Need help with scaling

[champion]

I guess what I'm asking is how to take the 0 and 1 bit from one number and put them into a seperate number so that I can further scale just those two bits. Thanks!!!

[SteveB]

I guess what I'm asking is how to take the 0 and 1 bit from one number and put them into a seperate number so that I can further scale just those two bits. Thanks!!!

This will give you a value with just the first two bits of the original byte:

TwoBits = OrigByte & %00000011

HTH,
Steve

[SteveB]

Here may be a more complete answer to your question:

PotValue  VAR BYTE
Degrees   VAR WORD
Minutes   VAR WORD


Degrees = Potvalue * 14118    ' Convert to degrees and fractional Degrees
			      ' scaled by 10000.
	' It does not matter that we assign Degrees a value before DIV32,
	' since DIV32 uses only internal PBP variable in the calculation
	' It is important that DIV32 follows right after the Multiply
			      ' 
Degrees = DIV32 10000         ' Returns degrees
Minutes = R2                  ' Returns the remainder of the division,
			      ' which is fractional degress scaled by 10000
Minutes = (Minutes * 6)/1000  ' returns minutes

Here are a couple of nice threads by Darrel Taylor dealing with exploiting PBP's integer math:

Retrieving 32bit Multiply Result
Retrieving Div32 Remainder (This one explains "Minutes = R2". See the last post in the thread)
32-bit Variables and DIV32, Hourmeter 99999.9

Steve

[SteveB]

It seems a little over-kill to be getting minutes when the accuracy of your measurements is not even 1 degree. That, and a pot is also not likely to give you very consistant accuracy as well. Even if you used 10 bits, you are still looking at a LSB of .35 degrees (21 minutes).

Just my 2 cents,
Steve

[champion]

Thanks again for all of your help. I have tried this, and I get a two bit result, however they are not the last two digits of the potvalue * conversion. I will do more research inside the threads you listed to figure out exactly what these two bits are representing. Thank you very much for your help, and I will be sure not to forget it when/if you ever have a question in the future.

[champion]

Steve, actually I am trying to have accuracy down to 1 minute. I have come to the conclusion however, like you say, that a pot will not have the repeatability that I will need to take on this task. I have moved on to an optical encoder with 8192 pulses per revolution and a gear ratio of 3:1. Thanks for all of your help.