OK, my RF30 mill moves great with no load but, I am tired of getting %70 through a 25min real job then having the motors skip steps killing $50 worth of stock...
When I use backlash correction it confuses the issue more...

Like I said above, air cutting runs the job fine and going back to 0,0,0 is almost dead on, but add any load and its crap shoot...


Can I add encoders directly to the table?
I would like to:
- Track the X,Y, & Z actual positions via encoders.
- Keep the motors, drivers, and ESS if possible.


I am sort of new to this stuff but not afraid
I have, RF30, Mach3, ESS, wanati drivers, wanati steppers. (425 in/labs)
So I assume someone has done this?
Would be great to see video or step-by-step tutorial of this being done...

Thanks for you help.

Encoders won't keep you from losing steps, they'll only let you know after it's happened.  You'd be much better off fixing the actual problem, which is one or more of:

1) Your steppers are too small for the job, and don't have enough torque
2) Your stepper drivers are either not driving enough current, or your voltage is too low, preventing the motors from reaching full torque at speed
3) Your motor tuning in Mach3 is too aggressive - you need to reduce velocity and/or acceleration

With steppers, you can't test the maximum no-load performance, then use that in actual machining.  Whatever velocity and acceleration works with no load needs to be reduced 30-50% for actual use.  Rather than wasting time and money fitting encoders, fix the drives.  Stepper drive systems will operate 100% reliably when properly designed and operated.  If you're losing steps, then, by definition, you have either under-designed the system, and/or you are attempting to operate it beyond its capabilities.

Regards,
Ray L.

Thanks for the reply.
I agree with what you are saying about stepper useage, and I know it won't prevent lost steps.
Most people I have talked with run 425 oz/in motors on this mill with little or no issue at faster feed rates.

I "assumed" the motors were just to weak, So I went bigger.
I have moved up to  570 oz/in, and raised the current to the maximum these drivers will allow. (Slightly more than the motor lists)
I also have torn the system down, cleaned,lubed and adjusted until it is very easy to move, with no binding. (I even reduce some of the backlash.

But...
The unpredictability of the skipping combined with with the backlash of this mill adds to the mess. (I have not switched to ball screws yet still researching)
So even when it does not skip, the backlash is an issues even if I adjust the setting till I'm blue in the face.
My backlash compensation is great for some jobs like drilling etc. but radius pockets etc, no way.
Actually I get much more consistant milling results with backlash compensation off...


Summary:
I have motors I think are big enough.
And the only way to know where the matrial is really located, is to drive to a position measured off of the table and not count steps.

Did I make any sense?
And will mach3 drive a stepper to a position based on an encoder value?

Cheers

Mach is open loop so can not accept encoder feedback to use for positioning.
You can fit encoders if you want but they will only be used for your information. There is a board sold that can take encoder inputs and if a user set error is detected Mach will be faulted. Will that save your work? depends on how big the error setting is, how fast the axis is moving, how much of an error will render your work damaged.
Hood

Thanks Hood.

I did not realize Mach was only open loop, I was thinking it could do closed loop also...
Based on Mach being open loop, it seems like a chasing my tail to track it.

Maybe I just need to go overkill on the motors and put ball screws in...
OK, who has a step-by-step how to select the correct ball screw setup for a RF 30...


Randal

What is the inductance of your motors, and what supply voltage are you running?  If the motors are high inductance, and/or your supply voltage is too low, you'll never reach rated current at-speed, and therefore never reach rated torque.  Since you indicate others get good performance with the same size motors, and the same machine, you should be looking at what you did wrong, and fixing that, rather than throwing money at work-arounds.  NO controller can make a bad drive work.  Once a stepper can no longer keep up with the load, there is no "fixing" it, whether you have encoders or not.  If they can't produce enough torque to handle the load, you're screwed, and the only way to stay on-course is to slow down, which NO controller can do on-the-fly, as the trajectory is planned far in advance of being executed, and cannot be changed AS it's being executed.

Regards,
Ray L.

That is a good point, maybe the motors are not at full torque.  (More power )

The motors are 6.4 mH/phase and are rated at 3.5 Amp/phase powered by a 36 Volt 10 A supply.
My drivers are rated up to 50 VDC. 4.2 Amp peak.

I am also running at 1000 Pulse/rev (down from 1600) in an attempt to trade accuracy for power.
Unless I miss understand the concept,  less steps equals more power but less accuracy.
That said 1000 Pulse/rev still should still hold .00025 on my setup.

Maybe a 48 volt supply would help out, just don't want to bake the drivers. 

Randal

OK, who has a step-by-step how to select the correct ball screw setup for a RF 30...

Have a look at any of the ball screw manufacturers web site technical information on how to select a ball screw which is appropriate for your machine.
You will need to make a decision on the grade of ball screw, pitch, size. etc, and how to properly mount them relative to the load, size of machine, and anticipated
axis speed. Everythng acts as a "system" in the end so there are trade offs to be made.

RICH

That is a good point, maybe the motors are not at full torque.  (More power )

The motors are 6.4 mH/phase and are rated at 3.5 Amp/phase powered by a 36 Volt 10 A supply.
My drivers are rated up to 50 VDC. 4.2 Amp peak.

I am also running at 1000 Pulse/rev (down from 1600) in an attempt to trade accuracy for power.
Unless I miss understand the concept,  less steps equals more power but less accuracy.
That said 1000 Pulse/rev still should still hold .00025 on my setup.

Maybe a 48 volt supply would help out, just don't want to bake the drivers. 

Randal

And there's your problem - those are high inductance motors.  Proper voltage for driving those motors is 80V, so you're getting will under half the rated torque out of them.  So you can either change the power supply and drivers, or get better motors (and you'll still need a higher voltage power supply), and solve your problem properly.  Nothing else will make that machine perform reliably.  Changing the steps/inch will have essentially no effect.

Regards,
Ray L.