Anyone hear what happened to this or did the guy that was doing it just drop off the face of the earth after his tift?

 

This can be easily done by pointing the fuel pump voltage A/D channel read from channel 6 to the MAP input of choice for WB logging, disabling the fuel pump error code 54 and also editing the *.ads file to scale the datasteam value accordingly. I think you already have the address (287D).

MAP1 (Pin C11): A/D Channel 3: $30
MAP2 (Pin D8): A/D Channel 0: $00
Fuel pump voltage (Pin B2): A/D Channel 6: $60

Updated with additional changes that were discovered after the original post:

To disable error 54 in $6E, address 1E0, bit 5 should be set to 0.

Change the value $80 at address 28D0 to $7F to always input battery voltage to the injector offset table instead of the newly logged signal at L0080 that used to be the fuel pump voltage.

 

Sorry, I've been very busy lately. The code is done but I have to test it to make sure it works correctly.

 

The .ads multiplies the FP volts by .1 to output the volts.

My WB is reading ~ 3 volts (In Map1 by volt meter) but the byte # 35 is showing ~ 17 volts. This is with the .1 multiplier in the ads file.

If I multiply by .4 and add 8.8 in the ads file it reads real close to the volt meter input.

I will try to calibrate the ads file by doing a free air 5 volt input and set the byte 35 the read 5 volts then add the correction to read the AFR.

 

Sounds good. You just need to determine the correct offset and scaling for your WBO2.

For an LC-1 for example, the offset should be approx. 7.35 AFR (0 volts) according to the manual. 5volts should equal 22.39 AFR.

Scaling should be roughly 5v/255 *3 or 0.0588.

 

Cool, for a minute I thought you dropped off the board. We need as many code-literate people as possible. Eventually I hope to be one.

I ran the WB through the fuel pump pin for awhile but I didn't like the lack of resolution it had there.

 

With free air the number rolled up to 25.5, this is with 5 volts (meter measured) input Map1 C11 and the .1 "normal" volts scalar in the ads file, at byte #35 FP volts.

The ads scalar I tried was .04 not .4 and offset of 8.8. Without the scalar byte #35 reads 0 volts all the time.

It was okay around the cruise AFR, wot was 1.3 VDC measured volts and AFR of 11.5 from the chart.
Commanded AFR at wot was 11.92 and the WB output (byte # 35) corrected (scaled) was around 12 to 11.5 AFR.

I just am wandering why the volts out doesn't match the meter measured volts.

And how do you disable the low volts Error #54, I tried to lower the bypass volts to .1 (dec 1) but it still set the Check engine lite. I raised the less than to 255 and the 1.5 sec to 255.

Any suggestions?

 

To disable error 54 in $6E, I think its address 1E0, bit 5, but I haven't confirmed this. I normally use a $32B mask.

For testing the voltage, just try displaying 0-5 volts with a 0.0196 scalar and compare this to your voltmeter.

The "normal" fuel pump voltage signal and input circuitry is scaled for a 25.5 volt range for the A/D converter. There is a voltage divider upstream of the the A/D conversion. The 0.1 scaler is only applicable for the original 25.5 volt range circuit.

The MAP input that your actually using doesn't have a voltage divider in the circuit, so the 5 volt signal goes straight to the A/D converter. 5 volts will be converted to a decimal value of 255, so it should be scaled by 0.0196 to appear again as 5 volts when displayed by the ads.

With the original (0.1v) scaling the 5 volts will still be converted to 255 by the A/D but then will be multiplied by 0.1 by the ads, hence the 25.5 v that you observed.

 

1E0 is the right spot, Flag 4 is missing from the 6E hack. I added it to my copy. From the 32B 1E0 and the 32 is 1D3.

32B and 6E are both at 1E0.

I will unset b5 to 0 from 1.

Map1 and Map2 are the same but different from the FP volts input.

It is interesting how you can adjust things either by the hardware or software, to get what is needed.

The attached gif shows the difference in the hardware.

