The problem with stepper drives, is that you may well put out the pulse to the motor from a controller card, it may have even turned, but due to toothed belts, worn lead screws, inertia, and anythink else you can think ,of the table may well not have achieved the requested position.

Therefore a better system is to take feed back via linear encoders that tell ' Mach3' where the table is and then Mach 3 can then decide how many extra pulses it should supply to get to the required position i.e. closed loop control. 

Can Mach3 do this

Can Mach 3  drive servo motors rather than steppers ??

I am the new guy on the block so appologies if this has been discussed already and I missed it.

Les C

Hi Les - welcome

I'll take your points in reverse order.

Yes it has been discussed MANY times here.

Yes Mach can drive servos. Mach puts out step/dir signals - it neither knows nor cares what's attached. Note for servos the loop is closed at the drivers NOT at Mach.

No Mach does not have closed loop control - even if it did it would not work by sending out "extra" pulses.

That is NOT a problem with steppers it is the problem of a poorly designed or maintained or badly implemented system. On a well designed system where the steppers are sized and driven correctly they NEVER miss a step.

Hey Les, welcome to the forum.

Mach is not closed loop.  you can feed positional data back to Mach and have it monitor the desired/actul position with allowance and have mach throw a fault if it gets out too far.  You can run servos with Mach.  Some hardware will allow you to run analog input drives (CS labs), others will only allow step and direction (smoothstepper).  I think most that do what your wanting is let the drives keep track of the position feed back and set the drive to send a fault signal to Mach if it gets out more than their following error will allow.

Brett

Thank you, Chaoticone and Stirling for your responses they are well received.  I have been using a Seig 3 miller of recent times with Mach 3 software and found a number of short comings of the machine (not the software) , which I won't waste your time on here.

I have now purchased an Alcock and Shipley miller ( Bridgeport purchased them I beleive )  and want to convert the machine from purely manual operation to a CNC Mill.

Do you know of anyone supplying a decent kit,  or any knowhow which may help me in my task. Instictively I would go for the servo and close loop to the drivers with error's being reported to Mach 3 option, but what would you go for ??

I have a resonably open mind on this but am concerned about worn leadscrews, backlash etc introducing error's into wok pieces that a half decent machinest would take out / not allow. Hence my stance on knowing where the table is and not on how many pulses went to a stepper motor.  I also don't want to incur a lot of expense on refitting the machine with new leads screws etc, but will do if it becomes paramount to a successful system.

Les, I understand what your saying about replaceing screws and an operator being able to take out backlash but on a cnc backlash is a huge no no to get the benefits of cnc.  The difference between the operator and any control driveing the machine is timeing.  I havent seen any manual operators be able to interpet a 3 inch circle with a 1/2 inch cutter at 30 IPM.  Nor have I seen any controller be able to take any amount of backlash out at any speed and still get flawless results.  Maybe acceptable, but never flawless.  The big question is how right does it have to be for you?  Backlash will cause you a lot of grief and should be prioriety one I think.

I'm partial to servos myself but I have a Bridgeport clone running steppers.  It will never miss a step unless I ask it to do something it simply can not.  If I had any intentions of putting an operator (button smasher) in front of the machine I would consider Servos a must.

I had Rockford Ball Screw make my scews.  A Bridgeport kit wouldnt work in my knockoff.    I think McMaaster Carr had a bridgeport ballscrew kit.  Google will find you lots of options most likely.

Advise, drive the quill for the z rather than the knee.  I did my knee and its painfuly slow.  I will keep it for tool height comp. and drive the quill for the z at some point.

Brett