Which parameter to adjust idle fuel mixture?

[1FastCSX289]

My CSX (2.4 conversion) has a very erratic idle.......the car starts very easily and is quite rich but idles steadily with a bit of chug typical of a rich mixture. As soon as the computer starts looking at O2 feedback the idle is all over the place and the computer is obviously trying to compensate. The problem is I dont really know which way its trying to compensate. Logic tells me that if its really rich prior to o2 feedback then its probably trying to lean it out, but I often see numbers on the wideband that indicate its running very lean. What is the easiest parameter in dcal to adjust to play with that idle mixture for a noticeable change either way? IF I can find a parameter that seems to change the idle mixture substantially I can just play with it and figure it out.....

[ShelGame]

I'd start by turning OFF the spark scatter idle control, forcing the AIS to control idle.

The primary means the Chrysler ECU uses to control idle is spark and fuel changes. So, it can be tough to get the idle fuel right while the spark scatter is running. I think all you really need to do is set IDLTMP and SSTEMP (in the 'Spark - Scatter Control' group) to FF to disable spark scatter. Though, you may want to copy over the complete 'Spark - Scatter Control' group from the 2.2 MP template. I'd start with IDLTMP - it controls the spark scatter fuel. Then try the SSTEMP to dis-able spark scatter completely.

The other effect is, this will likely keep your revs higher during shifts because the AIS will be wide open. So, you may also have to play around with the rolling AIS tables to bring the RPM down during shifts (if you need/want to).

[1FastCSX289]

The other effect is, this will likely keep your revs higher during shifts because the AIS will be wide open. So, you may also have to play around with the rolling AIS tables to bring the RPM down during shifts (if you need/want to).

Im already having problems with the RPMs being too high between shifts and the idle doesnt settle until I am at a dead stop. Could you have turned off the spark scatter originally when you wrote the base cal? If not, then why would I be experiencing the increase revs?

Also, once the spark scatter is disabled, then what is the best way to adjust idle fuel? Pumping efficiency?

EDIT: Is the spark scatter adjustable in Dcal? I dont have MP tuner because my dang laptop is old school and running Win98.

[ShelGame]

Im already having problems with the RPMs being too high between shifts and the idle doesnt settle until I am at a dead stop. Could you have turned off the spark scatter originally when you wrote the base cal? If not, then why would I be experiencing the increase revs?

Also, once the spark scatter is disabled, then what is the best way to adjust idle fuel? Pumping efficiency?

EDIT: Is the spark scatter adjustable in Dcal? I dont have MP tuner because my dang laptop is old school and running Win98.

I don't think I dis-abled it.

You should be able to do it in D-Cal. Just look for those same names I posted above. There won't be groups, obviously.

The increase revs could be because the spark scatter simply isn't tuned for your engine, yet.

Once you get it shut off, I'd start with the pump eff to reduce the idle fuel.

[1FastCSX289]

Thanks Rob......I did look and didnt see those names, but I will go back and have another look to be sure. If not ill just load MPtuner on my desktop computer and tune from there.

[1FastCSX289]

Nope....those items are not on the list on dcal. I think ill have to start messing with MPtuner.

[turbovanman²]

Interesting, my idle is a bit lumpy due to my cams, so I'll try playing with the spark scatter, even though my idle AF is 14.5:1.

[ShelGame]

Nope....those items are not on the list on dcal. I think ill have to start messing with MPtuner.

You know what - I didn't build your cal with a template. I used a stock 2.5 SBEC cal to start with. T-SBEC wasn't ready for prime time when I originally built that cal for you.

Look for address 0x8199. If it doesn't exist, define a new one (let me know if you need help with this in D-Cal). This is IDLTMP. SSTEMP is at 0x8157. They're both 8-bit constants (scaling factors don't matter really). Set both of them to FF to dis-able. Maybe remember what the stock settings were so you can re-enable later.

