Ignition timing - initial setup problems.

[misterforager]

I am really close to firing up an installation of an 'early' AE86 4AGE (1985). All of the wiring checks out ok and there are no diagnostic codes from the ECU. Problem is that I can't get a stable ignition position when cranking the engine (no fuel in the system yet - to be safe). What happens is this:

1 - Crank engine over and timing light shows spark at (say) 5 degrees retarded and remains there. Switch off.
2 - Try again and spark is occurring at 25 degrees advanced - and I haven't changed anything!
3 - Subsequent retries will lead to it being -5, +25 or completely off the scale of the timing light (on the other side of the crank pulley).
4 - Once it has established a position (while cranking) it does 'lock on' to it i.e. it's not scattering around.

Notes:
Valve timing is spot on.
Coil lead is welll away from ECU signals (not an RFI problem).
Distributor signals are fully shielded
ECU is unknown insofar as I've never run an engine with it.
Same goes for distributor and coil/ignitor.
Coil mesurements (resistances) are ok
Dizzy pickup coils ok and gaps to 'teeth' on RPM and spark wheels are within spec.
All earths check out good.
ECU part = 89661-12060
Dizzy = 16040 (last 5 digits)
Ignitor = 12270 (last 5 digits)

I dare not put fuel in and try starting the car fuelled up until I have a stable cranking spark position. Next step will be to connect 'scopes and Logic analyser so as to measure relationship between dizzy pulses on signals Ne & G with respect to ECU output IGt (I believe that IGf is the rev limiter?). I'll stop here as somebody out there may have an answer already. My suspicion is with the dizzy or the ignitor...

Anybody seen this problem before and have a suggested solution?

Many thanks in advance.

[totta Crolla]

It might be worth trying shorting the diagnostic plug, certainly you need to do this when the engine is running and the timing is set.

[misterforager]

Thanks. I already tried that and it makes no difference. The ECU should not be commanding such massive changes in advance on cranking retries. Shorting the T-E1 terminals simply disables the internal ESA look-up map so that you can set the dizzy position when idling. I do appreciate your reply, though.

[misterforager]

The problem is solved - and I feel really stupid! Explanation is this:

1- I bought a low mileage Japanse import engine/T50 box (vintage 1985, 3-rib blue top) in 1990. The loom was cut in many places.
2 - Bought loads of bits from Roger Moran and (the sadly, late Rex Paddock) e.g. loom, coil, ignitor, ECU and clutch bits - all UK spec parts. These guys were terrific and the parts were exact for a UK AE86.
2 - Project was shelved for many years.
3 - Built new loom and all checked out ok.
4 - The dizzy has 3 wires to relay the loom signals back to the ECU. These are White, Red and Blue. On this Japanese dizzy, the white is common to both pick-up coils (G-) with the Ne and G1 signals being on the red and blue.

I foolishly assumed that dizzy red matched to (UK loom) red, white to white, and blue to black.

Not so! The dizzy common white should be connected to ECU red, the dizzy blue to ECU white, blah blah blah.

The poor ECU had no chance of working out the initial timing point (under cranking) 'cos it was not getting the correct relationship between the Ne and G1 signals, DOH!

I got some really useful diagrams and knowledge form this link and I am grateful for the insight into Denso/Toyota signals and nomenclature.

http://www.autoshop101.com/autoshop15.html

Spark is now spot-on and no excuses now for getting petrol primed up and ready to run.

Hope some of this is useful.