Holding single axis
25 Oct 2021 13:52 - 25 Oct 2021 13:52 #224264
by andypugh
Replied by andypugh on topic Holding single axis
I think that I would do this with a custom HAL component that holds the X encoder position when triggered.
But I think it can probably be done with standard components, just not as neatly.
You need to accumulate an offset in a HAL signal and then subtract that from the encoder position number before passing it to motion.
(Note that subtraction in HAL is done by using the sum2 component with a negative gain on one input)
The built-in sample-hold component in HAL only works with integers, and this will need to work with floating point numbers. The workaround is to use the mux2 component looped back to itself.
I think that this structure of 2 x mux2 and 2 x sum2 will do what is needed, but my only "testing" has been looking at the picture and imagining the numbers on the pins
But I think it can probably be done with standard components, just not as neatly.
You need to accumulate an offset in a HAL signal and then subtract that from the encoder position number before passing it to motion.
(Note that subtraction in HAL is done by using the sum2 component with a negative gain on one input)
The built-in sample-hold component in HAL only works with integers, and this will need to work with floating point numbers. The workaround is to use the mux2 component looped back to itself.
I think that this structure of 2 x mux2 and 2 x sum2 will do what is needed, but my only "testing" has been looking at the picture and imagining the numbers on the pins
Attachments:
Please Log in or Create an account to join the conversation.
25 Oct 2021 19:22 - 25 Oct 2021 19:27 #224292
by andypugh
Replied by andypugh on topic Holding single axis
I did admit that I hadn't tested it.
After some trial and error, this seems to work. It is a standalone config that you can try with "halrun -I test.hal" if you save it as test.hal. Then you can "sets encoder 100" to change encoder counts and "sets hold 1" to set the hold input.
After some trial and error, this seems to work. It is a standalone config that you can try with "halrun -I test.hal" if you save it as test.hal. Then you can "sets encoder 100" to change encoder counts and "sets hold 1" to set the hold input.
loadrt threads
#load mux2
loadrt mux2 count=2
addf mux2.0 thread1
addf mux2.1 thread1
#load sum2
loadrt sum2 count=2
addf sum2.0 thread1
addf sum2.1 thread1
setp sum2.0.gain1 -1
setp sum2.1.gain1 1
setp mux2.1.in1 0
setp mux2.0.sel FALSE
setp mux2.1.sel FALSE
net encoder sum2.1.in0 sum2.0.in1
net n_1 mux2.0.out => sum2.0.in0 mux2.0.in1
net output sum2.1.out mux2.0.in0
net n_3 sum2.0.out => mux2.1.in1
net n_4 mux2.1.out => mux2.1.in0 sum2.1.in1
net hold mux2.0.sel mux2.1.sel
loadusr halmeter signal output
loadusr halmeter signal encoder
loadusr halmeter signal n_1
loadusr halmeter signal n_3
loadusr halmeter signal n_4
start
Attachments:
Please Log in or Create an account to join the conversation.
Time to create page: 0.108 seconds