Limit&Home at one pin
28 Oct 2014 17:11 #52487
by andypugh
One thing to try would be to temporarily remove the limit switch connections, then home only the first axis in the sequence (X?) and make sure that the switch is clear at the end of the homing move, then do the same with Y etc.
it may be that one of the axes is finishing its homing sequence on its switch.
Replied by andypugh on topic Limit&Home at one pin
net all-home <= parport.0.pin-13-in-not
net all-home => axis.0.home-sw-in axis.0.neg-lim-sw-in
net all-home => axis.1.home-sw-in axis.1.neg-lim-sw-in
net all-home => axis.2.home-sw-in axis.2.neg-lim-sw-in
One thing to try would be to temporarily remove the limit switch connections, then home only the first axis in the sequence (X?) and make sure that the switch is clear at the end of the homing move, then do the same with Y etc.
it may be that one of the axes is finishing its homing sequence on its switch.
Please Log in or Create an account to join the conversation.
28 Oct 2014 19:05 - 28 Oct 2014 19:08 #52491
by Green#
Replied by Green# on topic Limit&Home at one pin
I made some test on pin setup as below and everything works perfect, I can do "home all axis" without any errors but in this case I have to use two pins: 0.13 and 0.10... I think it should be possible done on one pin,
net all-home <= parport.0.pin-13-in-not
net all-home => axis.0.home-sw-in
net all-home => axis.1.home-sw-in
net all-home => axis.2.home-sw-in
net all-lim <= parport.0.pin-10-in-not
net all-lim => axis.0.neg-lim-sw-in axis.0.pos-lim-sw-in
net all-lim => axis.1.neg-lim-sw-in axis.1.pos-lim-sw-in
net all-lim => axis.2.neg-lim-sw-in axis.2.pos-lim-sw-in
and all homing cycles are finishing outside the switch
net all-home <= parport.0.pin-13-in-not
net all-home => axis.0.home-sw-in
net all-home => axis.1.home-sw-in
net all-home => axis.2.home-sw-in
net all-lim <= parport.0.pin-10-in-not
net all-lim => axis.0.neg-lim-sw-in axis.0.pos-lim-sw-in
net all-lim => axis.1.neg-lim-sw-in axis.1.pos-lim-sw-in
net all-lim => axis.2.neg-lim-sw-in axis.2.pos-lim-sw-in
and all homing cycles are finishing outside the switch
Last edit: 28 Oct 2014 19:08 by Green#.
Please Log in or Create an account to join the conversation.
28 Oct 2014 19:33 #52494
by andypugh
I am fairly sure that it is possible, as that is how my machine used to be set up (until I really ran out of pins and switched to a 7i43 board instead).
One _possible_ cause of what you see might be switch-bounce.
With the shared pin, does the homing sequence always fail at exactly the same point? (and which point is that?)
Replied by andypugh on topic Limit&Home at one pin
I think it should be possible done on one pin
I am fairly sure that it is possible, as that is how my machine used to be set up (until I really ran out of pins and switched to a 7i43 board instead).
One _possible_ cause of what you see might be switch-bounce.
With the shared pin, does the homing sequence always fail at exactly the same point? (and which point is that?)
Please Log in or Create an account to join the conversation.
28 Oct 2014 19:56 #52496
by cncbasher
Replied by cncbasher on topic Limit&Home at one pin
as well as Andy mentioned over switch bounce , also concider noise and long wires , especialy if voltage free ,
hence why closed switchs and fed from say 5v is better than 0v and open ) especialy with high voltages around such as Plasmas ..
use sheilded cables .
hence why closed switchs and fed from say 5v is better than 0v and open ) especialy with high voltages around such as Plasmas ..
use sheilded cables .
Please Log in or Create an account to join the conversation.
28 Oct 2014 20:10 #52497
by andypugh
Replied by andypugh on topic Limit&Home at one pin
Here is how to software-debounce your limit/home switch using the "debounce" component (www.linuxcnc.org/docs/html/man/man9/debounce.9.html )
It will use two debounced outputs from the ssme input. False-tripping of limits is annoying, and the machine really won't go far in 10mS, but if the home is that long then the homing accuracy will suffer.
