Spindle control woes
21 Nov 2018 11:07 #121113
by Nitram
Replied by Nitram on topic Spindle control woes
Thanks with all you insight on this Andy.
In short, a negative command does not run the vfd in reverse after initial establishment of sign for direction. As the 7i76e spindle output is an analogue wiper signal it can never be less than zero, with direction ultimately being commanded by the direction pin.
I have a few things I would like to check as well, specifically why the feedback velocity is jumping around like it does. I wonder whether I have an index-enable signal/pin hard set true somewhere causing the velocity to reset to zero each turn making the velocity algorithm jump around causing unneccesay fluctuations.
I would also like to explore a limit2 on the pid/spinout output to keep it positive.
Interestingly I think that a negative command i.e. S500M4 shows up as -500 on the gmoccapy screen, but if the spindle is direction is already established running off an m3 command, anything negative will not reverse direction again mainly because I suspect that direction is set initially with the direction pin and thereafter velocity is simply controlled via the analogue signal which itself can not be negative.
IIRC limiting minlim to a slightly positive value rather that mirroring maxlim but as a negative figure prevents an m4 being executed. In other words, it seems the command needs to see and respond to the negative in the command in the initial instance, but thereafter can only control the velocity of the m4 via a 0 to 5v signal (in absolute terms). Similarly for an m3 command. The initial positive output confirms direction (m3) and thereafter the actual rps is solely controlled by the 0-5v signal. In other words a pid commanded negative signal after the initial direction has been established does not seemingly change direction.
Similarly while I would like to try it, I therefore have my doubts about a limit2 function locking the commands to the positive (greater then zero) side being an answer. Without an initial negative command (m4) as the initial output to establish the direction pin, an m4 could never be executed.
Buuut perhaps I'm reading waaay too much into it?
Thanks again Andy and moreover thanks for sticking with this most verbose post
In short, a negative command does not run the vfd in reverse after initial establishment of sign for direction. As the 7i76e spindle output is an analogue wiper signal it can never be less than zero, with direction ultimately being commanded by the direction pin.
I have a few things I would like to check as well, specifically why the feedback velocity is jumping around like it does. I wonder whether I have an index-enable signal/pin hard set true somewhere causing the velocity to reset to zero each turn making the velocity algorithm jump around causing unneccesay fluctuations.
I would also like to explore a limit2 on the pid/spinout output to keep it positive.
Interestingly I think that a negative command i.e. S500M4 shows up as -500 on the gmoccapy screen, but if the spindle is direction is already established running off an m3 command, anything negative will not reverse direction again mainly because I suspect that direction is set initially with the direction pin and thereafter velocity is simply controlled via the analogue signal which itself can not be negative.
IIRC limiting minlim to a slightly positive value rather that mirroring maxlim but as a negative figure prevents an m4 being executed. In other words, it seems the command needs to see and respond to the negative in the command in the initial instance, but thereafter can only control the velocity of the m4 via a 0 to 5v signal (in absolute terms). Similarly for an m3 command. The initial positive output confirms direction (m3) and thereafter the actual rps is solely controlled by the 0-5v signal. In other words a pid commanded negative signal after the initial direction has been established does not seemingly change direction.
Similarly while I would like to try it, I therefore have my doubts about a limit2 function locking the commands to the positive (greater then zero) side being an answer. Without an initial negative command (m4) as the initial output to establish the direction pin, an m4 could never be executed.
Buuut perhaps I'm reading waaay too much into it?
Thanks again Andy and moreover thanks for sticking with this most verbose post
Please Log in or Create an account to join the conversation.
21 Nov 2018 12:31 #121116
by andypugh
Replied by andypugh on topic Spindle control woes
OK, so you are saying that the problem lies with your spindle drive, in that it doesn't respond to a direction command once it has started?
Is this configurable? It's not a great feature, and might make rigid tapping impossible. Perhaps this is a configuration setting in the drive?
There are ways to work round this in HAL, I feel sure.
My first idea would have been to use a pair of mux2 to set the limits of a limit1, but that won't work as the limit1 limits are parameters not pins.
So the answer is probably to use two limit1 components, one for forward and one for reverse then choose between them using a mux2.
In fact it might be better to use a mux4. Then connect sel0 to spindle-fwd and sel1 to spindle-rev. That means that you can setp the output to zero on the unused (0 and 3) outputs too.
Is this configurable? It's not a great feature, and might make rigid tapping impossible. Perhaps this is a configuration setting in the drive?
There are ways to work round this in HAL, I feel sure.
My first idea would have been to use a pair of mux2 to set the limits of a limit1, but that won't work as the limit1 limits are parameters not pins.
So the answer is probably to use two limit1 components, one for forward and one for reverse then choose between them using a mux2.
In fact it might be better to use a mux4. Then connect sel0 to spindle-fwd and sel1 to spindle-rev. That means that you can setp the output to zero on the unused (0 and 3) outputs too.
Please Log in or Create an account to join the conversation.
Time to create page: 0.165 seconds