[ynneb]

Hey Frans, I never could work mine out mathematically, so I went for the trial and error method.
Draw 3 lines on your table, one at zero and one at 1000 mm and one at 2000mm ( As I know your table is at least this long)

Then zero your axis at the zero line.
Set the steps per unit at a rate that you know would be less than it should be. For example you could set it at 10,000 steps per mm of travel.
Then drive your axis 1000mm buy typing in g0x1000 (replace the x in this code with the letter of the axis, x, y, or z. )
Keep you finger near the emergency stop button and send the axis moving. If the axis passes the 1m mark then activate your emercency stop. This means you will need to increase the steps per unit in your motor tuning set up. If it doesnt reach the 1m mark then you will need to decrease the steps per unit. If for instance the travel is only 500mm then that means you will need to halve your current steps per unit setting.
You will need to do a few tests of different settings until you get it to travel the correct distance.
Once it is about right send your axis down to 2m and make any fine tuning in the steps per unit setting that is required.

I know this method does not sound as good as doing it mathematically but in some ways it is more accurate. Lets say your pulleys are not exacltly made to what you think they are, then all the maths you use will never get it right. The trial and error method gives real world results.

I know it sounds like a lot of work and fussing to do, but just remember you will only have to do it once. Then write down and record the appropiate figures if you ever need to set up mach2 again.

Caution. Because you are using a 27:1 reduction, you may find that you need to use the step multiplier switches on your Rutex cards. If you do switch them then you will have to reset your steps per unit in mach2. Do not just flick those switches and then send you axis travelling. It will result in the axis over traveling and crashing.

EDIT: Because I know you are using the same sized encoders that I use and that I only use a 3:1 reduction and it works ok, I would suggest that you use the step multipliers so that they multiply 4times. Any resolution greater than this would be serious overkill for what you are trying to do, and may result in the axis been driven too slow as a result of Mach2 not been able to keep up. Just a thought.

EDIT: The way to know deside on the amout of step multiplying, is to descide on the acceptable resoltion you require. For cutting timber as you are planing to do, I would think that 10 steps per mm should be more than adequate. That will mean you can cut down to 1/10 of a mm. It is highly possible that your machine does not have those sorts of tollerances built into is anyway, so picking a even higher resolution is just a waste. With this in mind, you work out what step multiplying you need in oreder to give you a figure that will drive that axis at a close number to 10 steps per mm.

I hope all this makes sence?