'7747 CTS in A/D Counts?? (ALDL)

Darrell N. darrelln at datalog.ab.ca
Wed Jul 11 21:13:17 GMT 2001


> Is there a formula to convert the A/D counts into degrees celsius? I've
> found tables on both Dan and ECMGuy's pages that show the correlation, but
> no formulas.

There isn't a formula per se, but it's not that hard to calculate.  Usually 
the A/D converter reads the voltage at the center tap of a resistive 
divider (formed by a fixed resistor and the sensor) connected between 
5V and ground.  Looking at the 747 schematics, the fixed resistor is a 
1K ohm.   From this, you can come up with a  temperature vs voltage 
table (using the CTS sensor data) with the formula:

                                                                     Rsensor
Voltage at a given temperature [Vt]  =  5V * --------------------------
                                                                Rsensor + 1000

Then, calculate the A/D counts (for an 8 bit converter) by:

A/D counts at a given temperature  =  51 * [Vt]

The magic number 51 is simply the A/D resolution (8 bits in the 
example) divided by 5V.  If you have a 12 bit converter it would be 819 
(4095/5).  

You can now plot this table using Excel, graph it, and simply visually 
extrapolate temperatures for A/D count values in between your 
calculated values from the graph.  It comes out as sort of an S-shaped 
graph, and comes very near to 255 counts at the low end, and around 
32 counts at 100C.  Note that the voltage value will never equal 0 or 5V, 
and the curve will be very nonlinear.   This makes it a bit difficult to 
come up with a formula unless you're real good at math...

If you feel adventurous, you could try using the best fit curve and 
equation solver functions in Excel to come up with an overall formula for 
temp vs A/D counts, but it would probably be a 2nd or 3rd order 
polynomial.

Hope this helps.  I can email you the spreadsheet if you want to reply 
off-list, or if there's more interest, I can post it to the website.


darrell n
Calgary, Alberta, Canada
----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org




More information about the Gmecm mailing list