RotorDream

Ok, I know alot of people have had trouble with the idle learning process, but what seems to be happening is the exact opposite. A few weeks ago I had my AC fixed by the dealer, and I needed to relearn the idle since it never learned it with a working AC. The process went great, 750 idle, 800 under load, and 900 under AC. Inj duty and Ing timing were steady. So I shut the car off and went to bed. I woke up the next morning, and as soon as I started it it jumped to 1900 for about a sec then fell to 1300 then dropped back down to 750 and settled. Seemed weird, but ok. after about 10 minutes of driving, it would idle at 800-820 with no loads and wouldn't drop below that with electric load it will be around 920-930 and with AC it will stay around 1100. I figured it didn't learn right, and I'm running the base mod map, until I can get to Garland for a tune, so I reset and reran the learn, once again it went great with the same 750-800-900 idles, drove the car around again and 10 minutes later same symptoms. Reset it ant tried again for about 6 hours trying to get it right, but still same symptoms.
I've tried
checked TPS voltage and it is correct
adjusted air bleed screw below throttle body from fully in to 1 turn out
adjusted settings under Rev/Idle on PFC
adjusted the dashpot, top idle screw, wax pellet, and every other screw i could find
Ive tried starting the idle learn process with the engine hot and cold
Compressions good
spark plugs and wires are brand new <200mi
Checked ISC with DMM and seems fine

This is starting to **** me off pretty bad, never had any problems with her until now. It just refuses to remember what it learned

I think my PFC has the memory of a goldfish!

cewrx7r1

iTrader: (2)

How much did the water temp vary from the beginning of the leaning process to when it starts screwing up? There is a WATER TEMP CORRECTION table that richens up the AFRs for cold starting and after 60C water temp it turns off if it is setup correctly.

Normally it is done on a cold engine with no elect and no AC.
When fully warmed up and the fans come on, turn on the headlights.
AC is done last.

Make sure that the cold start hot wax unit is engaging when the engine is cold and dis-engages when the engine heats up.
If not it with screw up TB air flow.

RotorDream

just pulled the ISC off and checked it again, it is within specs in the FSM
FSM says 10.7 to 12.3, and mine right now is at exactly 12.3 it is within specs, but does anybody know what being at the higher end of the specs would do?

RotorDream

Pulled the TB and UIM off tonight to see if I could find any visible reason for the symptoms, no luck, although I did find out my AWS solenoid is reading about 1.5 Mohms (yea thats 1,500,000 ohms) so im pretty sure its not any good. makes a good block off plate though!

cewrx7r1

iTrader: (2)

You still have not mentioned the hot wax rod unit that operates the cold fast idle cam.

RotorDream

Damn I really need to get an inspection mirron checking that thing sucked! but from what I could see it was in between the 2nd and 3rd line position B and C in the FSM, which is where it should be at ~13C and this is at 70C on the PFC, the motor wasnt running but its about over 35C (96F) outside. Ill adjust it and see if it helps. Thanks for the info.

arghx

iTrader: (3)

You can physically rotate the fast idle cam up and zip tie it away, or at least on the 2nd gen TB's you can and the FD TB is very similar in design. That will verify that the fast idle cam isn't stuck/misadjusted. You can see the VTA1 and VTA2 tps values change as you rotate the cam with your finger.

Try this:

loosen the dashpot so the plunger does not contact the throttle at all. Since your idle speeds are 750-800-900 , set your fuel cut speeds to 850-900-1000 . Those would be the lowest allowable fuel cut settings for those idle speeds (they must be 100rpm apart). Your ISC should kick in a lot later on decel with less likelihood of the idle sticking. In fact, it might start stumbling a bit when you decelerate, but you can then incrementally increase the fuel cut speeds to help correct that.

Also, make sure O2 feedback is still off. And I like to keep idle IG control on with the ISC still installed, I think it does help with idle stability based on my logging.

RotorDream

Any negative side effects to ziptieing the fast idle cam/setting the screw so it is disengaged when cold?
The dashpot is already set to where it is disengaged after about 3mm of movement

arghx

iTrader: (3)

here's what you can do to see if the fast idle cam is affecting warm idle speed. With the engine cold, bring your Commander into the engine bay. Go to etc.-->sensor/sw check. note the VTA1 and VTA2 values. Now rotate the fast idle cam with your finger. The VTA values should drop, especially VTA2 which is your narrow range sensor that is designed to detect throttle angle at idle. Return the fast idle cam to its normal position.

Now warm the engine up (get it to 85 C or so) and cut it off (you don't have to cut it off but it may make it easier to work back there). Look at your VTA values--they should be lower than before. Perform the same cam rotation test. The VTA values should NOT change as you rotate the cam away from its resting position on the now-extended rod. If they do change, your fast idle cam has not fully rotated away, most likely due to adjustment of the cam adjust screw:



adjust the fast idle cam screw so that, with the engine completely warm, VTA voltages do not change as you physically rotate the cam. Does that make sense? You are just trying to make sure the fast idle system isn't making the throttle plates stick open once the car is at operating temp. I'll admit that my original suggestion of ziptying it really isn't necessary and isn't a very precise diagnostic procedure. The way it works is that the hot coolant makes the rod extend, which rotates the cam. That screw mostly sets the angle of the cam when the rod is fully extended.

I'm not sure how hard it is to get to that screw, I've only adjusted a fast idle cam screw on a 2nd gen turbo throttlebody which has the fast idle cam in a different orientation.

