I had an interesting day. I found the Scangauge CANSF xgauge coding for cat temps, and programmed them in. The following photos show the results:
Starting with a cold (51F) engine, the UG cat bank 1 sensors 1&2 were showing 214F and 134F, while the corresponding SG cat temps were 2499 and 1727.
When the cat reached light-up temp per the UG (sensor 1 = 749F, sensor 2 = 553F), the SG showed unbelievable 7853F and 5889F sensor 1&2 values.
The SG values were roughly 10 times higher than the UG values. I visited Linear-Logic's website and found their xgauge manual, which explained the MTH (math) coding. It explained the MTH coding had three two byte hexadecimal numbers. The first pair made a multiplier, the next pair made a divisor, and the third pair made an add/subtract value. The MTH coding for all the Fahrenheit cat values was 0009 0005 FFD8. The MTH coding for all the Celsius cat values was 0001 0001 0000.
Hmm. The Celsius values get multiplied by 1 and divided by one, so they're unchanged. The Fahrenheit values should get converted by the formula F=9/5C+32. But the values are ~10X too high. The 0009 is the 9s multiplier, and the 0005 is the 5s divisor. The FFD8 is wrong. It uses the high bit to indicate a 2s compliment to be subtracted. I want to add 32, and drop the overall value by a factor of 10. So what happens if I change the MTH coding to 0009 0032 0020? I'm multiplying by 9, but dividing by 50 (0032 hex = 50), and adding 32 (0020 hex). Here are the results.
So, I solved the SG coding error. The SG cat temps followed the UG cat temps closely when my xB was idling. Tomorrow I'll take it for a drive, monitoring cat temps with both SG and UG.
__________________
Darrell
Boycotting Exxon since 1989, BP since 2010
Have you ever noticed that anybody driving slower than you is an idiot, and anyone going faster than you is a maniac? George Carlin
Mean Green Toaster Machine
49.5 mpg avg over 53,000 miles. 176% of '08 EPA
Best flat drive 94.5 mpg for 10.1 mi
Longest tank 1033 km (642 mi) on 10.56 gal = 60.8 mpg