On Tue, 1 Mar 2005, James Newtons Massmind wrote:

> Too cool of you to post this! Thanks!
>
> In the language you are using here, is 0 or 1 considered true? Also, what
> does the "tag[j]" on the line after the print " $j" do?

It is a pseudo language and my rendering of it is full of 
inconsistencies (like using bin(i) and tag[i], both being array 
references).

> If we are not to "add anything for open reading or zero reading" and we are
> "using -1 as reading for an open reading" then "R (i+1lj+1) != +inf" needs
> to be something like "(R (i+1,j+1) != +inf) && (R (i+1,j+1) > 0)" correct?

assuming +inf is represented as -1 as I proposed, make that R(i+1,j+1) != -1

> In javascript, I did
> 			r = R(i+1,j+1);
> 			x = ( isNaN(r) ? 0 : parseFloat(r) );
> 			bin[i] += (0<x ? x : 0 );
> Where isNaN returns true if r is not a number so x becomes 0 if r is empty
> and the numeric value of r otherwise.
> bin[i] is incremented only if x is greater than 0.

Yes, that is the idea.

> "slice = (min * 11) / 9" seems to work better than "slice = (max + min) / 2"
> but why not just "if (bin[i] <= slice) " rather than "if (bin[i] < slice) "
> since this also managed the case where min == 0. Then slice can just be set
> to min. Or am I missing something there?

min is not supposed to be 0, unless there is a short. Getting users to 
take meaningful readings for < 1 ohm windings with a $5 DVM will be 
interesting anyway. The idea of slice is that the measured values will 
be all around R and 2*R, where R is the approximate winding resistance. 
Then after min/max slice will be approximately R+2*R/2 = 3/2*R and the 
farthest from either group's strays (iow poor measurments) (I hope).

> I also added some checks for coil resistance readings that vary from one
> phase to the next and for phase to phase resistances that aren't double the
> coil resistance.
>
> The javascript/web page translation is at
> http://www.piclist.com/techref/io/stepper/wires.htm and seems to work up to
> 6 wire motors.
>
> This still doesn't find the ORDER of the phases, but I can't see a way to do
> that with out applying power or using a dual trace 'scope.

That is correct, you cannot find the order easily, but I have something 
in mind that involves a $1 magnetic compass and some thin wire coils 
wound around it ... let it sit for a while it isn't just there yet.

Peter
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist