EGT dropout on M800 PnP w/OKM

[mrsaturn7085]

Alright, more progress...

I extended the Diag Hi value of the sensor from 4.9 V to 32 V just so I could get a raw reading of what was happening past 480 deg C. It appears the Diag Hi value was the cause of the problem... which now begs the question - what does the OKM actually output to the M800, and what are the correct Diag Lo/Hi values? I assumed a standard 0-5 V reading but it doesn't look like this is the case.

I ran the sensor up to 1000 deg C with a handheld propane torch and the voltage was approximately 10.3 V. 250 deg C was approximately 2.5 V. In fact, as I watched it creep down, the temperature is almost a direct 100x multiplier on the voltage. This would explain why the 4.9 V Diag Hi went into error at around 480 deg C.

Since the actual temperature values are reasonable both on the car and for when the Inconel sheath started to glow cherry red, the "#65 Temp(C): OEM K Type Amp" calibration seems to be correct.

[mrsaturn7085]

After some further research - I've determined that the (roughly) 100x scaling for voltage to deg C is correct for the OKM board output to M800 plug-in.

"#65 Temp(C): OEM K Type Amp" very clearly confirms that this is how the OKM amplifier was designed. The table limits are set as 0 to 12.52 V and I will therefore be extending my Diag Lo and Diag Hi values as follows:

Diag Lo: -0.48
Diag Hi: 13.00

This should give me an EGT range of -48 to 1300 deg C.

I have set a warning value to 900 deg C and left my default value as 0 deg C.

The information about the EGT amplifier on-board the OKM board is almost completely non-existent in all MOTEC white papers that I have come across - hopefully this will assist anyone else that runs into this issue, in the future!

[SprinterTRD]

The OKM will have the same calibration as MoTeC's thermocouple amp

[mrsaturn7085]

The OKM will have the same calibration as MoTeC's thermocouple amp

No, it doesn't - I'm not just throwing this information out... this has all been tested and verified in the last 48 hours.

The #65 calibration is the only one that works with the M800 plug-in w/OKM, and looking at the table for the #65, it is a 0 to 12.5 V scale. The TCA is a 0 to 5 V. This was confirmed w/MOTEC Systems USA, though they were just as surprised as I was.

[Scott@FP]

Sort of behaving like an ungrounded probe that is grounding, or a cracked probe microweld.

[mrsaturn7085]

Sort of behaving like an ungrounded probe that is grounding, or a cracked probe microweld.

In my case, the problem was ENTIRELY one of incorrect settings in MOTEC due to lack of data in white papers - the Diag Hi voltage limit was set too low. The sensor functioned well into the 1000 deg C range when tested with a torch (after extending the Diag voltage range).

[Rude Engineering]

I would suggest setting your Diag Lo value to something like 0.1 if you do not want to see any errors. You are using the correct calibration and it does go up to about 12.5V full range.

Logging Error channels is a useful way of detecting this type of problem. Unfortunately the default diagnostic settings are not always suitable for all sensors. In some instances a sensor / sensor wiring fault will not trigger an error at all.

[mrsaturn7085]

The default Diag Lo is 0.1 V, however, due to the K type being able to read negative temperature values, I selected -0.48 V as the Diag Lo.

I do NOT know if the OKM will output below 0 V, but this will at least prevent a 0 V signal (which would be entirely possible in 0 deg C weather) from locking the Error Flag (and therefore, the warning light) on cold start in Winter.