The capacitors to ground reduce noise and the bigger 4.7uf on the FP volts will give a more averaged reading. The smaller .47uf on the Map inputs will track the ups and downs better, only filtering off some spikes (noise).

 

With 287D changed the fuel is way rich.

Is this location for 6E or 32?

 

Which bin (or bcc) are you using? Note: These addresses are for "standard" bins, not the modified WB bins

287D is the correct address location for $6E (ARAP based). For an AUJM based $6E version, try 2885 instead. 286C is for $32B.

Did you change the original value of $60 to $30 or $00? (depending upon which MAP input that you're using for the WB). Verify this first.

 

Yes I am using Arap original not a WB one, the only change that affected the tune was the 287D. It was hex 60 or 96 dec changed to 48 dec or 30 hex. Just that one change messes with the fuel at idle, it is way rich. Change it back and all is well again. ?????

 

Is it possible that you're failing the greater than 16 volt (#160) check and branching to LC7EF? You could try increasing the #160 threshold to #255 to test.

;
; CK FUEL PUMP VOLTAGE
;

C7A9: JSR LE87C ; To Pump Subroutine

C7AC: LDAA L0080 ; Pump VDC, (A/D)
C7AE: CMPA #160 ; 16.0 VDC
C7B0: BCS LC7EF ; IF GT 16 VDC. EXIT via HU CK
; ... else
;
; CK DIAG MODE STATUS
; (IF DIAG V gt 0.8V & lt 2V, SET FACT TEST)
;
C7B2: LDAA #$70 ; SEL A/D CH 7
C7B4: JSR LF2F7 ; To A/D CODE

C7B7: CMPA #100 ; 2.0 VDC
C7B9: BCC LC7EF ; IF GT 2.0 VDC, EXIT via HU
; ... else
C7BB: CMPA #40 ; 800 Mvdc
C7BD: BCS LC7EF ; IF LT 800 mvdc, EXIT via HU
; ... else

C7BF: BSET L0049,#$80 ; Set b7, FACT Tst MODE. <---**
C7C2: BSET L0032,#$04 ; ENABLE SPK

;------------------------------------------------------
;
; DO GM IEEE SUM, $C000 --> $FFFF
; $4000 IN LENGTH
;
;
C7C5: LDX #$C000 ; START OF EPROM
C7C8: LDD #$4000 ; LENGTH OF EPROM
C7CB: JSR LF467 ; To Ck Sum Subroutine

C7CE: STY L0168 ; IEEE CK SUM

;
; INIT IAC & IGN OFF TIMER
;
C7D2: LDAA #204
C7D4: STAA L016E ; IAC MOT
C7D7: STAA L0033 ; IGN OFF TIMER


C7D9: LDAA L004B
C7DB: ANDA #$03
C7DD: BNE LC7EC ; BR IF NOT b0 & b1
; ... else
C7DF: LDX #$0000
C7E2: LDAA #$AA
LC7E3
C7E4: LC7E4 STAA 0,X
C7E6: INX
C7E7: CPX #$002F
C7EA: BNE LC7E4

C7EC: LC7EC JMP LC90A



;
; CK IF HEADS UP PRESENT
;

C7EF: LC7EF LDX L5803
C7F2: CPX #$7E58 ; HU ID CODE
C7F5: BNE LC7FA ; BR IF NO HEADS UP
; ... else
C7F7: JSR L5800 ; H.U. ROM Addr

 

Another thought: The MAF faults also reference the fuel pump voltage. Perhaps your setting a MAF fault before the engine is running and getting default airflow based fueling. This may explain the rich fueling at idle. You could try disabling the MAF faults, or editing the voltage threshold at LC214 (214).

