When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
I have the car running pretty normal but I could never find the closed valve % before the engine stalled. So the %range I have it is between 35 and 80%
At around 33% I'm already at like 700rpm, if I lower the % the engine keeps slowing down until it stalls.
Does this mean that the throttle body plate position is too closed / the idle screw is too closed?
I think that my throttle plate is all the way closed, if not close to it. When I removed the throttle cable to install the VTPS I had that initial resistance to open the throttle as if it was all the way closed.
I wouldn't care about this if I wasn't tuning out the idle hunting I have. And I think this would also keep the engine speed within operating range no matter what the idle valve is doing, right?
The stock ECU controlled the idle... OK, if I kept the throttle sliiightly open for a few seconds and then i let it go, the engine would stall. It's like the ISCV was always holding the idle up, there seems to be no "mechanical" idle if that makes sense.
So, should I play with the throttle body? Would opening the throttle a little bit also help with the AE since it would already be slightly cracked open and the air rush would be smoother?
Having the mechanical stop set so you can idle with the IAC closed is really working around the issue of tuning the IAC PID's correctly. If done right you'll have a rock solid idle no matter what. Once I have the idle setup correctly then might set the throttle stop so it was a backup but it's really a moot point
Having the mechanical stop set so you can idle with the IAC closed is really working around the issue of tuning the IAC PID's correctly. If done right you'll have a rock solid idle no matter what. Once I have the idle setup correctly then might set the throttle stop so it was a backup but it's really a moot point
I respectfully disagree. In fact, it's the opposite. The IAC is used to help stabilize idle under varying loads and conditions, but is best viewed as "added value" to an otherwise stably idling car. The consensus is that initial idle tuning should be done in Open Loop (warmup) mode, which disables the IAC for all but idle warmup, meaning the car should idle with the valve fully closed. If your car cannot ilde without the IAC being open, then you have a mechanical or tune issue.
As you stated Nicolas, the car would not idle with the factory ECU. Assuming that the factory ECU was working correctly, there is a mechanical issue that needs to be sorted first, otherwise you'll always being chasing your tail and applying tune bandaids to a mechanical issue. The Miata can and should idle without the IAC being open. Then you can start tuning the IAC for warmup, load correction, variable conditions...etc.
Also disagree. Set the base idle just under where you want it with the IAC unplugged... set for warmup pwm then raise your idle DC to hit your target for the coolant temperature you're at... and use timing to reach your target... set up your timing table to stabilize idle by using rows just above and below your target. You can have a car that idles Pretty damn good on warmup PWM, which should have your IAC at a minimal DC when warm. You can have a 90% awesome idle using this method. Once you can't seem to improve it at all, THEN use closed loop idle to start fine tuning things.
... this is how I got my street car with 272 degree cams to idle rock solid, even on a hot summer day with A/C. Hell I can even slowly release the clutch from a stop, and watch the IDLE DC increase and maintain slow speed in traffic without touching the gas. At idle the IAC should not do anything unless load increases from electrical system, power steering, A/C, etc.
unplug the IAC and open up the bypass screw until you idle at 800 or so, and that's a start.
It's a strategy right? Your methods work as well but I can't help but wonder why both of you look at air that comes from the IAC as a bad thing. The quality of the idle has zero to do with the IAC, the proper fuel and timing will dictate how smooth the idle is. The IAC just sets idle RPM, and can do so under changing conditions.
To keep it simple let's talk about the different situations the IAC would need to compensate for. Really the only condition it wouldn't need to be open is a fully warmed up engine with no A/C, low electrical load and no power steering load. So with your strategy you would set your idle screw so you have zero IAC duty cycle in this condition and the IAC would kick in for any of the above conditions and warm up only. What I was suggesting was to simply turn the idle screw to just below your target idle so you have 5 to 10% duty cycle to hit your target. There is zero reasons why the car wouldn't idle identically like this vs. with zero duty cycle. My IAC doesn't hunt, the idle doesn't surge. The benefit is that at lower elevations where the atmospheric pressure is higher or in colder temperatures where the air is more dense I'll have the IAC capacity to lower the idle.
