced bad oscillation when the bins around Idle aren't the same/very very similar.

Also doesn't this advice (static AFR) fly in the face a bit of advice to have low rpm areas higher in timing and richer to help combat stalls?

I'll certainly give it a shot though.

Just to check though. AFR moves, it's never static, how much of a swing will cause issues? a full point i.e. 14:1 - 13:1 more than that or less than that?

[muythaibxr]

It depends... the leaner you run your idle, the more stable it has to be, and the more a small change in AFR will cause a larger change in RPM.

If you're trying to idle at above 14:1, then you need to make AFR nearly flat during load increases, etc... otherwise you'll get the PID code fighting with the RPM change that going rich then lean and back and forth causes.

I usually try to keep AFR fluctuation less than a few tenths, and I try to "box in" the idle... so I'm not idling on a load or rpm point, I'm idling right in the middle of 4 points that box in the idle. I do that for no-load idle, and then again for high-load idle (all accessories on) and get the AFR where I want it in both cases.

Another thing that can help here is to use slightly less than optimal timing at idle. This tends to make the engine a little less responsive to other changes, and makes it easier for the PID code to control the idle speed.

What I usually do is with warmup-only idle speed control, I get the engine to the point where when I turn on any accessory suddenly, it doesn't bounce the idle around and slowly oscillate to a stable point. Instead I tune fuel and spark so that when I suddenly turn on an accessory, load goes up and idle drops and just stays there at the new spot. Once it gets to that point, the closed loop idle speed control code will no longer fight with AFR and timing changes, and instead will be able to properly respond to changes.

Ken

[richyvrlimited]

Cheers for that Ken, I'll print it off and try it tomorrow/Friday very handy advice.

My final goal is to idle at 14.7:1 but at this point I'm idling richer than that as I wanted to dial in idle before I dialled in the fuelling 100%

FWIW I took this datalog on my way to the shops earlier, just an example of Idle dip and unrecoverable oscillation (unless you manually rev the car in which case it settles down and idles steady).

Cheers for the advice/help!

[xr2wishy]

i have the same problem as it goes, i thought i had a really nice idle, but a bit of load sends it all over the place, i think i'll block in the bins like ken said and see how it goes from there.
as in closed loop i can get the revs to rise to target, but then they drop off for a bit, then pick up and then repeat the sequence.
doesn't tend to be as bad when cold, so must be engine load related.
good advice to follow from you guys.

thanks
sean

[elaw]

Just to throw one more thing into the mix: in a lot of cars, the alternator output is marginal at idle. So when idling, as RPM goes up and down a little, the battery voltage will vary. And if your battery voltage correction in injector setup isn't exactly right, that means the amount of fuel injected will vary with the voltage, thus varying with RPM and causing all sorts of mayhem with your AFR.