[1FastCSX289]

You know what - I didn't build your cal with a template. I used a stock 2.5 SBEC cal to start with. T-SBEC wasn't ready for prime time when I originally built that cal for you.

Look for address 0x8199. If it doesn't exist, define a new one (let me know if you need help with this in D-Cal). This is IDLTMP. SSTEMP is at 0x8157. They're both 8-bit constants (scaling factors don't matter really). Set both of them to FF to dis-able. Maybe remember what the stock settings were so you can re-enable later.

You are speaking a little greek to me here (im a big newb with tuning), but I think I got the idea.......I would have to open the bin file in the willem burner software for this, right?

[ShelGame]

You are speaking a little greek to me here (im a big newb with tuning), but I think I got the idea.......I would have to open the bin file in the willem burner software for this, right?

Nope, just open D-Cal, load up your cal file, and press the 'new' button. In the 'table name' and 'chrysler name' field just put the names I used above; in the 'hex address' field, put the addresses from above. Make sure the checkbox for 'constant, unsigned 8-bit' is checked. Do that for both of the above locations and then set them to FF or 256 using the slider...

[1FastCSX289]

I just did the IDLTMP and I can only slide the slider as high as 255.....that reads: FFFFH. Is that right? Or am I missing something? Forgive my ignorance on this......

Edit: Just did both of them and slid them up to 255 (FFH) but couldnt get 256. SSTEMP was already set at 255.

[ShelGame]

Yeah, that's right. 255, 256, whatever it takes...

[Aries_Turbo]

do you know the constants to do this with the LM?

Brian

[1FastCSX289]

Yeah, that's right. 255, 256, whatever it takes...

Thanks for your help. So if I reset these to 255, I will eliminate the ability of the computer to control idle with the timing and therefore eliminate one variable. Will the computer still try and compensate by adding and subtracting fuel (making it a bit difficult to see exactly where the idle mix is at) or should I be able to just go in and dial it in pretty easily with the pumping efficiency table now?

[ShelGame]

The SSTEMP looks like it was already set to FF. That will dis-able the spark scatter. But, the IDLTMP is the one that adds fuel to control the idle speed. Now that that one is off, you should be able to adjust the pumpeff table effectively. After that, maybe try setting the SSTEMP to a value. If you set the sacle on that table (double click on the table name to bring up the properties) to -200 to 260 deg F, you can set it to ~100F and see if that helps bring down the RPM between shifts (assuming you want it to).

[1FastCSX289]

The SSTEMP looks like it was already set to FF. That will dis-able the spark scatter. But, the IDLTMP is the one that adds fuel to control the idle speed. Now that that one is off, you should be able to adjust the pumpeff table effectively. After that, maybe try setting the SSTEMP to a value. If you set the sacle on that table (double click on the table name to bring up the properties) to -200 to 260 deg F, you can set it to ~100F and see if that helps bring down the RPM between shifts (assuming you want it to).

AWESOME! Thanks I will try that soon and update the thread.

[GLHNSLHT2]

when I put my ported head on the car it was great till the adaptives started kicking in, then the idle went really lean as it was was trying to set the fueling. I plugged in the scanner and watched it trying to add fuel. As soon as it got up to around 20%+ more fuel than in the curve it would start to get irratic. I just bumped up the curve in the fuelnothrottle table around my idle vacuum till the and watched the adaptives on the scanner. I can't remember off hand how much idle fuel I added. It was was between 25 and 50%. Anyway that solved it.

[1FastCSX289]

when I put my ported head on the car it was great till the adaptives started kicking in, then the idle went really lean as it was was trying to set the fueling. I plugged in the scanner and watched it trying to add fuel. As soon as it got up to around 20%+ more fuel than in the curve it would start to get irratic. I just bumped up the curve in the fuelnothrottle table around my idle vacuum till the and watched the adaptives on the scanner. I can't remember off hand how much idle fuel I added. It was was between 25 and 50%. Anyway that solved it.

