My 88 auto intermittently idles at 900 in gear w/no codes. I have isolated it to the ECM and am looking for a vendor to repair it. It is definitely the ECM causing the issue as when the instant mpg displays valid numbers (before stopping) the idle is normal (650). When the instant mpg displays 0 (before stopping) the idle is high. If someone has had this issue or just knows of a solid vendor to fix it, please advise me. Thanks in advance.

 

What’s the IAC duty cycle at hot idle,

 

That sounds very similiar to my issues. I found a refurb ECM in the For Sale subforum here and the problems went away.

 

Sounds like excess min air (adjustment). Warm/hot idle speed in gear for auto trans cars is typically 550-575 rpm.

The IAC control is rather complicated. This is only a brief summary: The idle speed is not really controlled above 1.8 mph. Otherwise, the IAC will follow the throttle and will be parked in an intermediate position while off throttle while moving (dependent upon coolant temperature and history, fan, etc). With the converter locked the engine speed will generally follow the vehicle speed (for a given gear). With the converter unlocked the engine speed will vary with the airflow through the throttle and IAC (and any vacuum leaks) and will also be influenced by the vehicle speed, gear ratio and converter slip (see the N/V ratio).

Here is a log from an 89 showing some example IAC behavior with respect to throttle position and vehicle speed. The car comes to a stop near the end of the log, however the driver almost immediately shifted into park (or Neutral) at 0 mph, causing the desired idle speed to increase by 50 rpm. The idle speed is only being controlled (PID control) in this region of the log.

https://datazap.me/u/tequilaboy/logs...08&axes=shared

If the rpm is consistently high (overspeed error), it may not enter PID control.

;------------------------------
; PID ENABLE CONDITIONS
;------------------------------
LC543 FCB 3 ; 1%, ENAB T/F IF TPS GT THRESH%
LC544 FCB 0 ; 0, DONT STORE NEW TPS IF ID DIFF VAL < THRESH
LC545 FCB 6 ; 1.8 MPH/3.2, ENAB IAC PID IF Vss > THRESH
LC546 FCB 4 ; 50 RPM/12.5. MIN RPM UNDERSPEED ERR TO ENABLE PID DURING TRANSITION DELAY
;----------------------------------------------

;----------------------------------------------
; IAC TARGET SPEED vs COOLANT TEMP
;
; Dissassemby of ARAP, LINES = 17
;
; TBL = .08 * RPM
;----------------------------------------------
ORG $052E ; RPM Deg c COOL
;----------------------------------
LC52E FCB 84 ; 1050 -40
LC52F FCB 84 ; 1050 -28
LC530 FCB 84 ; 1050 -16
LC531 FCB 84 ; 1050 -4
LC532 FCB 84 ; 1050 8
LC533 FCB 76 ; 950 20
LC534 FCB 72 ; 900 32
LC535 FCB 60 ; 750 44
LC536 FCB 48 ; 600 56
LC537 FCB 48 ; 600 68
LC538 FCB 46 ; 575 80
LC539 FCB 46 ; 575 92
LC53A FCB 46 ; 575 104
LC53B FCB 46 ; 575 116
LC53C FCB 46 ; 575 128
LC53D FCB 46 ; 575 140
LC53E FCB 46 ; 575 152
;----------------------------------------------

 

Thanks for your help, but there is nothing mechanically wrong with the car. This issue has been isolated to the ECM. I'm just seeking a reference to someone to repair the ECM.

 

The reason that the instantaneous mpg may display zero as the vehicle speed is decreasing is due to the running distance counter not updating in time to show that any distance was traveled during the calculation time interval (presumably one second). The mpg calculation is done by the dash, not the ecm.

Running distance and running fuel counters added to the display (same log): https://datazap.me/u/tequilaboy/logs...08&axes=shared

Note: the above log is slightly misleading since the running distance was only displayed to two decimal places in the adx file that was used prior to exporting this log to csv. The running distance counter should increment every 2.64 feet (or 0.0005 miles), so iin reality it will ramp more smoothly than the steps displayed above.

Below 1.8 mph (2.64 ft/sec) you can expect the mpg to go to zero, which is coincidentally the entry speed threshold for PID control.

 

Another possibility is that you're intermittently entering ALDL (160 baud diagnostic) mode due to a possible short between ALDL pins A & B that mimics a 10k resistor. If so, this will replace the (4 byte) 160 baud serial data string that is sent to the dash (plus Mode Word 2):


;======================================= =======
; 160/8192 BAUD SERIAL DATA PARAM'S
;
;======================================= =======

LC611 FCB 4 ; DIS STRING LENGTH

;======================================= =======
; DIS ADDRESS'S , XMIT
; (4 BYTE, 160 BAUD MSG)
;======================================= =======
LC612 FDB $C009 ; Num of Cyl
LC614 FDB $0123 ; RUNNING TOTAL OF SUPPLIED FUEL
LC616 FDB $0127 ; RUNNING TOTAL OD DIST TRAVLED
LC618 FDB $C620 ; Gal/sec SCALE FACTOR
LC61A FDB $0000 ;
LC61C FDB $0000 ;
LC61E FDB $0000 ;
;---------------------------------------------
with the diagnostic (25 byte) 160 baud ALDL datastream and it will also set the desired idle speed to 1,000 rpm independently of the coolant temperature or gear position.

Check the voltage on ALDL pin B.

 

Tboy, I'm afraid I have to disagree with your statement about the cluster calculating the instant MPG. If that were so, the cluster would have to have a throttle sensor input and it doesn't. The ECM has the throttle, vehicle speed and RPM inputs. This allows the ECM to calculate the instant MPG and send it in the serial data to the cluster where it is parsed out to the instant MPG display. My car has been thoroughly diagnosed and the ECM is the culprit here; it is not sending the correct info in the serial data stream and it is affecting the idle.

 

Perhaps your diagnostic sources are overconfident and not familiar with the C4 ECM variants, and are quick to say it's the ECM? I'm not saying it's not, but I will say that your diagnostic logic is a little bit flawed.

The ECM estimates MPG from:

injector pulse width (IPW)
injector flow calibration constant
number of injector events
vehicle speed sensor (VSS)

The instantaneous MPG calculation is basically:

MPG = Fuel Flow Rate / Vehicle Speed
​
And fuel flow rate is derived from injector on-time, TPS is irrelevant.

If the ECM accurately knows:

how large the injectors are
how long they are open
how often they fire

It can estimate fuel used per unit of time accurately... then Vehicle speed from the VSS is then compared against calculated fuel flow.

TPS and RPM are indirect/contextual inputs (irrelevant)