Idle Control - methods?

[richyvrlimited]

Ok, I'll look there. I'd played pretty fast and loose with the tables around/under idle trying to get higher AFR's after I had CL idle in place.

It seems when it's cold, a higher dash-pot would solve some issues. Or mask them

And yet when cold i want zero dashpot, (causes an unnecessarily high idle) but need it when warm.

Dashpot is just going to have a transient affect on the idle, it adds duty for a short amount of time to make it a soft landing into the idle set point

Well aware of what dashpot does. When cold and you already have a high idle, there's absolutely no reason to need more idle valve duty.

As a double whammy, when cold the increase in idle speed from the amount of dashpot adder to the idle valve is far greater than when warm.

When cold my dashpot amount increases idle speed by 350rpm or so. When warm it increases it by 150rpm

[arran]

Isn't that what the initial value table is going to dictate?
You can select air temp (preferred) or coolant temp to specify how much initial duty to use when entering closed loop.

[muythaibxr]

Isn't that what the initial value table is going to dictate?
You can select air temp (preferred) or coolant temp to specify how much initial duty to use when entering closed loop.

You can do that and then just turn off the dashpot altogether and build it into this table.

Ken

[richyvrlimited]

But with using the initial value table you're scuppered when say you turn the lights on due to the added load etc.

Plus it's all well and good using coolant as a load value, but air temp changes the amount of idle duty required, and vis versa.

Initial values would be good, but there aren't enough 'adder' inputs available to get an OEM like tune

[AbeFM]

That's just what I'm looking for. Put in gas pedal, since it's cold, it kinda "overshoots" the idle conditions. Maybe I could raise the RPM things kick in at - or maybe it's just a vehicle speed thing. Basically, stuff which makes it behave nicely when warm are too aggressive and make it stall when cold.

[muythaibxr]

But with using the initial value table you're scuppered when say you turn the lights on due to the added load etc.

Plus it's all well and good using coolant as a load value, but air temp changes the amount of idle duty required, and vis versa.

Initial values would be good, but there aren't enough 'adder' inputs available to get an OEM like tune

You can get AC and fans at the very least as adders. Surely PID itself can handle your headlights coming on. The initial value table should be set a bit high anyway, so I doubt you'll stall, or even drop the RPM that much from turning on your headlights.

You should not use CLT as the value, that is already covered by changing the target with CLT. Use MAT and you should be all set.

Ken

[richyvrlimited]

Surely PID itself can handle your headlights coming on.

It can, but the point is you're not in PID yet.

The initial value table should be set a bit high anyway, so I doubt you'll stall, or even drop the RPM that much from turning on your headlights.

You should not use CLT as the value, that is already covered by changing the target with CLT. Use MAT and you should be all set.

Ken

'Should', 'doubt'

Why aim for okay when you can get better? I'm aiming for OEM idle quality, all of the above is just acceptable, not as good as it can be.

As stock my car has idle ups for the Power steering, HAVC fan blower, A/C, Radiator Fans, and probably others i don't even know about. MS3 has more than enough inputs/outputs to accommodate this, and they're not being used for anything else by me. Maybe others are I/o limited, but I suspect most people have quite a few spare.

[AbeFM]

I personally feel the issue is how it is entering a controlled idle, that if there were more control over that, things would be better off.

Then again, many of my cold stalls see the valve shut very far and staying there when cold, even when below target. Perhaps the function is winding up (not going to say which variable since they don't seem to really correlate in the traditional sense), either way, somehow I can see the idle about to crash and gas it in order to save it... So why can't the PID loop?
Right now my biggest stalls are on starting. I've got some logged, I'll try to thread them together into something meaningful to post if I can't get an answer myself.

But to respond to Richy:
I've long since thought that taking an analog input, wiring several circuits to ground (a 50 ohm, 100 0hm, 200 ohm, etc), putting some voltage across them each time a various input turns on, you should be able to get many bits of dynamic range (0.25 volts means AC is on, 0.75 means AC + PS), open the valve more for each step. This would let you arbitrarily change each and only require one input.
You could do that in generic outputs quite easily, I just never took the time to hook up all the various outputs. If I had, I feel my warm-up idle valve would work perfect.

[muythaibxr]

Then again, many of my cold stalls see the valve shut very far and staying there when cold, even when below target. Perhaps the function is winding up (not going to say which variable since they don't seem to really correlate in the traditional sense), either way, somehow I can see the idle about to crash and gas it in order to save it... So why can't the PID loop?

You still have not answered my last question. I need to know if you are actually in PID when you see problems. Also, if you are in PID when RPM is still dropping, you have not tuned the PID entry parameters correctly. The code is designed to enter PID with stable RPM above the target, then gradually bring idle down. No amount of PID tuning will correct for poor fuel/advance tuning either. I really need you to post logs if you want help.

Also, enough about how you don't think the PID parameters act right. There are many different types of PID loop, of which I chose the one that had the behavior I thought was most beneficial for idle control, especially before I added the initial value table. The current method does not have a windup problem like standard PID, which is largely why I chose it, and it does not require a bias table, which makes the "last known" method possible. There are slight differences in how the different types of PID work, so forget what you know about standard PID and do what works.

[muythaibxr]

.
As stock my car has idle ups for the Power steering, HAVC fan blower, A/C, Radiator Fans, and probably others i don't even know about. MS3 has more than enough inputs/outputs to accommodate this, and they're not being used for anything else by me. Maybe others are I/o limited, but I suspect most people have quite a few spare.

