Suzuki Swift 4 cyl ?
The start of any such investigation: upload vemslog showing the problem ! (Maybe it's more tricky than a slightly crippled HALL sensor).
- Despite a lot of wasted time, real investigation could not be started yet (lacking vemslog).
- Runs.
- runs well when cold
- runs well when warm, upto 5500 RPM
- primary trigger problem when warm, above 5500 RPM
- ECU was tested to sense good signal to 15x higher... So most likely the HALL sensor does not pull sufficiently to GND when warm, or does not sense the pulses reliably (home-made wheel is a primary suspect).
The engine crankshaft wheel was modified, and the ECU was also modified (apparently originally ordered with VR, HALL, but changed to HALL, HALL ?).
Maybe the pullup resistor was installed on the wrong side ? (not the input, but the internal side of the protection resistor ?)
- no sign of that. Should be good
ECU primary trigger HALL input passed test of 720000 pulses / minute
12000 RPM with 60-2 config, equivalent 60000 RPM with 12-2 config without any sign of problem.
Bad sensor test was not done. Eg. 1 diode can simulate a sensor that cannot pull down to below 0.4V (as a good HALL sensor), but only to appr 1V. That would be interesting...
Do you have oscilloscope ?
cranksensor: factory 36-2-2-2 HALL was replaced with
12-2 HALL.
- sensor type ?
- axial or radial sensor placement ?
- radial
- min and max distance (as wheel rotates)
- wheel geometrics ?
- diameter ?
- thickness ?
- Photo of the wheel ?
- window size ?
- material ? not stainless, right ?
- If I understand right, almost factory wheel, but teeth grinded off
v3 ECU has HALL, HALL input
- primtrig HALL = 620mV when measured with 1k pulldown (so internal pullup is appr 7k equivalent, although this is a bit unusual ... 2k7 .. 4k7 is usual).
When engine is warm, trigger error (and misfire) above 5500 RPM.
Clearly related to warm HALL sensor, not ECU temperature (ECU was cooled but no effect: engine cooled fixes the primtrig problem).
It is suspected that primtrig HALL sensor does not pull to GND (or to <0.5V) as it should but perhaps to 1.4V or so ?
- they claim the sensor was replaced (with another factory HALL sensor of same type) and problem persists, so it is reproducible
beware: invert primtrig edge if installing an "inline" inverter-HW, else TDC position might change. If it works with an NPN inverter (eg. the PDTC114ET inverter we provided free of charge, appr. 1.8 .. 2.5V threshold), it confirms that the factory HALL sensor had OUTPUT signal level problem (not pulling to GND sufficiently).
- otherwise most likely HALL sensor output pulses completely missing for some teeth
- a scopeshot should be captured. Real scope is best, an USB soundcard is better than nothing (we provided one free of charge).