I've tuned my '165 close to stoich in open loop, but it's targeting 13/1 in closed loop. All the O2-related tables are stock and it refuses to run anywhere near the target afr. It's a 2.8 v6 with a '165 ecm running $32b using an LC-1's NB signal for an O2 sensor. The ecm (or maybe the sensor?) seems to think that 13/1 is stoich. I've done a free air calibration to make sure the sensor's not off. Suggestions?

 

sounds like a reincarnation of my issue. had to dump my WB's simulated NB output and go with a 3 wire (ACDelco AFS 74) and change O2 values.

does it feel a little sluggish? 13.0 AFR is great when you're rippin in PE, but you can feel the engine laboring under normal driving with the super-rich AFR.
(my experience anyway)

see what your O2 mv's are in closed loop when your AFR says 14.6/7. then use that as your Mean rich/lean value and adjust the lean and rich tables to compliment

just an idea, it worked for me.

 

I'll try that. It doesn't feel too sluggish, but I can tell it's rich when I give the throttle a stab and it hesitates. Any idea why it's so far off from the target afr? In open loop it runs at 14.6 when the target afr is 13.5. Maybe the BPW formula is off?

 

that's a possibility

 

My thought was that since part of the formula incorpirates a variable based on the # of cylinders, and I have a v6 instead of a v8 the formula would be off, but changing the code 41 parameter should've adjusted for that.

The formula in $32B is BPW/5 = 89 x KINJXBC x LV8 / (1024 x KLVMSC) where KINJXBC = (inj flow in g/sec x 5) x 256 and KLVMSC = ?
Anyone know what KLVMSC is?

 

Do you mean it's "commanding" 14.7, but it's "getting" 13:1? i.e. the wideband shows 13:1?

That would mean there's probably a problem with the narrowband output of the LC-1 - maybe the default LC-1 calibration in it has the switch point set to 13:1, rather than 14.7. Try reprogramming the LC-1 to switch at the correct AFR. I only use LM-1's so I don't know how to program an LC-1... I think it's possible.

 

Yep, it commands 14.7 and gives me 13/1 on the wb. Already tried changing the NB signal, didn't seem to have much of an effect. Here's a screenshot of what I have programmed in the LC-1.

 

Sounds like it's not in closed loop. Can you see the INT changing up and down? Otherwise maybe you're not actually programming the LC-1 when you think you are... or it's bugged.

 

It enters closed loop- the Int unlocks and the loop status changes to closed in the datalog. I'm going to contact Innovate and see what they have to say. Maybe it's a bad ground, on their website they make it seem like that's the cause of nearly all problems.

 

Does the MAT do any kind of conversion to the signal? I have the WB signal going into pin C12 (MAT) for datalogging.

 

Probably has a voltage divider in it... It normally measures a resistance only.

 

So are wb readings off then? Can I compensate for that in my coversion formula somehow? My current formula is (X x .058980) + 7.35. Or is there a better pin to connect the wb?

 

Sorry, I don't know. I've never messed with that. I just log with the serial cable, and mix the logs together with excel (but I'm using my company's ecu and software, in China - I'm now somewhere in Inner Mongolia province (Ya Ke Shi), -32C this morning).

I think it would have to go to an analog input, perhaps something that measures a 12V signal, such as fuel pump voltage, since the 5 volt reference is not exactly valid.

 

Maybe I'll just plug it into the egr diagnostic pin. It has a linear 12v based input so it should work, and I disable the egr during tuning anyway.

-32*C in northern china hu? And I've been complaning about -12*C here in denver!

 

Grrr! Innovate wan't any help. I explained my situation and all the guy had to say was "Well, good luck." Feels like I'm beating my head against a wall with this. Checked out the ground- fine. Grounded it to the block- no change. Tried hooking the wb through the egr pin (controll and diagnostic) and didn't get a read through it either. Went to innovate's forum for help and that didn't help much either.

I'm stumped.

 

Try to see what it does on the fuel pump voltage pin. Isn't that the one everyone uses on the SAUJP (I haven't looked at the schematic, just going by a fuzzy memory)? EGR is not an analog input (neither one).

 

im using SAUJP V4, if i recall im using E10 for my wideband input. it did require i make a mod my ECM, i had to cut a trace to do away with the 5 volt pull up.
it matches almost perfectly what my wideband software shows, i think the difference is because the ZT2 software logs a little faster than what i get from the data stream of the ECM.

 

He's using a 165 not a 730/749 . SAUJP is for $8D only.
Just as a test of your system, setup a controllable 5 volt source and see if the voltage you input is being converted to the correct value within the ECM.
Put in 2.5 volts and see if your "counts" are 128.
The "actual" numbers don't matter and can be scaled/offset accordingly so they match the engineering units (AFR) you want to see.
If you are not doing any assembly code work within the ECM program then you could build a lookup table for all 255 data points within the ADS file.
Granted, only needed if you have something skewing the signal and can't correct it properly.
Jp

Edit: You could use D8 (TPS2) if its not already used for something.
If you have a hac of the 32b you could work with the source code to make any input work for you. There is a thread that I went throught this for 6E that worked.
Do a search for "wbo2 closed loop"

 

someone me, i deserve it for not reading through the thread again before posting.

 

It's cool, any help is good help.