Maybe instead of sitting around idling you should just drive the car? (joke)

In all seriousness though, the number of idle-related settings and code is already huge. The point of having a PID loop is to account for things that many cars have no feed forward control for and reduce the number of settings needed. With as many feed forwards as you are asking for you could probably just turn off PID altogether on the idle valve and just do some PID on the ignition advance to fine-tune things a bit.

Maybe that is the right way to go though, so I will have to give that some more thought. It is certainly a lot to tune.

In any case that won't work for older cars because none of those things went through the computer back then. There weren't even electrical outputs on the systems for HVAC fan or power steering on my car. On those cars PID control is really the only option, and in those cases it avoids the alternative (very low idle or stalls).

[AbeFM]

Also, enough about how you don't think the PID parameters act right. There are many different types of PID loop, of which I chose the one that had the behavior I thought was most beneficial for idle control, especially before I added the initial value table. The current method does not have a windup problem like standard PID, which is largely why I chose it, and it does not require a bias table, which makes the "last known" method possible. There are slight differences in how the different types of PID work, so forget what you know about standard PID and do what works.

Yes, Ken, definitely trying to keep out of a philosophical argument. I'm already drinking the kool-aid, just hoping to get into functioning-idle heaven here. My experiences have been cases where I'm not getting a correction in the way I expect, likely this is what you're saying - not triggering PID for some reason.

I will definitely post logs - but since I haven't even looked myself it seems unfair to have you do my work for me. A lot going on in my non-megasquirt life. Thanks again for the help, as soon as I have something worth pinging you on, I'll post it here.

[JAM]

wow, this conversation is getting deep.... I like it! I have race cars so my applications are generally higher idle no lights and accessories so not as complicated.

Trying not to get too off topic, i have been messing with mine and studying some logs which is where i have been learning cause and effect- pretty much - guess and check - and then - a little less of a guess and check - until i was more of an educated guess and what worked. thats about as complex as my tuning idle gets, no hooking up series of circuits etc here

I do have one question that will show my ignorance:

in all of my factory efi daily drivers, there are 2 different idle rpms (when warm) - in gear and out of gear when the rpm drops. why is that? why and how does that go low and stay a little lower in most factory cars while idle in gear? why would you want it lower in gear than out of gear? i would assume that the IAC could pull it up if it was tuned to do so or? naturally it did this with an old carburetor because thats just as good as it worked and the load pulled it down.

when i tune my race car with MS3x, ( no input for shifter position) i put it in gear and IAC it naturally pulls the idle up to the idle setting where it was out of gear. I actually want it that way, it idles good about where i want it in gear.

maybe i am missing something but one of my intents was to not have the idle drop in gear and it does - so i like it

[muythaibxr]

I will definitely post logs - but since I haven't even looked myself it seems unfair to have you do my work for me. A lot going on in my non-megasquirt life. Thanks again for the help, as soon as I have something worth pinging you on, I'll post it here.

Fair enough and much appreciated.

Ken

[muythaibxr]

in all of my factory efi daily drivers, there are 2 different idle rpms (when warm) - in gear and out of gear when the rpm drops. why is that? why and how does that go low and stay a little lower in most factory cars while idle in gear? why would you want it lower in gear than out of gear? i would assume that the IAC could pull it up if it was tuned to do so or? naturally it did this with an old carburetor because thats just as good as it worked and the load pulled it down.

My factory EFI car works opposite to that. It keeps the RPM lower if I'm sitting in neutral with the clutch out, but as soon as I push in the clutch and put it in 1st, it idles up some. I always assumed this was to make the engine respond better to my right foot or something when I do finally push the pedal, but I don't know the "real" reason.

Ken

[AbeFM]

My Suzuki TL1000R does the same, it even runs different timing maps in neutral just to keep cooler.

[JAM]

in all of my factory efi daily drivers, there are 2 different idle rpms (when warm) - in gear and out of gear when the rpm drops. why is that? why and how does that go low and stay a little lower in most factory cars while idle in gear? why would you want it lower in gear than out of gear? i would assume that the IAC could pull it up if it was tuned to do so or? naturally it did this with an old carburetor because thats just as good as it worked and the load pulled it down.

My factory EFI car works opposite to that. It keeps the RPM lower if I'm sitting in neutral with the clutch out, but as soon as I push in the clutch and put it in 1st, it idles up some. I always assumed this was to make the engine respond better to my right foot or something when I do finally push the pedal, but I don't know the "real" reason.

Ken

My Suzuki TL1000R does the same, it even runs different timing maps in neutral just to keep cooler.

ahhh... the mysteries of factory efi. I have a few friends who mess with factory ecus and HP tuners. apparently there are many things they dont have access to change. One of the many reasons I like the MS platform, there seems to be a known reason for everything and a way to adjust almost anything. And many times when i have wanted added features it has happened. hard to beat in my book.

[AbeFM]

So.... It took me long enough, but I finally have some logs of my car starting and inexplicably dying on a warm start.

You can see towards the end you start getting what looks like I term analog control of the duty cycle, but you also see the valve not opening as much as it could early on when it's stalling.

As you can see, the idle target seems to be all over the place, from pretty low to pretty high. It's as likely a settings snafu as anything, perhaps you can figure it out and tell me what I'm doing wrong.