Spindle feedback at high spindle speeds
13 Aug 2011 06:03 #12469
by PCW
Replied by PCW on topic Re:Spindle feedback at high spindle speeds
So if you chose the MSB you will have one pulse per revolution, next to MSB would be 2 pulses per rev and so on, so you can just chose one that not too fast to count at 10K RPM (166 RPS)
Please Log in or Create an account to join the conversation.
13 Aug 2011 12:27 #12471
by andypugh
Replied by andypugh on topic Re:Spindle feedback at high spindle speeds
frequency wrote:
14 bits of io is rather a lot to "spend" on a spindle encoder, but it would be very easy to interface to EMC2 if you do have 14 spare input pins.The value is in a binary format with differential line drivers for each bit.
Please Log in or Create an account to join the conversation.
14 Aug 2011 08:12 #12478
by frequency
Replied by frequency on topic Re:Spindle feedback at high spindle speeds
But i think thats a good idea!
An cheap additional PCI parallel port card for the position information i think will solve my problems.
EMC then will always know the position of the spindle at low speeds and ignores it if the position is not needed at high spindle speeds!?
An cheap additional PCI parallel port card for the position information i think will solve my problems.
EMC then will always know the position of the spindle at low speeds and ignores it if the position is not needed at high spindle speeds!?
Please Log in or Create an account to join the conversation.
14 Aug 2011 11:39 #12483
by andypugh
Replied by andypugh on topic Re:Spindle feedback at high spindle speeds
frequency wrote:
Configured for input, a parallel port gives you input pins. You could simply not wire the LSB, and would have 2048 resolution of the spindle.
You could wire the individual bits into the HAL "weighted sum" component to get the position back to 0-2047 numbers. You would probably need a custom component to convert the numbers to a count. (ideally stored internally as 64 bits, to prevent wrap-around problems)
One thing a bit inelegant with this is that the parallel port driver reads in the port state as an integer, then cuts it up into bits, then the weighted-sum would put them back together as an integer again. A custom component _could_ read the data direct from the port register if you wanted to, and didn't intend to use the extra parport outputs.
An cheap additional PCI parallel port card for the position information i think will solve my problems.
EMC then will always know the position of the spindle at low speeds and ignores it if the position is not needed at high spindle speeds!?
Configured for input, a parallel port gives you input pins. You could simply not wire the LSB, and would have 2048 resolution of the spindle.
You could wire the individual bits into the HAL "weighted sum" component to get the position back to 0-2047 numbers. You would probably need a custom component to convert the numbers to a count. (ideally stored internally as 64 bits, to prevent wrap-around problems)
One thing a bit inelegant with this is that the parallel port driver reads in the port state as an integer, then cuts it up into bits, then the weighted-sum would put them back together as an integer again. A custom component _could_ read the data direct from the port register if you wanted to, and didn't intend to use the extra parport outputs.
Please Log in or Create an account to join the conversation.
Time to create page: 0.076 seconds