i did some scans of my 730 tpi ecu, and i've got a few questions. the system is from a '91 camaro, obviously speed density type.

first, what should the reference pulse value be while the engine is idling? mine jumps all over the place(the idle will not stabilize). maybe something in the distributor is bad? the idle will not stabilize, and it'll won't idle below 800rpm. my knock sensor is disconnected, and i was still getting knock counts & knock retard.....how is that possible?

the idle jumps all over the place, and i have no vacuum leaks. the cam is what came with the crate motor with a hydraulic non-roller followers.

i thought that the reference pulse was the basic timing pulse from the distributor, and should vary smoothly as rpm changes.

are there any good, detailed theory of operation for the GM ecu available anywhere?

TIA.

 

If you disconnect the KS you need to put a resistor between the cable and ground otherwise the ECM will pull back timing.... if its a 90-92 KS you should put a 3.9kOhm resistor to ground...

The timing at idle on a 730ECM jumps all over the place +/- 5 deg, its in the PROM (SA adder/remover vs rpm error at idle)

/N.

 

i know the timing varies some at idle, but what exactly is the reference pulse, and what should i see on the scan?

 

The timing on the scan will jump around, here are a pic of the idle spark parameters on a 90-92 AUJP bin.

 

What is base timing set at right now? Not to belittle you, but are you sure the knock sensor is what is unplugged and not the fan switch? Folks get them mixed up sometimes.

 

the timing is set at 7 degrees advanced, and the system is from a '91 firebird installed into an '81 el camino with a custom harness.

there are no electric fans installed, but the wiring is in place.

i had to swap the location of the coolant temp sensor & knock sensor because of interference with the headers & the knock sensor. besides, the knock sensor is a single wire connector, and the fans are not.

i finally got carig mostes' software to run on my win98 laptop (i had to run the software from the floppy as opposed to the hard drive), and after using panalyzer to graph the results of the scans, i'm going to connect the knock sensor wire to ground through a 3.9k resistor, which should give me no knock detection.

if the problem was/is the knock sensor detection, the motor should smooth right out instantly.

i'll let the board know the results. hopefully, it won't rain tonight.

TIA.

 

well, i 'fooled' the knock sensor, and the difference is amazing. the idle is about 90% better, but i still think i need an IAC.

so, don't even think about disconnecting the knock sensor and expect things to reasonably work. when the missing sensor causes all sorts of knock indications, the timing changing can/will cause really eratic idle fluctuations.