Again this is just a strategy, tune your car anyway you want.
Bronson, reread your first post. You did not suggest that Nicolas set the idle screw, as you did in your second post. The others were right to correct. What you said next is much better advice.
Nicholas,
1) Throttle should be just closed and slight slack in the cable. In other words, the stop screw should set the position.
2) AE has been proven by Ford and Toyota to be a wall wetting issue. Even though carburetors fixed it with Accel Pumps, the assumptions of fuel lagging air in the ventui was false.
3) Make sure timing is correct and that you are working with advance at the nominal value when you set steady state idle air.
4) With AIC closed, set idle screw to give abou 50 RPM less than your minimum target.
5) Now proceed with finding minimum IAC, etc.
Pro Tip: idle is controlled mostly by timing followed by IAC and correct AFR
Last edited by DNMakinson; 04-04-2017 at 07:47 AM.
Reason: Spelling correction
Ehh not really I just explained it better. In my mind if I have 5 to 10% duty cycle on the IAC it's no different than 40 to 50%, the result is the same the ECM is in control of the idle speed via the IAC valve. Tune the PID loops so it doesn't have any bad behaviors and you'll learn to love the fact your motor hits your idle speed no matter the conditions.
I wouldn't mind some more info on this. I have no Idle control valve, no air con and massive cams. The current smooth idle is 2000rpm and if I drop below that the engine goes into extreme lumps. Violent enough to break things or vibrate bolts loose. Due to my gearing it also makes idling through the pits extremely hard so it would be nice to try and get it a bit smoother at a lower rpm.
To get the engine to idle at 1300rpm or so do I advance the timing in the cells below and retard above to try and ramp it to the target idle? Is that the general ideal?
Ehh not really I just explained it better. In my mind if I have 5 to 10% duty cycle on the IAC it's no different than 40 to 50%, the result is the same the ECM is in control of the idle speed via the IAC valve. Tune the PID loops so it doesn't have any bad behaviors and you'll learn to love the fact your motor hits your idle speed no matter the conditions.
Yes, he could simply set his minimum DC% to that which results in, say 750RPM without ever touching the screw. However, best practice would allow car to run even if the AIC Valve were to fail.
OK so I think I was planning towards the right direction then. I have not done it yet because my apartment complex has a rule against working in your car and I'm surrounded by old ladies that have nothing better to do than stick their noses in other people's business.
The stock ECU would make the car idle, but it was very clear to me that it was being sustained by the ISCV, because whenever I disconnected the battery, it took the ECU a few tries to "learn" the idle, and sometimes if the ECU did not react quick enough, the idle would drop so to like 300rpm or stall and then the ECU would pick it up all of the sudden overshooting, pretty much what I'm dealing with right now. So if the ISCV doesn't "catch" the idle going down, it just keeps going down to like 300rpm.
I managed to get the idle to run OK without the AC, by kicking up the dashpot adder to like 5, but whenever I have the AC on, if I don't roll in gear to the lower RPM side (basically holding the RPMs until the valve starts to open and I feel that the car rolls more freely and I know the valve has moved from resting position), then the ISCV does not react quick enough to hold the idle, it's not a smooth slide down towards idle RPM that the ECU has time to react. With the AC, the dashpot adder I have now is not enough and if I'm cruising at 3500rpm and press the clutch releasing the gas, the rpms just drop until the MS disconnects the compressor because the RPM are too low and then it tries to get back up, the PWM% starts going way up to like 50 while the engine is still struggling to get up, when it finally hits some decent RPM like 600 then it reacts to the PWM addition and overshoots to 2K rpm.
So, should I set it to warmup or disable the idle control 100% and get it to idle at like 780ish rpm only with the mechanical side and then enable the idle so it can play around that? I'm planning on idling at either 800 or 850, the stock ECU would hold it there and I could barely feel the engine running. It was smooth, like a round wave instead of a square wave, right now it's more like a sawblade wave hahah.
I'll check how to move the throttle stop, my cable has some slack, it's definetely closing all the way, perhaps I'll open it a tiny tiny bit.
