I have a customer with a PNP vems ecu that ever since swapping to a 7a flywheel is having some apparent trigger problems. The car ran without issues on the AAN flywheel.
The ecu IS equipped with a cam hall inverter (which is required for use with the 7a flywheel). On initial startup the car would not idle properly, with the tach displaying random high rpm values in vemstune. We were able to partially fix this by slotting the mounting hole on the vr sensor block so that the sensor reading the reference pin gets rotated closer to the face of the flywheel (the reference pin on the 7a is much smaller than the one on the AAN or other flywheels). By doing this, the car idles fine. However, around 1500rpm the same basic symptom comes back: the revs jump randomly to ~14000 rpm in the software and as you might expect, the ecu cuts fuel for the rev limit features.
Oddly, if the throttle is held in it is possible to 'rev past' this problem. Once the revs get up to about 2k+ rpm the car will rev smooth to redline no problems.
here is a zip file containing various .vemslog and .triggerlog files.
http://www.efiexpress.com/downloads/jason_trigger_logs.zip
the trigger logs seem to indicate that there are no problems with the primary trigger signal, so I suspect the problem is in some way related to the secondary trigger. The car is not local, so I'm playing somewhat blind here.
The next possible steps I've contemplated are shipping him another ecu in the hopes that an incorrect pullup resistor may have been fitted
- that can, and should be measured (DC bias voltage with ignition on)
- although I would have expected problems with the other flywheel if that was the case
- or we would expect that it does not start at all (most of the trigger problems with these VR sensors happen during cranking or cold cranking, when the amplitude is lowest)
- anyway, if it's close to margin, than - with this sensitive type, like any low toothcount VR - a tiny difference can cause it to not work
- or perhaps fitting a larger pin to the flywheel. I'd like to avoid that, since I do have customers running these flywheels without issues.
Suggestions?
This arrangement of sectrig-VR cap (2009-08 => 2010-02, appr ~2400-2900) is not as originally planned (this is a photo of ANOTHER board):
-- SN should be below the threshold for that problem: this one is 2106
Although it passes tests, and it works, it is not as intended. It would more likely cause trigger problems with the higher amplitude signal (with the AAN flywheel), if ever (possibly work well without symptoms).
Do you know how it was ordered ? If 1k pulldown, 1k5 pulldown or 12k pullup or other is used for sectrig-VR ?
-- it was likely ordered same as all the other ecus I order:
-- 22k pullup primary and 1k5 pulldown+mode 5V sec
-- I cant find the original order number though, perhaps you can lookup that info from the webshop backend to confirm?
In any case, it would be good to know.
- What is ECU serialnr ? (order-ID or order date ?)
-- ECU SN is 2106
- Also, what is sectrig-VR sensor resistance ?
-- should be in the order comments, assume it is 1k5 but check the order comments.
- what is measured DC voltage on sectrig-VR sensor ?
-- did not measure, is this externally measurable by the customer? across what pins? At the moment I'm probably just going to ship him an exchange box and measure the details of the 'problem' box when I get it back.
Although it's very likely a HW problem, we should inspect everything, including sw.
- firmware ?
-- 1.1.67
- saved .vemscfg config ?
you should be able to extract that from the .vemslogs I provided in the above zip.