03whitegsr

iTrader: (8)

I've tried to avoid writing any of my own scaling as I've been a little unsure about a few things.

First off, when you write a scaling parameter, do you need have your "inc" value be a multiple of your scaling? (i.e. 1 bit increment = 1 increment in number form)

Secondly, is the displayed value a truncated value or a rounded value?

Third, why are most toexpr/frexpr reciprocals but a few others are not? Primarily the AFR and InjectorScaling (that was part of the xml from the v1.4 of ECUflash) where the toexpr and the frexp are identical.

Finally, do the max/min values actually limit the range you can input or is it just used for the map color scaling?

l2r99gst

iTrader: (2)

I knew the answers to all of these, but would have to double check and don't want to give the wrong information.

But, for your #3:

The to and from expressions are just relationships of the data in the ROM to your display and then from the display back to your ROM.

Any 1/x relationship is going to be the same to and from.

For example, y=1/x
To find in terms of x, x=1/y

If your to display is x-40, your from is x+40
y=x-40, x=y+40

That might help you think about it a bit easier.

03whitegsr

iTrader: (8)

Ha, yeah, now that I just thought about it, it's pretty obvious that it's right. It just didn't look right as most others are an inversion because as you said, they often have offsets. Where that particular one already is a reciprocal...

I need to go through all my scaling values and make sure they convert back and forth correctly.

03whitegsr

iTrader: (8)

One additional question, what is the difference between "int8" and "uint8"?

int8 seems to turn bit values >128 into negative numbers with 255=-1 and 129 = -127

uint8 seems to just be 0-255

Deepnine

iTrader: (4)

unit = unsigned integer = only positive numbers from 0-254
int = integer = positive and negative number from -127 to +128

codgi

iTrader: (22)

What Deepnine said. Signed ints using 2 complements representation:

http://en.wikipedia.org/wiki/Two%27s_complement

That's the phenomena you are seeing above. You can use "calc" to see it better.

Put in -1 in Decimal mode and change it to binary, you get a row of 1s right? Now put in 255 also in decimal mode and change it to binary another row of 1s right? We tell the computer how to interpret bits so depending on the interpretation that row of 1s can mean 255 or it can mean -1.