This is exactly whats happening. The idle starts out nice and smooth running at about 12.5:1 or so and then it starts to lean out......and lean out....and lean out until it gets so lean it wants to stall, then the idle races and settles down. Then the cycle repeats itself. Im starting to get frustrated. I tried to disable the spark scatter as per Rob's instructions and its still doing the same thing. My question would be why would the idle be going lean if the computer is adding fuel? Are you saying the computer is trying to add enough fuel but it needs more than the parameter allows?

[1FastCSX289]

I dont have a "fuelnothrottle" parameter that I can see.......can I add this similarly to the SSTEMP above?

[Aries_Turbo]

FuelBaselineFromMap is the same thing as FuelNoThrottle

[GLHNSLHT2]

This is exactly whats happening. The idle starts out nice and smooth running at about 12.5:1 or so and then it starts to lean out......and lean out....and lean out until it gets so lean it wants to stall, then the idle races and settles down. Then the cycle repeats itself. Im starting to get frustrated. I tried to disable the spark scatter as per Rob's instructions and its still doing the same thing. My question would be why would the idle be going lean if the computer is adding fuel? Are you saying the computer is trying to add enough fuel but it needs more than the parameter allows?

Yes the ECU can only adapt 24.7% either way. How long do you have a nice rich idle? 10-60 seconds? While you're still on the cold start table...?

But yes that's what I'm saying, you need more fuel than the computer is giving you. I added more fuel at idle than anywhere else in the calibration when going to the ported head setup.

[ShelGame]

Sounds like the idle adaptives are out of whack (or can't keep up). You need to dis-able them as well to see how lean you really are.

I think if you set 'MNRPO2_MinimumRPMOverIdleForO2Feedback' to 0, it should completely dis-able to O2 feedback (forcing the system to go open-loop). I think this includes skipping the adaptives. That value should be at 0x8a25 and it will be set to 0xBC (188 decimal).

[1FastCSX289]

FuelBaselineFromMap is the same thing as FuelNoThrottle

Gotcha. Thanks.


Yes the ECU can only adapt 24.7% either way. How long do you have a nice rich idle? 10-60 seconds? While you're still on the cold start table...?

But yes that's what I'm saying, you need more fuel than the computer is giving you. I added more fuel at idle than anywhere else in the calibration when going to the ported head setup.

Yes, it idles nice for the 20 seconds it takes for my innovate wideband to warm up and then about another 10 seconds after that......then it goes to crap.


Sounds like the idle adaptives are out of whack (or can't keep up). You need to dis-able them as well to see how lean you really are.

I think if you set 'MNRPO2_MinimumRPMOverIdleForO2Feedback' to 0, it should completely dis-able to O2 feedback (forcing the system to go open-loop). I think this includes skipping the adaptives. That value should be at 0x8a25 and it will be set to 0xBC (188 decimal).

So, if I dont have this parameter than I need to create it like i did previously and the address is 0x8a25. No scaling or anything fancy? Will I get a slider bar that I set to 0xBC?

Thanks again for your help guys.....id be really frustrated without the guidance....:thumb:

Im hoping to get another chance to play with it on Sunday.....ill post back.

[ShelGame]

Gotcha. Thanks.



Yes, it idles nice for the 20 seconds it takes for my innovate wideband to warm up and then about another 10 seconds after that......then it goes to crap.



So, if I dont have this parameter than I need to create it like i did previously and the address is 0x8a25. No scaling or anything fancy? Will I get a slider bar that I set to 0xBC?

Thanks again for your help guys.....id be really frustrated without the guidance....:thumb:

Im hoping to get another chance to play with it on Sunday.....ill post back.

I doubt you'll have that parameter ID'd in the table file. So, yes, like the others you'll have to create it. It should already be set to 0xBC, change it to 0x00. If you want, you can set the Y=axis scale to RPM, 0-8192 rpm. This is the spped over idle where the O2 feedback is turned off. I think this is new in the SBEC code. I don't remember this parameter from the SMEC code. It's basically there (I think) to force a richer mixture at high rpm. Though the stock code has it set so high that it proabably never kicks in...

