2500 - 3000 RPM bucking problem, at steady state throttle, on level
ground or even more when going down gentle grade. Always reproducible
when engine is hot (milder and not as reproducible when engine cold).
No noticeable flat spot when accelerating, and no bucking when gas
held steady from 3000 RPM to redline


Basic tune items good/checked

Key wiring ran, and grounds/earths clean and secure

All sensors within specs

No Check Engine Light codes

Fuel pressure steady & O2 sensor healthy/fast V fluctuation (w/DMM)
during bucking

Injectors working with CEL test, but can't be sure what they're doing
at problem RPM

--

2 possible faults found:

- no ground at AMM #4 (should be one, per Bentley, and indicates LH
ECU is not internally grounding AMM #4: both LH ECU grounds are good,
as is continuity between AMM and LH ECU)


- ignition timing good at idle, however, when rpm held steady ~2700
RPM, timing _retards_ (< 10 BTDC) instead of advancing, suggesting
problem in EZK control module (and/or may be faulty signal from ECU to
EZK). Timing advances to + 20 BTDC with faster throttle blips thru
problem rpm.

KS is putting out voltage, and no apparent problems above 3K RPM

Any thoughts/suggestions? Seems like ECU and EZK failures are pretty
rare, and usually don't manifest themselves this way...


91 900S 5spd LH 2.4.2
Thanks,
Lance


p.s. there appear to be 4 wires that connect 6-pin LH 2.2 AMMs to ECU,
and 3 wires for 5-pin LH 2.4+ (w/the 4th wire being a ground, and 5th
being batt V/power)

One of the 3 or 4 wires (pin #3 on my 91 n/a [Bosch 019]), AFAIR,
sends a variable non-linear voltage to the ECU - something in the 2.2V
- 2.9V range - depending on engine RPM

Does anyone know what the other 2 wires do?