;
; KEEP TRACK of SERIAL DATA & AIRFLOW HERE
;
CB97: LDAB L00B8 ; MAF Hi TMR (100Ms)
CB99: LDAA L0035 ; Get mode wd 1
CB9B: BMI LCBBF

CB9D: BRSET L0034,#$10,LCBBC ; BR IF b4, IGN OFF,
; { MNR LP MD WD 1}
; ... else
CBA1: LDAA L0080 ; Pump Volts, (A/D)
CBA3: CMPA LC214 ; 10 vdc, MIN FOR ERR 34
CBA6: BLS LCBBC ; BR IF VOLTAGE GT 10 V
; ... else
CBA8: CMPB LC211 ; ___ Msec MIN QUAL TO
; SET HI MAF ERR 33
CBAB: BLS LCBB2 ; BR IF MAF HI TIME GT ___ Msec
; ... else
CBAD: BSET L0040,#$80 ; SET b7, (MAF HI ERROR)

CBB0: BRA LCBBF


CBB2: LCBB2 LDAA L00BD ; Get Air Flow value
CBB4: CMPA LC210 ; If Air Flow <= 45 gms/sec, disable ERR #33
CBB7: BLS LCBBC ; BR IF AIR FLOW GT __ gms/S
; ... else
CBB9: INCB ; INCR HI MAF TMR

CBBA: BRA LCBBD



CBBC: LCBBC CLRB ; CLR ERR 33, MAF HI TMR
CBBD: LCBBD STAB L00B8 ; ERR 33, MAP Hi TMR, (100Ms)

 

With the 287D changed it reads the correct volts from the WB. It worked for a test run once, then I started getting this rich idle condition the WB reads lean and then dumps a bunch of fuel. It kind of motor boats a few times each time it just shows more lean and more rich till it dies. I have a prominator and if I switch to the unchanged 287D, it is fine. I even tried just changing the loc287D back reloading it in the same spot and all goes back to okay.

Since it worked once, I would agree that I am triggering something, I had to raise the Maf limit to 60 from 45 because of the 383. This was done 1 year ago.

I am also running NB simulated, I can try hooking the heated NB back up as I have two Bungs a foot apart in the same pipe.

Thanks for the replies, I am not as good at code as I need to be, but I understand how the data flows, all the jumps, and loops. I am just not real up on this ECM.

 

I don't know why you would use the FP input when he made a new bin that works for the WB on another input.
Using the FP input is obviously causing issues because fueling is changing with WB voltage. Disabling the error will not make the jumps and code operation change. Those would have to be delt with accordingly (once identified)
Q: what is the voltage range on the FP input?
It can't be 5V / 255, it must be higher and dropped using a voltage divider. So what would the upper scaled value be based on the hardware installed? 18V, 20V?

 

I see something else that may be the culprit:

This portion of the hac isn't clearly commented, but it appears that the fuel pump voltage (L0080) can be conditionally input to the injector offset calculation, instead of vbatt (L007F) depending upon the BITB #$20 (bit 5) -> BEQ LE871. (BEQ: Branch if = zero). So if bit5 is 0 then it will branch and bypass loading the battery voltage.

This could be screwing up the injector voltage offset compensation.

It looks like if error 54 is bypassed previously by branching to LE8C5, then bit 5 isn't set before it is checked later.

One method to test this idea, would be to set all of the values in the injector vbatt offset table equal to the value at 14.4 volts for example, to eliminate the influence of changing voltages on the injector pw.

Another even simpler method to test would be to simply load accumulator A with battery voltage (Use LDAA L007F instead of LDAA L0080) at E8CF before the BEQ instead of pump voltage, so that battery voltage would be used regardless of the bit 5 condition.

E8BE: LE8BE ORAA #$0010
E8C0: LE8C0 BSET L0051,#$20

E8C3: ORAA #$0020

E8C5: LE8C5 TAB
E8C6: ANDB #$0004
E8C8: ANDA #$00F3
E8CA: ASLB
E8CB: ABA
E8CC: STAA L0041
E8CE: TAB
E8CF: LDAA L0080
E8D1: BITB #$0020
E8D3: BEQ LE8D7


;----------------------------------------------
; INJECTION OFFSET VS BATTERY
;
; TBL = 0.065536 * usec
; TBL = 65.536 * Msec
;----------------------------------------------
E8D5: LDAA L007F ; BATTERY VDC

E8D7: LE8D7 LDX #$C3B5 ; INDEX INJECTION OFFSET TBL
E8DA: JSR LF3FF ; Call 2d Lk Up, (No Offset)

E8DD: LE8DD STAA L00D2

E8DF: RTS

 

To simplify the previous post:

I think that changing the value $80 at address 28D0 to $7F will correct the fueling issues.