So, the HAL config below sets home debounce to 20 base-thread-periods and limit-debounce to 200 base-thread-periods. Feel free to adjust the numbers.
It might be that stepconf knowns how to debounce, I can't recall.
It will use two debounced outputs from the ssme input. False-tripping of limits is annoying, and the machine really won't go far in 10mS, but if the home is that long then the homing accuracy will suffer.
So, the HAL config below sets home debounce to 20 base-thread-periods and limit-debounce to 200 base-thread-periods. Feel free to adjust the numbers.
loadrt debounce size=1,1
...
addf debounce.0 base-thread
addf debounce.1.base thread
...
setp debounce.0 200
setp debounce.1 20
...
net limit-home-in parport.0.pin-13-in-not => debounce.0.0,in debounce.1.0.in
net all-lim debounce.0.0.out
net all-lim => axis.0.neg-lim-sw-in axis.0.pos-lim-sw-in
net all-lim => axis.1.neg-lim-sw-in axis.1.pos-lim-sw-in
net all-lim => axis.2.neg-lim-sw-in axis.2.pos-lim-sw-in
net all-home debounce.1.0.out
net all-home => axis.0.home-sw-in
net all-home => axis.1.home-sw-in
net all-home => axis.2.home-sw-in
It might be that stepconf knowns how to debounce, I can't recall.
The following user(s) said Thank You: Green#
Please Log in or Create an account to join the conversation.
29 Oct 2014 02:20 #52510
by Green#
Replied by Green# on topic Limit&Home at one pin
Great idea, i think it should work. I will make some tests and give an answer. Thank you in advance
Please Log in or Create an account to join the conversation.
02 Nov 2014 04:20 #52675
by Green#
Replied by Green# on topic Limit&Home at one pin
I made some tests on slow machine driven by trapeze screw and stepmotors and Your solution works perfect. I had to adjust debounce time only. At this machine I got acceptable effect with debounce set to 2 for home switch and 7000 for limit switch. Tests I made only for checking a code, simulated on simple milling machine, not real plasma. In close future It will be plasma machine drive by servomotors. Thank You Andy!
Please Log in or Create an account to join the conversation.
02 Nov 2014 04:49 #52676
by andypugh
With more advanced hardware you can use things like automatic torch-height control for example.
It may be possible to use this device (as an example) with a parport, but I don't think that the results would be as good as with one of the inexpensive alternatives.
store.mesanet.com/index.php?route=produc...oduct&product_id=128
Replied by andypugh on topic Limit&Home at one pin
At that point I would strongly recommend not using the parport any more, and not only to have more home and limit switches.In close future It will be plasma machine drive by servomotors.
With more advanced hardware you can use things like automatic torch-height control for example.
It may be possible to use this device (as an example) with a parport, but I don't think that the results would be as good as with one of the inexpensive alternatives.
store.mesanet.com/index.php?route=produc...oduct&product_id=128
Please Log in or Create an account to join the conversation.
02 Nov 2014 05:08 - 02 Nov 2014 05:11 #52677
by Green#
Replied by Green# on topic Limit&Home at one pin
It will be machine with torch height control. I'm preparing design now an electronic board consist of THC with RS232 transmission, input/output optoisolation, and some more stuff, I think it should work but we will see. Project is based on one lpt and rs232.
Last edit: 02 Nov 2014 05:11 by Green#.
Please Log in or Create an account to join the conversation.
02 Nov 2014 05:14 #52678
by andypugh
Feel free to make things for fun, but tried and tested commercial alternatives exist
Replied by andypugh on topic Limit&Home at one pin
It will be machine with torch height control. I'm preparing design now an electronic board consist of THC with RS232 transmission, input/output optoisolation, and some more stuff, I think it should work but we will see. Project is based on one lpt and rs232.
Feel free to make things for fun, but tried and tested commercial alternatives exist
Please Log in or Create an account to join the conversation.
Time to create page: 0.193 seconds