Hey all, the stats. 747 ecm, $42 mask, BBC TBI 90 pph injector, 350 chevy, vortec heads.

My problem/question. Why does tuner pro rt read the temp scale backwards?? IOW, when I am datalogging, the temp says 22* C when the car is cold, when it warms up it goes to 70-73 C, (which is right according to the scanner) when warm?? Anybody else have this issue?? Just wondering if this could be affecting the tuning of my bin through tuner pro rt.

 

It sounds like either there is a problem with TunerPro or more likely, the ads file. In the ads file, for the temp conversion, is it a table look-up and interpolate or some formula? The more accurate way is to do the table look-up since the aldl datastreams the coolant temp in raw ad and raw inverse ad. If you plot out the ad you'll notice that no "simple" formula would cover the whole range let alone the points between freezing and overtemp!
EBL...

 

The temperature in the older ECMs is an inverse term based on the resistnance of the sensor. The raw A/D value is an exponential curve that rises as teh temperature falls. It doesnt display right in some programs due to the fact that you need to use a formula to interpret it.

The newer ecms/pcms have tables in them that are used to get the temperature from the raw A/D reading, so its easier to display the temp.

 

I don't think it's exponential. From all of the data online it appears to be a polynomial.
Inv.Temp = 2E-11x6 - 9E-09x5 + 9E-07x4 + 0.0002x3 - 0.0465x2 + 3.7387x - 33.383

 

IIRC, that is one of the ADS files I saw that was not set up correctly. It was displaying raw counts instead of the table values. The table is there, but the data "operation" just needs to be changed to "Use Lookup Table" and then point "Lookup" at the table. I've extended the 7747 ADS file I sent to Mark and Craig WAY beyond any other ads file I could find, along with fixing some conversions like this (as well as providing Counts, *F, and *C outputs for temp). Haven't heard yet whether they have posted it or not...

 

For a generic RTD its an exponential type of relationship. For an actual one in use they probably used curve fitting to get the relationship.

 

ok, I have to look at the ads file, maybe I could get my hands on the ads file that was modified?? I'm not REAL good with changing the way the source code interpretes things, YET. I'm a newbie at the prom tuning. But it would be nice to see the close to actual temp as the vehile is warming up because I still have some issues to figure and tune out in those areas.

 

I sent it to you in an email. Just remember, I'm also a newbie and AFAIK, that file has only ever been used by me, so "use at your own risk" and all that. Let me know if you find any errors, but I tried to be careful and test for valid values...

 

I just realized, I can attach it here, I just had to change the extension to bmp. Should be able to save the attachment and rename it to use it...

 

Use a lookup table.

 

It's not probably, it's really not exponential, I just checked myself.
The only formula that is exponential is the one that would convert ad inv to ad. ad to deg F (or C, don't matter) and ad inv to deg F is a polynomial.
Here is the deg F vs ad inv;I'd use a table look-up.

 

If it helps this is how the ECM 'sees' the CTS and IAT sensor. ADC counts (voltage) vs. temperature.

RBob.

 

What about the shunt resistor used to improve the accuracy (edit: I should say resolution, not accuracy) at both low and high temps? That should make it look like a sawtooth. I'm not sure exactly which ECUs have it, and I don't know how the ECU knows when the shunt resistor is in or not or what it does with that to get the normal A2D. Does anyone have that info?

 

well, I used the lookup table and had it converted to degree f and I works great now. thanx all. I appreciate the replies.