Injector current trace shows slower opening when hot
 

These are both taken with Vbatt = 14.4 V

COLD (30°C)
https://www.miataturbo.net/attachmen...1&d=1343708858

HOT (restart 5 min after coolant at 99°C)
https://www.miataturbo.net/attachmen...1&d=1343708858

These are with hi impedance 600cc injectors.

There is at least a 100 us difference in dead time cold vs. hot.
This 100 us represents a 15% change in fueling at idle.

The reason for the above is that copper's resistance goes up with temperature. I saw a shift from 15-19 Ω.
With a higher resistance, it takes longer for the current to build and reach the point where it can begin to move the pintle.

What's worse is that if actual dead-time at hot is set too short, when the ECU calculates a narrower pulse (like if O2 feedback is swinging back down), it will lean out way more rapidly than intended and can cause stalling.

This explains why if I set my dead-times for warmed up conditions, when cold O2 feedback goes tto -15 to -20%.

If I set my dead-times for cold, it tends lean misfires at low MAP, and stalls abruptly when O2 correction swings negative. AFR will go slowly leaner, than abruptly go very lean.

I have an idea for a circuit to address this problem, but it would be neater if Reverant put it into his new enhanced board. ;)

LOLz motoiq did the same thing with looking at current:
Tuners: Test Your Injectors!