Code: EGR and CTS conversions

[1project2many]

Who knows. I'll have to take a look and see what's going on. The PC was locked up the other night so I lost a bunch of work in the programs that were open.

[EagleMark]

Nevermind, graph doesn't like it either.

The equation 1P2M came up with plots properly in the graph. Don't know what TP problem is.

(-.0315x + 4.5)(-.0315x + 4.5)(-.0315x + 4.5)+(x*x*-1*.00045)-.135x+73.00000

Just put that in and got 64.88c?

[EagleMark]

EDIT: I will get the center of the curve as accuate as I can and to worry too much about the extreme ends i.e. above 150C and below 0C.

OK, this one gets it within about 5C all the way from -30 to around 95C...I THINK! And it is the format TP 'should' recognize?

((-.055x+6.5)(-.026x+4.6)(-.024x+4.0))-((.049x+1)(.028x-2.3))+.04x+53

Tried this one and got 56.36c ?

[EagleMark]

Nevermind, graph doesn't like it either.

The equation 1P2M came up with plots properly in the graph. Don't know what TP problem is.

(-.0315x + 4.5)(-.0315x + 4.5)(-.0315x + 4.5)+(x*x*-1*.00045)-.135x+73.00000

Was missing a "*" then it came out the same 35.59C

[1project2many]

So it's good now?

[EagleMark]

No, I just tried Gregs and got a different number, but he said yours was correct. I didn't understand how I got a different number? But his was missing a *.

Still

Verified hex at $DA and decimal 218 which is 10C on AD count chart. But temp in TunerPro was 35.59c ?

[1project2many]

Ok, then. See previous post. ;)

[gregs78cam]

I think we are dealing with an 'order of operations', or something that TP doesn't like to handle. In the graph these all work out the same. If you follow this link and then click on the "trace" button on the graph you can click anywhere to see the (x,y).

(-.0315x+4.5)^3+(-.00045x^2)-.135x+73.

(-.0315x + 4.5)(-.0315x + 4.5)(-.0315x + 4.5)+(x*x*-1*.00045)-.135x+73.00000

(-.0315x+4.5)^3+(-.00045(x*x))-.135x+73.

http://fooplot.com/#W3sidHlwZSI6MCwi...IsIjIxMCJdfV0-

[Mangus]

There are no implicit operators in TunerPro like there is in hand-written algebra. You're missing a few * operators. For instance, this:
(-.0315x + 4.5)(-.0315x + 4.5)(-.0315x + 4.5)+(x*x*-1*.00045)-.135x+73.00000

Needs to be

(-.0315 * x + 4.5) * (-.0315 * x + 4.5) * (-.0315 * x + 4.5)+(x*x*-1*.00045)-.135 * x+73.00000

[EagleMark]

Thanks Mark.

I put the new conversion in the footplot link and it too worked.

But when inserted to TunerPRo Parameter Conversion it changed calculated to 0.00 and hex to 80h. Restart TunerPRo and hex was 66, change to calculated and went back to 0.00, change to hex and again it was 80h. Before it was off, not it just freaks out?...

[Mangus]

Thanks Mark.

I put the new conversion in the footplot link and it too worked.

But when inserted to TunerPRo Parameter Conversion it changed calculated to 0.00 and hex to 80h. Restart TunerPRo and hex was 66, change to calculated and went back to 0.00, change to hex and again it was 80h. Before it was off, not it just freaks out?...

Most likely a result of this being a multi-root equation. I haven't plotted it to see what it looks like, though. TunerPro converts from the calculated value back to the raw value stored in the bin by doing a binary search. If there are multiple points in the graph where that binary search results in a correct answer, it may land on a different one each time. It could also be an error in the equation still. Is that last ")-.135*x+73" correct? Should there be another * there, or do you mean to subtract 0.135 * X from the result of the previous results? Order of operations in TunerPro are respected. (Edit: just looked at the graph, and it should work. I'll have to take a closer look, but also, don't forget about the number of decimal places. This is a cubed function from a low-resolution binary input - you might try increasing to 3 or 4 decimal places in TunerPro.)

[EagleMark]

Here's a link from above for the plot, it's a common AD count to celcius from many dissasemblies.

Tried increasing the decimal places in TP but same results, temp in calculated is 0.0000c and right click to show hex changes hex to 80h from 66h

Greg or Shannen will be around shortly to check the math you corrected for TP.

[RobertISaar]

all of this work, just because GM chose to compare against an A/D count instead of a linearized variable....

[1project2many]

TunerPro converts from the calculated value back to the raw value stored in the bin by doing a binary search. If there are multiple points in the graph where that binary search results in a correct answer, it may land on a different one each time.

Could you explain this? Does this refer to the mechanics of committing a value calculated in the "Edit parameter" window to the .bin?

all of this work, just because GM chose to compare against an A/D count instead of a linearized variable....

True, but... All this work is really about trying to make a linear equation for an "Aggressively non-linear" sensor. The actual equation to calculate temp from resistance uses the natural log e raised to a third order polynomial. I honestly don't even know how to code that for an HC11. I can see where a lookup table is an acceptable solution.

Tried increasing the decimal places in TP but same results, temp in calculated is 0.0000c and right click to show hex changes hex to 80h from 66h

I didn't check this thread before I worked out the formula. I noticed a couple of things that TP didn't like. Using a negative coefficient such as -.0315 caused problems. Breaking -1 out as a separate value caused TP to interpret coefficients less than one as zero. So .0315X became 0. TP didn't even recognize X as a variable in the "Variables" window. And as previously mentioned, the implied operators needed to be made explicit, .0315X needed to be .0315 * X. FWIW I did that with Greg's first formula but for some reason it seemed like it wasn't needed the second time. Silly me...

Anyway, this one's working on my pc. Lowest displayed temp should be -34.77 and highest should be 164.13 $DA = 8.92 deg C.

(-1*.0315*X + 4.5)*(-1*.0315*X + 4.5)*(-1*.0315*X + 4.5)+(X*X*-1*.00045)-(X*.135)+73.00000

If TP works with logarithms I might be able to get the actual Steinhart-Hart equation into play.

[chevmasta]

I have zero idea what you guys are talking about.....

but this is why I love this forum. Keep it up! I'm sure I'll benefit in the end!