Author Topic: Measuring backlash on rotary table  (Read 451 times)

Offline Dell

  • Jr. Member
  • **
  • Posts: 87
  • Country: gb
    • anniversary-clocks.co.uk
Measuring backlash on rotary table
« on: March 26, 2024, 07:26:56 AM »
am motorising my rotary table using stepper motor and World of wardís controller and it has backlash compensation in the settings but how do I measure it , on the handle vernier itís just over 1 and a 1/2 notches but thatís not very accurate is it so I would like to be more precise .
Dell
https://www.worldofward.com/
Old man but still learning

Offline Muzzerboy

  • Full Member
  • ***
  • Posts: 125
  • Country: gb
    • The Mighty Shiz
Re: Measuring backlash on rotary table
« Reply #1 on: March 26, 2024, 01:21:24 PM »
Can you not adjust out (or at least reduce) that backlash? Many / most rotary tables have an eccentric adjuster on the wormwheel for this purpose.

Backlash adjustment in software assumes the table is driven against a friction-like drag load that resists angular movement but you may well have a machining load that biases the wormwheel in the other direction, in which case the anti-backlash may actually double the error when you change the direction of movement.

To measure the angular error, perhaps you would want to lock the table and then apply a controlled torque to the wormwheel in each direction. It's not a perfect science, since there is inevitably a degree of spring so that the backlash will depend on the torque you apply. If you applied no torque but simply moved the wormwheel in alternate directions, you'd get a smaller backlash - but not necessarily what you will see with a real world load applied.....

Offline philf

  • Hero Member
  • *****
  • Posts: 1113
  • Country: gb
Re: Measuring backlash on rotary table
« Reply #2 on: March 27, 2024, 01:42:50 AM »
Another possible problem is that the backlash could vary over the 360 degree movement. In that case I doubt backlash compensation would be much use. As Muzzerboy has suggested have you tried to adjust the meshing of the worm and wheel to reduce the backlash to a minimum?
[On my Vertex dividing head it's impossible to get rid of backlash. The wheel must be slightly eccentric as trying to minimise the backlash at one point will mean it can lock solid at another]
« Last Edit: March 28, 2024, 07:51:29 AM by philf »
Phil Fern
Location: Marple, Cheshire

Offline Noitoen

  • Full Member
  • ***
  • Posts: 143
  • Country: pt
Re: Measuring backlash on rotary table
« Reply #3 on: April 05, 2024, 08:04:49 AM »
With that controller, you don't really have to know/worry about the backlash. This is only important when you reverse the direction of movement. You program your favorite direction in the parameters and work always in that direction. When you reverse, the controller will go past the desired position by the amount of steps programmed and re-approach the final position on the correct count. If in doubt, program a high value and it will be ok.

Offline Muzzerboy

  • Full Member
  • ***
  • Posts: 125
  • Country: gb
    • The Mighty Shiz
Re: Measuring backlash on rotary table
« Reply #4 on: April 05, 2024, 08:25:58 AM »
With that controller, you don't really have to know/worry about the backlash. This is only important when you reverse the direction of movement. You program your favorite direction in the parameters and work always in that direction. When you reverse, the controller will go past the desired position by the amount of steps programmed and re-approach the final position on the correct count. If in doubt, program a high value and it will be ok.

It depends what you are doing with the table. If you are simply positioning a work piece eg to drill holes in a plate, the friction of the movement combined with the backlash compensation will probably eliminate most of the backlash - although any non axial machining load may move the table within the unconstrained backlash angle, as there is nothing actually holding the table from moving. But worse than that, if you are milling a feature in such a way that the cutting forces try to move the table further in that direction, you may actually double the total backlash by applying "compensation".

A similar-ish issue arises on linear axes when you transition from conventional to climb milling, which can happen simply by increasing the optimal load (width of cutter engagement). The forces on the ballscrew are reversed, yet the compensation isn't. You can feel this effect for yourself in manual milling in "conventional mode" - there comes a point where increased engagement results in negligible feed force being required and further engagement beyond that point results in the work "self feeding" or pulling itself into the tool. At that point if you have backlash, it will result in a sudden step feed into the tool as that backlash is suddenly taken up, which can be disastrous for the tool.

The bottom line is that you are much better trying to minimise / eliminate the mechanical backlash in the first place rather than hoping to rely on some "clever" software to magically dial it out for you.


Offline philf

  • Hero Member
  • *****
  • Posts: 1113
  • Country: gb
Re: Measuring backlash on rotary table
« Reply #5 on: April 06, 2024, 03:50:49 PM »
Dell,

To determine whether you could even consider backlash compensation I'd try the following:

Move the handwheel about a tenth of a turn. Check the backlash (using the divisions on the handwheel)and repeat until you've done a full revolution. If you get variations in backlash then your worm is running eccentrically.

Next revolve the table a tenth of a turn (whole numbers of turns of the handwheel) and note the backlash. Repeat for one revolution of the table. If there's a big variation then your wheel is running eccentrically.

You may find that at some point there's zero backlash and in that case there's no way of adjusting any of it out - without machining!

As has already been said you probably needn't worry about the backlash compensation but, if there are inconsistent results from the above tests, you'd probably introduce more problems than you solve.

Phil.
« Last Edit: April 06, 2024, 04:18:03 PM by philf »
Phil Fern
Location: Marple, Cheshire