 

Thankyou for the replies. I will try your sugestions TB and also the other Arap WB bin.

I track my logs using the run seconds for the data range in excel, so I would have to use something else.

Just for my understanding, to use the TunerPro hex editor, roll down to 2870 and look over to the, third to the last column, this would be 287D as the last column would be 287F.

The 287D changed from 60 to 30 and back.

The row reads 70 71 72 73 74 75 76 77 78 79 7A 7B 7C 7D 7E 7F. This is the address.
The row reads 1D 15 3C 01 B7 01 B1 27 03 14 C3 01 86 30 BD F2. This is the value (data)

Is this correct, I have been just loading 287D as a constant value and it displays in dec. I make the conversion and change there.

287D is 96 dec or 60 hex this is changed to 48 dec or 30 hex.

 

Pandin,

In the TunerPro hex editor, you are correct that the last column in the line starting with 2870 represents the 287F address. The next to last column would be 287E. You should be changing the value at 287D.

Changing the value shown above as $30 (from the original $60) as you have shown is correct. This changes which A/D channel is read. $30 means A/D channel 3. $60 means A/D channel 6, etc.

I ran a quick test with $32B to always load vbatt instead of fuel pump voltage for the injector offset table. It seems to work fine in this manner.

I also checked a hard-copy source code that I have on hand. It confirms that the fuel pump voltage is the primary input for the injector offset. So I learned something new as part of this exercise, afterall.

Vbatt or ignition voltage is used as a back-up input if there is a malfunction present on the fuel pump voltage signal. (This is the bit 5 check). This would explain the behavior that you have reported.

If you change the value $80 to $7F at 28D0 as indicated in my previous post, it should work properly.

 

JP86SS,

This presents an alternative approach, that can be done by means of simple bin editing as opposed to an actual hac or relocation effort.

I've been using this technique to log other 5 volt signals for some time now, but I wasn't aware of the injector voltage offset impact, due to the small offsets that I typically use with my 42 lb. injectors. The problem was present, but not significant enough to be noticed.

Note: The fuel pump voltage input circuit isn't used, only the fuel pump read logic, storage, and datastream implementation is used. You're right, the fuel pump input circuit does include a voltage divider for a 25.5 volt range, which makes it undesireable for this purpose.

This method makes use of one of the MAP input circuits, similar to the WB modified bins.

If it were desired to log multiple 5 volt inputs, both the new bin and this approach could be combined together. Say if you wanted to log a WB and a 2 bar map sensor, or whatever. Its another way to skin the same cat.

 

It worked fine thanks tequilaboy.

Add one wire to the harness, change two code numbers, and adjust the ADS file to read correctly. This is a real simple way to add the WB to the data stream.

I am still going to try the NB with the WB reading to the data stream, to see if there is any difference.

To JP86SS I am glad that you have shared and helped to get this WB data logging to work. It was not my intention to belittle 400Z28Racer and your work. I am just not up to speed on the disassemble and reassemble of the code, so I was looking for a "close enough" way to do it.

This would be a good place to read the EGT.

 

This is a Dec change from 128 ($80) to 127 ($7F).

 

I take no offense to such things. Its just harder to troubleshoot where there are unknowns in the code/hardware to deal with.

Yep, running the $8D I still haven't got mine to read right, mainly due to the way I have it setup and the board I bought to read the TC. Even with the pull up resistor removed there is still an impedence problem. I added an op amp as a buffer and it still is giving me fits. I finally just installed a gauge so I could see it but it would be much nicer to log it as well.