The fuel pump signal pin works, but the lack of resolution was disappointing. I narrowed the wb's signal to 12-17:1, but even then it's only using 50 of the 255 datapoints. And for some reason it seems to be running a bit rougher. It is targeting 14.7 just like it should, so the MAT pin was what scrambled things up.

What does the '165 use map input for?

 

Edit- nevermind, doing a helps sometimes. LOL

 

I'm curious if your base pw is being calculated or if you are using the table?

This switch setting also has the effect of substituting the MAT in place of coolant temp in some of the tables (start-up enrich, crank pw, AE cool factor etc.). This could possibly explain why using the WB on the MAT input was impacting the AFR.

I'm not clear on which switch setting (bit set or cleared) results in the increased dependency on the MAT. Perhaps your answer will help clear things up. I'm assuming at this point that using the table results in more influence from the MAT vs. coolant temp.

 

That's interesting, didn't think the ecm would switch sensors like that. I'm letting it do the calc'ing, I haven't really had the time to get into tuning with the BPW table. Alot to do there because the stock v6 BPW table isn't anywhere near accurate, infact the table's 2 dimensional in the v6 ecm. No RPM adjustment whatsoever!

 

Here's one of the relevent excerpts:

;-------------------------------------------------
; VOLTAGE COMP, COOL or MAT
; LK UP DECELL & ACCEL Enr Cool Factor vs COOL
;
;--------------------------------------------------
EDE0: LEDE0 LDAA L005D ; COOLANT
EDE2: LDX #$C35D ; COOL MULT Tbl
EDE5: JSR LF3F8 ; 2d Lk Up (W/LINE CNT)

EDE8: STAA L00EE ; Accel Enr Cool Factor


EDEA: LDAA L005D ; COOLANT

EDEC: BRCLR L0044,#$80,LEDF2

;
;
;
EDF0: LDAA L0061 ; MAT
EDF2: LEDF2 PSHA
EDF3: CMPA #208 ; CK MAX LK UP VAL, 115c
EDF5: BLS LEDF9

EDF7: LDAA #208 ; USE 115c MAX FOR LK UP
EDF9: LEDF9 PSHA
EDFA: LDAB L0035
EDFC: BMI LEE4F
; ... else
EDFE: LDAB L0001
EE00: BITB #$08 ; b3
EE02: BNE LEE4F
; ... else
EE04: LDX #$C3ED
EE07: JSR LF3FF ; Call 2d Lk Up, (No Offset)

EE0A: PSHA
;----------------------------------------------

And another:

;----------------------------------------------
; IDLE SPEED TIME OUT IAC STEPS vs COOLANT or MAT
; (SEE LC016 b6 FOR OPTION)
:
; TBL = STEPS
;----------------------------------------------
; STEPS Deg c COOL/MAT
;----------------------------------
LC59F FCB 30 ; 30 -40
LC5A0 FCB 30 ; 30 -28
LC5A1 FCB 30 ; 30 -16
LC5A2 FCB 30 ; 30 -4
LC5A3 FCB 32 ; 30 8
LC5A4 FCB 10 ; 30 20
LC5A5 FCB 10 ; 30 32
LC5A6 FCB 10 ; 40 44
LC5A7 FCB 10 ; 40 56
LC5A8 FCB 10 ; 45 68
LC5A9 FCB 32 ; 50 80
LC5AA FCB 60 ; 60 92
LC5AB FCB 60 ; 60 104
LC5AC FCB 60 ; 60 116
LC5AD FCB 60 ; 60 128
LC5AE FCB 60 ; 60 140
LC5AF FCB 60 ; 60 152
;---------------------------------------------

 

Thanks, I'll keep that in mind when I start adjusting the BPW table. One more question, what table or constant controls idle AFR offset? I'd like to get a lean idle going, about 15:1. Every drop counts nowadays.

 

After going through the 32b hac there doesn't seem to be any direct way of controlling the idling afr in open loop. I could crutch it with the maf tables or maybe the low pw offset table, but that would compromise other areas. Changing the open loop afr %change tables doesn't seem to have an effect.

 

The open loop % change vs coolant temp should have an effect on the target AFR in open loop. Negative values should move your target AFR in the direction of 15:1 or greater as desired.

0% should result in a stoich AFR just like in closed loop (assuming the % change vs. load table isn't adding fuel).

Perhaps the open loop % change vs load is complicating your results. I tend to forget about this one, since I have this table zero'd for the idle load region for more stable idle fueling.

 

Actually, with no compensation (in either table) it targets 17:1. I have no compensation at idle region in the load table and minimum setting of 14% in the coolant table, which gives close to 14.7:1 target afr anywhere other than idle. I don't seem to have any control over idle afr, so there must be some other constant controlling it.

 

Has the stochiometric AFR constant been changed?

Address: LC3D2 in $32B (normally 0x1BD or 445 decimal for 14.73:1)

AFR = 6553.6/ARG

 

No. I'll take a second look at the coolant table... maybe I got the conversion factor wrong somewhere when I changed my copy of 32b for the v6.

 

I've got the conversion factor right, but it was running far enough out of the wb's range that I couldn't see any changes. It's wandering between 16.5 and 17:1 right now! I didn't notice how lean it was until I started concentrating on idle. I found the constant that controlls idle afr, it's at LC425 and it's called "open loop a/f bias for tbl below (open loop a/f %change vs coolant temp)".