Spindle encoder and spindle gearbox backlash
20 Mar 2019 07:09 #129044
by Henk
Spindle encoder and spindle gearbox backlash was created by Henk
HI
I am in the process of retrofitting a Deckel FP5 Milling machine. All the major things are complete, but I now have a challenge with the spindle encoder. There is no space in the gearbox to mount an encoder or proxys that will give me sufficient resolution and a 1x pulse. The gearbox have 8 speeds.
The only option I have is to mount a encoder on the gearbox input shaft.
I think I will be able to adjust the encoder scale according to which gear is selected, but I am concerned about the backlash in the gearbox. I will also be faced with the issue of multiple z-pulses whenever the gear selected is not 1:1.
So my question is:
1. Would it be possible to compensate for the backlash in the gearbox so that it doesn't break taps when rigid tapping (on reversal)
2. If the gear selection is say 4:1 reduction, can I somehow :discard" 3 of the z pulses and make G33 only react on every 4th one so that it simulates 1x spindle rev?
If this is possible, what would be the best approach to do this?
Thanks for your assistance.
Henk
I am in the process of retrofitting a Deckel FP5 Milling machine. All the major things are complete, but I now have a challenge with the spindle encoder. There is no space in the gearbox to mount an encoder or proxys that will give me sufficient resolution and a 1x pulse. The gearbox have 8 speeds.
The only option I have is to mount a encoder on the gearbox input shaft.
I think I will be able to adjust the encoder scale according to which gear is selected, but I am concerned about the backlash in the gearbox. I will also be faced with the issue of multiple z-pulses whenever the gear selected is not 1:1.
So my question is:
1. Would it be possible to compensate for the backlash in the gearbox so that it doesn't break taps when rigid tapping (on reversal)
2. If the gear selection is say 4:1 reduction, can I somehow :discard" 3 of the z pulses and make G33 only react on every 4th one so that it simulates 1x spindle rev?
If this is possible, what would be the best approach to do this?
Thanks for your assistance.
Henk
Please Log in or Create an account to join the conversation.
20 Mar 2019 13:31 #129056
by pl7i92
Replied by pl7i92 on topic Spindle encoder and spindle gearbox backlash
the encoder is to be placed on all surcomstances direct to the spindle
Please Log in or Create an account to join the conversation.
20 Mar 2019 14:08 #129067
by Henk
Replied by Henk on topic Spindle encoder and spindle gearbox backlash
Unfortunately that is not possible in this case. Especially not for the horizontal spindle, which I forgot to mention.
Please Log in or Create an account to join the conversation.
20 Mar 2019 15:50 #129076
by PCW
Replied by PCW on topic Spindle encoder and spindle gearbox backlash
For the index issue, can you mount a proximity sensor near the spindle to sense a notch or stud? This could either be used as the index itself or as a index mask for the encoder counter
The backlash is more troublesome, you might be able to create a backlash simulator component to feed the encoder count through, though the output of said component would have infinite acceleration when the simulated backlash was taken up and might need some filtering so it did not generate following errors in Z...
The backlash is more troublesome, you might be able to create a backlash simulator component to feed the encoder count through, though the output of said component would have infinite acceleration when the simulated backlash was taken up and might need some filtering so it did not generate following errors in Z...
Please Log in or Create an account to join the conversation.
21 Mar 2019 14:02 #129171
by andypugh
My milling machine is not an FP5, so this might not apply.
As my machine has a 1:1 ratio in the vertical head I am able to work with an encoder that is only on the horizontal spindle.
I mounted it at the back of the machine, where the draw-bolt is accessed.
photos.app.goo.gl/PTVhKU46ghp3iYa29
It is driven by a set of nylon gears, but a belt would also work.
More info and pics n CNCzone: www.cnczone.com/forums/vertical-mill-lat...c-2.html#post1212335
Replied by andypugh on topic Spindle encoder and spindle gearbox backlash
Unfortunately that is not possible in this case. Especially not for the horizontal spindle, which I forgot to mention.
My milling machine is not an FP5, so this might not apply.
As my machine has a 1:1 ratio in the vertical head I am able to work with an encoder that is only on the horizontal spindle.
I mounted it at the back of the machine, where the draw-bolt is accessed.
photos.app.goo.gl/PTVhKU46ghp3iYa29
It is driven by a set of nylon gears, but a belt would also work.
More info and pics n CNCzone: www.cnczone.com/forums/vertical-mill-lat...c-2.html#post1212335
Please Log in or Create an account to join the conversation.
21 Mar 2019 14:50 #129180
by Henk
Replied by Henk on topic Spindle encoder and spindle gearbox backlash
Thanks Andy.
Ons the deckel, the spindle to be used is automatically engaged/disengaged when the vertical head is swiveled up or down again when changing between horizontal or vertical.
On my FP4nc I mounted an encoder driven by a belt just below the drawbar but that doesn't work when using the horizontal spindle.
My idea was to write a userspace comp that intercepts the encoder position and velocity signals that would normally be netted to motion.spindle-position and ....velocity. The comp would then compensate for the gearbox backlash and output new position and velocity signals to be connected to motion.spindle.......
The index pulse similarly could be filtered in a base thread comp so that multiple z pulses are masked and only one per rev of the spindle is let through,depending on which gear is selected of course
Do you think something like this can work or do you think it will be a waste of time trying.
Thanks
Henk
Ons the deckel, the spindle to be used is automatically engaged/disengaged when the vertical head is swiveled up or down again when changing between horizontal or vertical.
On my FP4nc I mounted an encoder driven by a belt just below the drawbar but that doesn't work when using the horizontal spindle.
My idea was to write a userspace comp that intercepts the encoder position and velocity signals that would normally be netted to motion.spindle-position and ....velocity. The comp would then compensate for the gearbox backlash and output new position and velocity signals to be connected to motion.spindle.......
The index pulse similarly could be filtered in a base thread comp so that multiple z pulses are masked and only one per rev of the spindle is let through,depending on which gear is selected of course
Do you think something like this can work or do you think it will be a waste of time trying.
Thanks
Henk
Please Log in or Create an account to join the conversation.
21 Mar 2019 15:08 #129185
by Henk
Replied by Henk on topic Spindle encoder and spindle gearbox backlash
For a bit more clarity on why I want to do it this way .....
The only convenient place to mount an encoder is on the gearbox input shaft that is belt driven by the spindle motor.
In the highest gear, the speed ratio is 1:1 between this shaft and the spindle. In the lowest gear it is something like 6:1.
This will result in 6 index pulses if the lowest gear is used.
The backlash through this gearbox is also too much for rigid tapping which is why I would like to compensate for it.
Henk
The only convenient place to mount an encoder is on the gearbox input shaft that is belt driven by the spindle motor.
In the highest gear, the speed ratio is 1:1 between this shaft and the spindle. In the lowest gear it is something like 6:1.
This will result in 6 index pulses if the lowest gear is used.
The backlash through this gearbox is also too much for rigid tapping which is why I would like to compensate for it.
Henk
Please Log in or Create an account to join the conversation.
21 Mar 2019 15:30 #129188
by andypugh
Replied by andypugh on topic Spindle encoder and spindle gearbox backlash
The first question is. how much backlash is there? Is it enough to cause a problem?
Maybe a floating tap holder is the pragmatic answer?
The index might not matter, as long as you don't plan on doing peck-tapping.
Maybe a floating tap holder is the pragmatic answer?
The index might not matter, as long as you don't plan on doing peck-tapping.
Please Log in or Create an account to join the conversation.
Time to create page: 0.081 seconds