[1FastCSX289]

I doubt you'll have that parameter ID'd in the table file. So, yes, like the others you'll have to create it. It should already be set to 0xBC, change it to 0x00. If you want, you can set the Y=axis scale to RPM, 0-8192 rpm. This is the spped over idle where the O2 feedback is turned off. I think this is new in the SBEC code. I don't remember this parameter from the SMEC code. It's basically there (I think) to force a richer mixture at high rpm. Though the stock code has it set so high that it proabably never kicks in...

Ill give it a shot and see how we do.

[Aries_Turbo]

rob, i took seans cal (just to follow along) and added the constant for MNRPO2 to the table file.

8a25 is 0x01

8a26 is 0xBC

is it an 8 bit constant at 8a26?

Brian

[ShelGame]

It's 8-bit, and some of the constants are 1-byte offset from the dis-assy I'm using to reference these, so I guess it's at address 8a26, not 8a25....

[1FastCSX289]

rob, i took seans cal (just to follow along) and added the constant for MNRPO2 to the table file.

8a25 is 0x01

8a26 is 0xBC

is it an 8 bit constant at 8a26?

Brian


It's 8-bit, and some of the constants are 1-byte offset from the dis-assy I'm using to reference these, so I guess it's at address 8a26, not 8a25....

Ok....so 8a26 gets set to 0x00. Thanks.

[Aries_Turbo]

yes. :)

[Juggy]

FuelBaselineFromMap is the same thing as FuelNoThrottle

so this in combination with the pumping effieciency sets the fuel for idle??
im experiencing the same problems. i can get the car to fire up and run for a minute or so then it starts to lean out harder and harder as the engine warms up....

[ShelGame]

FuelNoThrottle sets the base PW for idle. There are many many factors that modify that value.

1. PumpEff - modifier based on engine RPM
2. Cold start curves A,B,C - based on coolant temp (A, and C are phased out with time, B is always active, but the factor is 1 at normal operating temp)
3. Spark Scatter fuel compensation - based on idle speed error and others
4. Adaptive fuel compensation

If you are leaning out as the engine gets warmer, it's likely to be due to the A,B,C Curves as the adaptives do not take effect until after the engine is warm and in closed loop. It could be that the A,B,C curves add too much fuel when cold, and/or that the FuelNoThrottle table or PumpEff table don't have enough fuel for that engine operating point (RPM/Vac).

[1FastCSX289]

I finally got to work on the CSX again after a few weeks of sitting in the garage. Thankfully I have this thread to remind me of where I left off. So I disabled the O2 feedback as instructed and it still wants to correct itself......dont know how or why, but it does.

It definitely seems like it wants to lean out almost to a stall and then the computer dumps fuel and the idle shoot right up and then the cycle repeats itself. So, I decided to try messing with the fuel enrichment A, B, and C. I adjusted them and now I have a steady idle but its PIG rich.....like 10:1. And there doesnt seem to be any O2 feedback or correction at cruise.......its just runs pig rich all the time. So, I think im gonna go back to an earlier revision of the cal before I started messing with the idle corrections and just make little changes to curve B (which is where i think I got the biggest change). Hopefully I can find the happy medium. Does this sound like a logical approach or is there another way???