I'm used to play with idle from having a Fiat with mild to aggresive cams running thru a 36-36 Solex that was a nightmare, so hopefully I can get this one right. Other than this detail my drivability has improved and gas mileage went from 25mpg with stock ECU to 31mpg with the last tune, so all that is good.
I wouldn't mind some more info on this. I have no Idle control valve, no air con and massive cams. The current smooth idle is 2000rpm and if I drop below that the engine goes into extreme lumps. Violent enough to break things or vibrate bolts loose. Due to my gearing it also makes idling through the pits extremely hard so it would be nice to try and get it a bit smoother at a lower rpm.
To get the engine to idle at 1300rpm or so do I advance the timing in the cells below and retard above to try and ramp it to the target idle? Is that the general ideal?
yes.. and don't be scared to run big advance numbers. I run 26-30 degrees advance at idle, and switch to a "driving" spark map at 1% throttle, otherwise I'd be driving around in my idle zones making off idle extremely jerky... do to the "feedback loop" I've tuned into the idle timing my advance gauge is bouncing all over at idle, but RPMs are smoooth
Thank you to everyone that's contributed to this thread. I've always had a tough time tuning idle and there seems like a ton of great info here.
I've seen a lot of references to "get your timing right". Would anyone mind sharing the idle portion of their timing map? I don't have a clue where to start here, and it'd be a nice to see a couple "correct" examples so I'll at least know what I'm working towards.
I also found this extremely helpful. Forget anyone else numbers.You need to understand the numbers yourself. Mega Miata: Closed Loop Idle REV 2.2
there is plenty of other great sources around.
sometimes its just that someone says something ever so slightly different. and the lights switch on. and there is someone home.
you can also find the old manuals online as well. Read them. there is a slightly different style of writing which I found very useful.
It took me ages to understand the numbers in the Closed loop initial values table or what the numbers in that table represent. Not saying you'll have the same issues. But you get my drift.
Thank you to everyone that's contributed to this thread. I've always had a tough time tuning idle and there seems like a ton of great info here.
I've seen a lot of references to "get your timing right". Would anyone mind sharing the idle portion of their timing map? I don't have a clue where to start here, and it'd be a nice to see a couple "correct" examples so I'll at least know what I'm working towards.
I started turning down the PWM% with the test thing, and then I brought back the RPMs with the idle bypass screo to around 800 to be on the safe side, I was off by like half a turn. The plan is to have the PWM to bring it to 850ish. The closed position is now 32 instead of 37, at 33 it already moves the idle up to 850, I corrected the initial closed loop values as well to reflect what the engine does based on the PWM%.
I'm not worried about high idle right now, I'm worried about losing steering power and braking power in the middle of a U turn in a busy intersection, so I rather have high idle than stalling.
All is good with the A/C off, I can probably drop the dashpot a few, but other than that is good.
Problem is with the A/C, not when it comes on, that works fine, but when it cycles to off and I'm idling, it will try to stall the car when the compressor DISENGAGES, it's weird, when the load is removed, the ECU kills the engine somehow, the RPMs drop A LOT and then they stabilize if the engine doesn't die.
I've tried to give it more fuel under idle load, over idle load, and more timing when **** drops, I tried to see where in the maps the ECU would walk through and correct those, it doesn't stall now, but it's patched, it's not a smooth transition, it will still try to die, I'm just holding it up with fuel and timing.
I've attached a log in case anyone wants to dig into it, I'm not super familiar with megalog viewer yet so I wouldn't be able to provide you with a good view.
And I've attached my full msq, it's not perfect, there is uneven **** in the maps because I'm doing autotune with my eyes and hand, parking it on values and using < > to correct, so I get one cell right and then i smooth out. My wideband readings are also .5 too lean in the MS compared to the gauge so I don't trust the autotune yet.
Here is a picture to show what's going on:
ADC6 is the AC Switch (either on the dash or the thermo/pressure/whatever switch, that one controls the same input in the MS)
ADC7 is the output to the compressor.
