View Full Version : Dual Screw Axis

11-24-2003, 11:29 AM
I am designing my second cnc router with a 60” wide gantry. I am using screw drive and feel that I need two screws to maintain proper alignment. I have seen machines on the net with two screws coupled by belts, this seems ideal, but b/c of the distance between my screws I am reluctant to use belts. I have also seen machines with two motors on the axis each driving one screw. I am using servos with geckos. If I tie the step and direction lines together will I see drift b/n the screws? I have heard many arguments against this b/c the motors will always be slightly fighting one another, but I have heard of many people doing it also. So.. Those who have done it, did it work well? Or am I about to make a huge mistake?

Tim T

11-24-2003, 01:28 PM
Hi Thielert,

I believe that to do this safely, will require a slave axis and a master axis. Servo motors rely on encoder feedback to work, so each motor will have to have its own encoder. For proper position monitoring, each motor must be driven as an independant axis, and the controller takes care of the master-slave aspect so that they move in unison. This is more or less just an opinion I have to offer. There might be some sneaky way around this.

11-24-2003, 03:12 PM
I think we are on a similar track. I am intending on using two servos /w encoders and each servo having it’s own gecko drive. I will take the step and direction lines from my break out board for the one axis and run them to both gecko drives.

I imagine that with each power up cycle I might see a single encoder stripe drift between the two servos. This would equate to 1.44 Degrees with 250 CPR encoders or .0004” on 10TPI screws. So, not a big deal, and random drift on each screw will keep them approximately zeroed.

Anyone have any experience, comments or anecdotes on this topic?
Tim T

11-24-2003, 03:15 PM
Considering quadriture encoders further, the drift would be event less than this, The phase shift, and the rise and fall of the pulse would allow only .0001” drift!
Tim T.

11-24-2003, 07:39 PM
I would think that small of a discrepancy would do no harm.

You'd need to know for sure that when a limit switch, overload, or encoder failure occurred, that both motors would stop at the same time. Homing might be a bit of a trick, too.

11-24-2003, 07:42 PM
I believe Mach2 is easily setup to run 2 motors on one axis, 1 Master and 1 slave. You can even set it up with 2 limit switches to automatically realign itself.


11-25-2003, 09:29 AM
Now that is a great bit of information Gerry.
It's almost as if they had this configuration in mind..

Tim T

CNC Brute
11-25-2003, 11:21 AM
From what I have read mach 2 is software?

I am also looking at running two threaded screws. Are there other options for tables over 5 feet wide? I couldn't think of any.

And I'm looking at the Axxus controller will that be able to control two servos?



11-25-2003, 11:43 AM
Probably the best solution when we get this big is to go with rack and pinion drive. There would no longer be a “shaft whip” issue and the pinions can be coupled across the table by a rod. I want to maintain the same drive ratios on my X and Y and can’t do pinion drive on my Y because I’m using a slide designed for screws, So I have to use a screw on my X also. So now I have a different bag of worms to sort out.

Tim T

11-25-2003, 11:57 AM
I think that the Axxus uses gecko drives, so yes, it probably can be set up for servos.
Or, you could buy the Axxus break out board, geckos, and a power supply to make you own.
The X and Y should be servos, for your Z you can use another servo, or a stepper. The Z doesn’t need much oomph so a nema23 will cover the bases.

This ebay auction is ideal. (http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2576218546)

The seller (me) can include a third servo for an additional $90.

Tim T