[Aries_Turbo]

you jumped the gun a little with the abc curves.

you seem to be lean at idle. the computer will dump in fuel and spark advance to try to raise the idle. i think thats when it goes lean and then shoots in fuel and the idle comes back. since you disabled o2 feedback and therefore the adaptives, it cant adjust the idle fuel trim itself over the long term, so it just blasts a little fuel in there with the idle control scheme.

so what you need to do.

go back to a previous cal before you started messing with the a,b,c curves.

start the car and let it get hot. whatever the setpoint on the tstat is. (so the a,b,c curves arent affecting anything)

disable the o2. it should start burbling and hunting again if it isnt already.

add fuel in the idle speed area of the pumping efficiency table till the idle hunting settles down and you have good afr at idle.

you may want to lower the boost as much as possible, go out and hammer the car against the brakes to make sure the pumping efficiency table is good in the rest of the RPM range too.

then let the car cool totally down outside. tangent time......we need to put quick disconnects on the heater hose lines of our cars and have a water pump (the blue harbor freight one i borrowed to pump fuel oil would work) and a huge radiator on a wheeled stand with a big fan on it to cool the engine down faster so we can set the a,b,c curves easier..... tangent over. :)

after the car is good and cold, (like NY october morning cold) start it up and wait at least 1 min seconds.

at this point, fuel curved a and c will be timed out and you can adjust curve B to have a good AFR as the temps rise. you may have to do this a few times to get it right

once curve B is dialed in so that when the car warms up (after that 2 min period) it has good solid afr's and the idle isnt hunting all over the place, then you can start dialing in curve A.

rob, we need the decay and hold memory locations for his cal. they arent defined.

after curve A is dialed in at both hot and cold temps, then you can dial in curve C if needed.

i have a feeling once the pumping efficiency is dialed in, curve A,B,C shouldnt need much tweaking.

curve B is in effect all the time until the car is warm.

curve C is the short term afterstart enrichment. its in effect for 5 sec and then its effect is phased out over a period of 8 sec or so based on the MP cal your cal is based on.

when the coolant temp is above 119deg Curve A is in effect for 26 seconds and is phased out over a period of what appears to be 89 seconds but im not 100% sure of that. its effect looks to be fairly small though.

when the coolant temp is below 119deg Curve A is in effect for 8 seconds and is phased out over a period of 10 seconds.

Brian

[1FastCSX289]

Well, I thought I could jump in the car and drive it to work and play with this on my lunch......but apparently I richened the crap out of it too much. I got about 1/4 mile down the road and it flooded out and I coasted off to the side. Of course, the plugs are all fouled so I had to walk home and get the truck so I could make it to work (ended up 5 min late). So, the car is stuck on the side of ridge rd. Ill have to go back with a new chip and fresh plugs to get it started and over to work. :mad: AAAARGH! When I get it here, ill follow your advice and let you know how it goes.

[ShelGame]

rob, we need the decay and hold memory locations for his cal. they arent defined.

Lemme see if I can make up a MP Tuner style table file for it real quick...

[1FastCSX289]

Lemme see if I can make up a MP Tuner style table file for it real quick...

Rob, would it help if I could send you the cal I have with the idle issues? The WOT and cruise is pretty decent.

[ShelGame]

Don't need it...

Here's a complete table file for your cal. This should be MP Tuner compatible, since I originally made it for converting stock bins over to templates. This one also has the table entries for the code hacks I added to the old cals...

You should be able to just change the name of this .tbl to match whatever your cal file name is now. Then open it up with MP Tuner or D-Cal.

[1FastCSX289]

Rob, is this a direct copy of the original cal you gave me except with additional parameters? Or is this an updated cal? I usually load a bin file on d-cal, so how would I do that?

[ShelGame]

The last post is just an updated table file. I would take your current table file (mycal.tbl) and rename it (mycal.old). Then, rename this table file to match your cal name (SBEC_24_1fastcsx.tbl --> mycal.tbl). Then you can open your cal in D-Cal or MP Tuner to edit it just like you have been. The only difference will be that the table names will match the 'new' naming conventions. And, you'll have 100% complete table data. This won't change the actual cal at all. Have you scaled your cal for injectors or MAP? This table file is setup for 2-bar MAP.

Just a warning - there are some tables used for the VNT boost control that will look like spagetti. That's because your cal is T1-based. It doesn't use those tables, so the data is all 0xFF'd out. It won't affect anything, just ignore the VNT boost control tables...