This is a full cycle, first goes off, does the weird ****, rpms go down to 360 at some point, recovers, then the AC comes back on, that part works fine.
Thanks to anyone that spends more than 0 seconds on my problems, I hope I can fix them and post back so we have the info for everyone.
Thanks!
EDIT: BTW, I still have the flapper, sometimes I wonder if that ***** up everything because it blocks the intake when flow is too little and then it kills the engine, just wondering...
Just getting back to basics, you want to adjust both timing and fuel for maximum vaccum. Stock the motor would idle with 30 to 35kpa. With timing and fuel optimized I'm seeing closer to 25 jpa. That is with timing in the 25deg range and AFR's in the 13.5 range.
I think your "Idle up duty adder" is waaay too high at 14%. So when you turn on the AC you get idle to jump up a lot. Not too bad. But then CL turns the IAC valve back down. And now, when you turn of AC, we take 14% away from the IAC value and the engine stalls. Lower that to maybe 3% or so and see if that fixes it.
I think your "Idle up duty adder" is waaay too high at 14%. So when you turn on the AC you get idle to jump up a lot. Not too bad. But then CL turns the IAC valve back down. And now, when you turn of AC, we take 14% away from the IAC value and the engine stalls. Lower that to maybe 3% or so and see if that fixes it.
I agree, also increase the delay of when the compressor kicks on so the RPM's have a chance to increase before the load hits.
I think your "Idle up duty adder" is waaay too high at 14%. So when you turn on the AC you get idle to jump up a lot. Not too bad. But then CL turns the IAC valve back down. And now, when you turn of AC, we take 14% away from the IAC value and the engine stalls. Lower that to maybe 3% or so and see if that fixes it.
12-14% is a pretty common value for NAs... but you should use what value your setup needs, if you need 15% more idle duty to maintain the idle with the extra load, then when you turn it off, the rpms should settle back to normal...because like math...
But this should be tuned to the actual increase in idle duty you require to idle at a slightly higher rpm with the extra load.
There needs not be a higher jump in RPM, it should be seamless:
The consensus is that initial idle tuning should be done in Open Loop (warmup) mode, which disables the IAC for all but idle warmup, meaning the car should idle with the valve fully closed. If your car cannot ilde without the IAC being open, then you have a mechanical or tune issue.
I don't subscribe to tuning the warmup table, because it's sorta wasted effort. I watch where my CL idle lands me, and start populating the initial duties table based on those values. But you could take the warmup table to do the same. I also have good basemaps with all this stuff worked out... HINT.
but the rest is true. I always aimed on a typical summer day, for the idle to maintain 600rpm at the lowest if the idle valve was completely closed. This would allow wiggle room for when it got cold.
here's another question, have these been tuned:
initial duties table
voltage compenstation table
adaptive timing turned on and tuned
min/max idle pwm% values and rpms
PID sensitivity
VE Idle table tuned for A/C load rows
initial duties table
voltage compenstation table
adaptive timing turned on and tuned
min/max idle pwm% values and rpms
PID sensitivity
VE Idle table tuned for A/C load rows
Initial duty table: Yes. I used the PWM test slowly to park the RPMs at the values that show up on the table and then placed the number from the PWM test into the table.
Adaptive timing is turned on, it's somewhat light though. For normal idle situations, without counting the AC, it works fine.
Min Max idle PWM yes, min is 32 and max is 71 altough I think I could let it go to 100%
PID sensitivity I'm using the basic for now, I have it where it doesn't hunt.
VE idle table I cranked up the values in the idle area for the AC. No AC idles at 57ve, AC idles at 65, trying to keep it somewhat rich to help, it's around 13AFR with those values. I've also corrected the values that are used when all this ON/OFF **** happens to keep them away from lean.
My AC duty adder is around that because if I put it lower the idle jumps pretty slowly to those extra 150-200 rpm I've set it to. With these values it kicks higher, but anyway, if I check the logs, when I turn the A/C off, it goes back to the values to hold 850rpm (well, 823 in that msq because I didn't notice that difference). So it drops back down to ALMOST closed PWM. I think that if the adder was smaller it would jump there anyway no?