Oil Temp Measurement Issues

Bubblehead · #1 10-07-2012, 09:12 AM

This might better go into the Glass Panel area - I'll let the moderators decide.

Since installing SkyView late last year I have had an intermitten indication on oil temp. The probe was still a GRT two-wire model from when I had an EIS-4000 in the plane. After reading up on things on this and the Dynon forums the most common diagnosis was "Bad Ground" so I rewired the ground with no improvement. I decided to change to the Dynon one-wire probe that came with the probe kit so yesterday prepared the water and stove and thermometer to check the calibration on the probe. TERRIBLE! 10 to 15 degrees off all the way up to boiling and back again, and yes I did change the sensor map to the Dynon 5/8" probe.

So I grabbed the original Lyc probe from when I had a Vans analog gage and tried it, again changing the sensor map to 1/8" probe, with the same results. I even tried both probes with the other sensor map settings with no improvement. Did both sensors again with different ground positions, and checked the continuity of the ground. The resistance is about .3 ohms which is supposedly a lot for this type of probe.

The last test of the day was with a new GRT two-wire sensor I had in the spares box. This sensor tracked beautifully right up to 150F and then failed! I turned off the stove and it started working again when the water bath temp dropped below 150. Turned the stove on - failed again.

I am not sure what to do now. The Dynon probe was inaccurate as heck and the GRT probes fail. The one thought I have is to create a custom sensor setup for the Dynon probe to compensate for the inaccuracy, essentially creating a temperature offset. Seems like a band-aid over a free-flowing wound though. Maybe the better way is to just adjust up the alarm setpoints and operating ranges on SkyView by 10 degrees and don't worry about it. That just does not seem right either. I really want to know the oil temp and have confidence it is accurate within a few degrees throughout the normal band.

Any suggestions?

az_gila · #2 10-07-2012, 12:13 PM

I think Lycoming makes several single pin sensors. The one I measured was resistive, and the 0.3 ohms you mention would be a pretty neglible error.

Here is my chart, ohms vertically and degrees F horizontal. I'm too high to get 212F as boiling...

For the certified SW gauges, the meter calibration point is the red square - 33 ohms at 240F.

Were your measurements similar?

Bubblehead · #3 10-07-2012, 04:43 PM

Great info as always Gil. You are always one of the first to post and always provide real substance to your posts.

i did not measure resistance vs temp, just compared the thermometer reading with the Skyview display. I just read up on the Dynon website how to use Excel to generate the coeficients to put into the sensor definition file and it's not too tough. I may go to the hanger and measure R vs F and try another definition.

az_gila · #4 10-07-2012, 06:03 PM

Quote:

Originally Posted by Bubblehead

Great info as always Gil. You are always one of the first to post and always provide real substance to your posts.

i did not measure resistance vs temp, just compared the thermometer reading with the Skyview display. I just read up on the Dynon website how to use Excel to generate the coeficients to put into the sensor definition file and it's not too tough. I may go to the hanger and measure R vs F and try another definition.

Thanks... perhaps you can do a few resistance/temp spot checks and confirm the curve you have is similar to the one I showed..

I was surprised at the steepness of the curve.

The curve on the graph is a Excel one generated from the points shown.

The 33 ohm reference came from the Grumman maintenance manual on how to test the SW type meter indicators. Replace the sender with a 33 ohm resistor and the indicator should read 240 F.

Bubblehead · #5 10-08-2012, 06:08 AM

Gil, if convenient would you email me your data for the graph. I'll put mine on the same spreadsheet for comparison. My email address is jdalmansr@gmail.com. Depending on workload today I will go to the hanger and make the measurements. Seems like the 33 ohm test would be very useful for checking the gage out and eliminating one source of trouble.