LX9000 Outside Air Temperature (OAT) Extremely Low (e.g. -79C) Troubleshooting?

[lonkelly]

Something is wrong with my LX9000 and outside air temperature (OAT) sensing/reporting. The OAT values in the IGC B-records are on the order of -79C. I have noticed the LX9000 was alerting below freezing temperatures every flight, but ignored the warnings up until now.

I've flown my LS8 for a year. It came to me with an LX9000/V5 that was installed in 2011. I have slowly been learning to use the LX9000 features and recently began using the IGC log files to overlay TAS on GoPro video clips taken in flight. In decoding the IGC B-records I noticed  the OAT recorded in the B-records was very low, and when I added a NAVBOX to a map screen and played back one of the IGC files in LXSim, the OAT NAVBOX was indeed showing very low temps. While in the glider I had noticed the LX9000 alerted to below freezing temperature every flight but ignored the alert up until now.

The OAT sensor appears to be stuck against the lower nose skin of the glider with sealant. There is no information in the installation manual about installing this sensor and it's not in the "Packing List" in Section 3 of the user manual. The installation manual wiring diagram for the vario harness does show the OAT sensor, labeled LM335Z.

LM335Z is a Texas Instruments temperature sensor, available for less that $1.00. I doubt the sensor has failed, but I suppose it could have.

There is a "Temperature Offset" in the Vario TE Compensation section of the setup menus, but it has a maximum value of 30C, which isn't close to enough to fix the issue.

The sensor attaches to the vario harness, and must get its power via the vario and convey its temperature signal (voltage) to the vario, but I don't see if the analog signal is digitized in the vario or main unit.

The installation manual says, "When the main unit is connected to the vario and other peripheral units, we can carry out a functionality test. After power up the vario unit should turn on. Other peripheral units have their own visual indication, so they will be tested via the main unit."

Is there some sort of peripheral testing going on in the main unit, and if so, how do I access the results?

Has anyone had a similar problem? Was it fixed, and if so, how?

Should I just replace the existing sensor and see if that fixes the issue?

Thanks for any help with this mysterious situation.

Lon Kelly

[cdeerinck]

I would suggest disconnecting the wires from the LX, and use a multimeter to test the resistance.

Check that against what the sensor is supposed to read for the current temperature per its specs.

It could be the sensor, or the wiring.  I believe (don't quote me on this), that cold means higher resistance, so a broken or smashed wire would appear as "very cold".

[lonkelly]

Assuming the sensor is what they show in the wiring diagram, it is acts as a zener diode and needs a bias voltage, so it’s not as simple as measuring a resistance. At least I don’t understand how that would work.

[lonkelly]

UPDATE: I have access to a new LXNAV vario wiring harness and have confirmed the sensor is an LM335Z or equivalent. I've wired up a test rig with an LM335Z (less than $2) and compared the performance of the sensor on the new harness with my test circuit and they work the same. When I get a chance to get under the hood of my glider I will be able to check the cabling and sensor independently of the main unit and vario and without cutting any wires or heatshrink. I'm hoping the problem isn't in the V5. It seems to me the sensor should be calibrated (via the vario parameters) near freezing so I was thinking of holding an ice-water ziploc on the fuselage at the sensor, though I'm a little concerned about causing the poly finish to crack, so I'll probably wait for a cold day if I can get the system to display something close to the actual temperature.

If anyone is interested in how the sensor works I think I understand now and can fill you in.

[lonkelly]

My issue was shorted leads in the heat shrink tubing at the sensor that comes with the vario wiring harness. I made a little test rig on a DB15 breakout board and deduced there had to be a short somewhere between the DB15 on the harness and the sensor, and when I pulled the sensor out and peeled back the heat shrink I found the outer (-) terminal leg was kinked and shorting to the middle (+) leg. I straightened the leg with a fingernail and it started working. I think LXNAV have a problem with the design of the harness where the OAT sensor is attached. They should either have insulated one of the legs on the LM335Z sensor IC along its full length and/or cut the legs short enough they wouldn’t bend so easily. As it is, there is a risk that handling the harness will short those terminal legs. It’s certainly possible that just heat-shrinking during assembly of the harness could cause such a short. Clear heat shrink over the sensor would have been better than black for troubleshooting. My fix is to insulate the + leg with a slit length of teflon tubing and then clear heat shrink over the assembly. Hot glue over the legs would arguably be better than the teflon as in my LS-8 the sensor lives on the nose side of the bulkhead behind the nose hook, so could be subject to rain or snow leading to moisture and corrosion under the heat shrink tubing. I just don’t have any idea how hot glue might perform in very cold temperatures and I’ve rarely flown in rain or snow.

[cdeerinck]

Great info!  I'm suprised they don't QA the cables after final construction, but before shipping.  

I would recommend silicone instead of hot melt glue, as it holds up fine in water and cold.  Just my 2 cents.