[1FastCSX289]

This won't change the actual cal at all. Have you scaled your cal for injectors or MAP? This table file is setup for 2-bar MAP.

...

You wrote my cal.....I would assume you scaled it b/c its running on 72's and a 3 bar.

[ShelGame]

I wasn't sure how much you guys had modified. And, I couldn't remember what all it was setup for. Lemme make a couple of changes to the MAP scaling and header for you. Gimme a couple of minutes...

[ShelGame]

OK, here's the table file with the MAP info and header updated for your cal...

[1FastCSX289]

OK, here's the table file with the MAP info and header updated for your cal...

Thanks Rob. Ill probably get Brian to give me a hand this weekend.

[Aries_Turbo]

yup^^^

sean you want me to come over a little earlier than 6pm to mess with Turbonator LM on the charger and your cal on the CSX?

brian

[1FastCSX289]

yup^^^

sean you want me to come over a little earlier than 6pm to mess with Turbonator LM on the charger and your cal on the CSX?

brian

Yea, we can do that.......although the fight doesnt start until about 9, so if we got together around 6 or 7, that would give us plenty of time to tweak.

[1FastCSX289]

I got to sit down with the cal and Brian's explanation of what the different warm up curves were really helped. I wound up adding a bunch of fuel to the pumping efficiency at idle and also on curve B, but I subtracted a bunch of fuel from curve a and c. I got it idling pretty well.....right around 14.5 to 15.0. I think its actually a bit rich because if you rev it up and then let it settle....it immediately settles to about 12.0:1 until the computer gets the o2 feedback and adjusts to 14. Ive got some more playing around to do but its waaay better then it was before.

[Aries_Turbo]

i was planning on being at laskys at 6pm.

brian

[Aries_Turbo]

good deal. glad it is working for you now.

brian

[1FastCSX289]

i was planning on being at laskys at 6pm.

brian

Ill just meet you there around then if all goes well on the homefront. Ill drive the charger.

[1FastCSX289]

good deal. glad it is working for you now.

brian

Yes, i got impatient and didnt want to wait for you to help me set up robs updates, so I just tried this on a whim and it worked. I guess I could sense what the computer was doing, i just needed the explanation of what each of those curves controlled. That helped a lot.

Id still like to get the updates on this cal on saturday though, so ill bring the computer along with me.

[1FastCSX289]

OK.....now ive got it idling decent. Its slow to rev down after initial start up and if I rev it up in neutral.....takes an extra long time for the revs to drop down to idle. What can I mess with to get it decel faster?

[ShelGame]

Well, it depends. When the idle hangs high, is the A/F rich, lean or just right?

[1FastCSX289]

Well, it depends. When the idle hangs high, is the A/F rich, lean or just right?

Ill have to make a note and get back to you. I dont have the car out today. Im hoping the weather holds out OK this weekend. Can you give me some possible scenarios or are there too many variables??.....

[ShelGame]

I think there's too many possibilities at this point. I could tell you how to adjust the idle speed, but I don't know if that's what's wrong.

[1FastCSX289]

I think there's too many possibilities at this point. I could tell you how to adjust the idle speed, but I don't know if that's what's wrong.

OK, ill let you know the AFR when the rpms hang. Thanks again for your help.

[Aries_Turbo]

there is a significant amount of timing in that cal iirc. i wonder if that is the issue combined with the fact that its a MP based cal and TempAboveWhichSparkScatterIsActive and ScatterInhibitBelowThisSpeed are set to keep the revs high between shifts according to the description in the table file.

Brian

[1FastCSX289]

there is a significant amount of timing in that cal iirc. i wonder if that is the issue combined with the fact that its a MP based cal and TempAboveWhichSparkScatterIsActive and ScatterInhibitBelowThisSpeed are set to keep the revs high between shifts according to the description in the table file.

Brian

Hmmmm. I did enable the spark scatter again after I wasnt having success with disabling it.