I can't test the X-series shift commands, but the w2 = w2 + w2 format (to shift left) appears to be about 10% faster than w2 = w2 * 2 and twice the speed of w2 = w2 / 2 (shift right). It's probably easiest to consider the denary equivalent: we want to divide say ABCD by EF but only have the capability of AB by EF. However, PICAXE does allow 16-bit by 8-bit divide ( and 16-bit by 16-bit ) and there are three overlapping groups of 16-bits in a 32-bit number.

Did you know that 100% cotton canvas marquees can insulate heat when it’s cold and retain a cooler temperature inside when it’s warm? Resilient and versatile, our extra-large 16ft x 32ft canvas marquees are rot-proof and made from white waterproof 500GSM cotton canvas. Changing belts is a breeze – you can remove a belt and install a different grit belt in about the same time it takes you to change the blade on a table saw (and a heck of a lot quicker than installing a blade on a bandsaw). I don't know if this may help or inspire anyone, but here's the code to calculate a 32-bit result from a 16-bit x 16-bit multiplication, "result = lhs * rhs", using only word and byte values. I am still hoping that someone will come up with a smart and compact long division mechanism that will to divide a 32 bit number by a 16 bit word where the values are set such that the answer will never be more that a 16 bit word.That's quite a timely request, I've just been drafting the explanatory text to go with the code, which I hope to post in the "code snippets" section later this evening.

pdf section 29) you will see that the instructions are for bit and byte manipulation with only add and subtract math functions so Rev Ed already have to impliment multiply and divide plus 16-bit math in firmware. I've been tidying up my above code, using word variables and a few extra lines to emulate an overflow/carry flag for the numerator.

Could you post the code of your cut down version as its finally worked out and I will try it in my run time application this weekend to check it out. ALLAN, I tried your code in my run time environment hoping it would just work , but got some unpredictable results which I have no time to look at just now.