Can someone explain how the sensor failure detection works?
In particular what I'm looking at is the EGT sensor. I understand about the failure rate and recovery rates and the default high and low voltages, but in this instance, how do the too low temp, too low time and too low run time settings work?
Basically the problem is that my EGT sensor is showing as failed when I first power on, but I think it's just because it's not up to temperature. Looking at the voltage the sensor definitely appears to be working.
Also, it's been hinted at in a previous thread, but can someone confirm that when using the k-type thermocouple sensor input, an internal calibration is used rather than the 0-5 volt linearisation curve or not?
Thanks!
Sensor failure detection
-
pat
- Syvecs Staff - Cleaner
- Posts: 356
- Joined: Fri May 23, 2008 10:23 am
- Location: Out there... somewhere
- Contact:
Re: Sensor failure detection
Tony,
There are two types of sensor failure that come into play...
The first is hardware failure, where we look at the voltage and decide whether it's "possible" or not. The failure and recovery rates are related to this, if a sensor reading is "impossible" then it needs to be that way for a little time before it's flagged as dead. If the signal returns into the realms of "possible" then it needs to stay in that realm for some time before the sensor is "resurrected"
The second type of failure, and the one you're most likely experiencing is a "plausibility" failure. In other words, is it plausible that 5 minutes after starting the EGT still isn't higher than 200 degrees ? This can flag a sensor as "untrustworthy", it is still failed, but it won't recover. Some of the default settings are perhaps a little too stringent for road cars. It is unlikely that a race car wouldn't be up to temp in 5 minutes, you'de be out warming up, but a road car could idle that long after start so it is possible, I do see this type of failure flag on water temp occasionally... you may just need to drop the threshold and/or increase the timeout.
With regard to the K-Type input, there is an internal calibration curve that is selected by having the entire curve as zero. You can still define your own curve if you prefer, but the internal curve should be fine for most applications.
Hope this helps,
Pat.
There are two types of sensor failure that come into play...
The first is hardware failure, where we look at the voltage and decide whether it's "possible" or not. The failure and recovery rates are related to this, if a sensor reading is "impossible" then it needs to be that way for a little time before it's flagged as dead. If the signal returns into the realms of "possible" then it needs to stay in that realm for some time before the sensor is "resurrected"
The second type of failure, and the one you're most likely experiencing is a "plausibility" failure. In other words, is it plausible that 5 minutes after starting the EGT still isn't higher than 200 degrees ? This can flag a sensor as "untrustworthy", it is still failed, but it won't recover. Some of the default settings are perhaps a little too stringent for road cars. It is unlikely that a race car wouldn't be up to temp in 5 minutes, you'de be out warming up, but a road car could idle that long after start so it is possible, I do see this type of failure flag on water temp occasionally... you may just need to drop the threshold and/or increase the timeout.
With regard to the K-Type input, there is an internal calibration curve that is selected by having the entire curve as zero. You can still define your own curve if you prefer, but the internal curve should be fine for most applications.
Hope this helps,
Pat.