RotorDream

just got done adjusting it right before you posted this. haha. ran through the idle learn procedure again, and it seems rock solid from about 7 minutes into each part (idle, EL, AC). I started the learn with a cold engine (about 30C) by the end of the first 10 min it was at about 60C the fast idle cam was disengaged, and ran through the rest of the procedure. Heres what the PFC was telling me during the process

No Load Idle
RPM 720 +/-10
Inj duty ~4%
timing ~ -5 -15
boost -400mmHg +/-20

EL idle
RPM 800 +/-10
Inj duty ~4.3% ? (not quite sure but I believe it was around there)
timing ~ +5 -5
boost -400mmHg +/-50

AC idle
RPM 900 +/-15
Inj duty ~5.0%
timing ~ +15 +5
boost -380mmHg +/-25

Peak hold during the process was 1009rpm during the change from EL to AC

Seems like everything is pretty much in check, but let me know if anything seems out of whack. I haven't gotten to check it driving, but I dropped the F/C settings down to 870-950-1050 and Ill adjust them up or down from there. Fans came on at around 60F but that's what they've always done, probably a variation of the fan mod, and they stay on until about 20-30 minutes after the car is shut down.

arghx

iTrader: (3)

keep us updated

RotorDream

Just got to work, and the drive pretty much pissed me off.

So I'm pretty sure the F/C settings were too low since it stalled backing out of the driveway, so I raised them to 1020-1100-1200 and didnt stall backing up. I got on the road, and as it got warmer, the idle would start to bounce alittle more. I got to a stop light and with the AC on, it would drop down to 600 and the catch itself and slowely rise to 920 where its set, and then repeat the process about 10 seconds later.

RotorDream

Ok. I'm on my lunch break, and here's what I noticed on the drive home.

Coolant Temp was about 65 when I got out there and started it up
I started it up and it goes straight to 1300 and then works its way down to 900 with the AC on. hangs at 900 for about 20 seconds, and then drops down to around 650, and wobbles around between 600-700. shut the AC off and it wobbles between 630-670. after the coolant hits around 80 it rises a bit and settles around 760 with the AC on (Its like 95F outside with all black interior, yea I need it). After it settles around 760 I start driving, and it runs fine when there's a load on it, boosts fine, runs smooth, no surging/stumbling. I get to a stop sign, and shift into neutral and the idle drops to around 500, and almost stalls out with the clutch depressed. I didn't throttle to see if it would recover on its own, and it didn't. It didn't stall and just kinda stayed right above stalling till I had a chance to get through the intersection. So I come to another stop sign, and this time it idles around 1050 No change in the way I shifted or slowed down. both times 35-0 with mild braking. I get home, park and let it sit for a minute. Coolant temps are now 92, and its idling decently but higher than the settings on the PFC. between 950-1050 turn the AC off and it drops to around 800

It seems to idle way too high sometimes, and way too low others. Except every time I run the learn procedure it goes perfect!!! WTF!!!!!!!!!!!

arghx

iTrader: (3)

This is when it gets tough and you just have to try different combinations of stuff. Try adding bypass air by loosening the air adjust screw slightly. Give it half a turn. If the idle sticks on decel, reduce fuel cut speeds. You need to adjust those fuel cut speeds in 10 or 20 rpm increments some times. I'd skip the idle learn for now.

it should start a little high, especially with water temps in that range. Sometimes the fast idle cam won't come down until 80 C.

RotorDream

yea I wasn't worried about the 1300 startup, as it only lasts about 5-6 seconds.

arghx

iTrader: (3)

this is why everyone has gone to drive-by-wire. you should see how amazingly simple the Rx-8 idle system is. It's just a throttle plate controlled by a computer. No idle air control valves, fast idle cams, cables, double throttle diaphragms etc to adjust or to fail.

RotorDream

Ok. Im back at work now, and the drive went a little better than last time, I adjusted the F/C levels while I was driving, and I think Im getting closer, but with alittle side effect. It seems like the lower I set the F/C rpm the less it hangs at 1300 on decel, but the trade off is it sometimes doesnt hang and drops down to around 480 before it recovers and comes back up to normal. and then sometimes it will just pop up around 1100 and hang out there for a few seconds then drop down to normal.

I do feel like I'm getting closer to what I would consider normal though haha

arghx

iTrader: (3)

Keep adjusting the fuel cut settings, you can at least mitigate this or otherwise make it tolerable. The PFC is very sensitive to those settings as I've said before. Adjust in 10-20 rpm increments at a time once you start to get close to where it needs to be. Also you may need to tune this setting completely independently for each mode (no load, electrical load, A/C load). In the end the idle speed may be slightly higher or slightly lower than you would like immediately following the lifting off the throttle. But you can still do your best to get it to settle to the idle speed you want as smoothly as you can manage.

Also a lot of cars will likely not drop to the target idle speed on deceleration unless the vehicle has come to a complete stop. The high idle on deceleration is intentional, to keep the car from stalling. But the specific logic used in this behavior is not controllable in the PFC, only in other computers like the AEM EMS. If it REALLY bothers you, you may have some luck depinning/cutting the vehicle speed sensor wire. I say that because I have noticed that the high idle behavior while coasting does not seem to occur as much without that wire hooked up. But that's not something I would really recommend at this point as it is only based on my own limited observations.

Are you talking about the idle hanging until the car comes to a stop? Again the best thing to do is keep playing with the fuel cut settings. Now if you are talking about the idle shooting up/sticking during small throttle movements (like parking manuvers), sometimes the dashpot can contribute to this. Loosen it as far as it will go so that the plunger does not make contact with the throttle.