question re. estop logic - (pncconf output)

More
30 Jul 2011 03:09 #12005 by 1:1
Hello,

Lots of questions today -

I set up both an EStop input and output in pncconf today and got the following hal set up (its been formatted to my way of reading)
#Estop_______________________________________
setp    hm2_5i22.0.gpio.049.is_output true
net 	estop-out     hm2_5i22.0.gpio.049.out   <=  iocontrol.0.user-enable-out
net 	estop-ext     iocontrol.0.emc-enable-in <=  hm2_5i22.0.gpio.048.in

#Enable__________________________________
setp    hm2_5i22.0.gpio.050.is_output true
setp    hm2_5i22.0.gpio.050.is_opendrain true
net 	enable     =>  hm2_5i22.0.gpio.050.out
net	enable     =>  motion.motion-enabled

#Chargepump_________________________________
loadrt 	charge_pump
addf 	charge-pump 		servo-thread
net 	estop-out       =>  charge-pump.enable
setp    hm2_5i22.0.gpio.051.is_output true
net 	charge-pump   hm2_5i22.0.gpio.051.out  <=  charge-pump.out

#Coolant___________________________________
setp    hm2_5i22.0.gpio.052.is_output true
net 	coolant-mist      	hm2_5i22.0.gpio.052.out <=  iocontrol.0.coolant-mist

The way I read it (in the manual) initially was that you'd have an output signal on say the 5i22 gpio pin 49. 5v which you'd link to your machines physical estop system so if you E-stopped it in Axis (or EMC decided it was appropriate) then whatever protection you had outboard would also kick in (active brakes, relays on power whatever...). The flip side was that if you hit a physical estop button that the same machine protection would kick in but also advise EMC2 it had done so. Basically whichever 'side' caused the stop the other would be aware, by a Normally On loop closing and breaking the 'not estopped' voltage.

First Question: Does this sound logical - is it how its done ? I want to get into machinery that operates around humans, so I want to get very particular about the safety logic.

The way I actually see it is different however - I'm getting 5v on the input and 0v on the output - EMC/Axis is live from word go and attempts to hit the Estop in axis do nothing - but connecting the two pins I get estop - pretty much the opposite of what I expected. I dont want to simply hack around with inverting things until it works as I might end up with a "two wrongs = a right" situation of which one of those wrongs will cause a problem elsewhere...

Another aspect to this is that I see no charge-pump in any state either - there is a connection (literally and figuratively) - I just dont get it yet (I understand charge pump, just not why it isn't on)

Enable and coolant are working - so you can be sure I'm checking the right pins this time :blush:

any help appreciated,
Nick

complete hal file attached for context if required

File Attachment:

File Name: XYZA.hal
File Size:9 KB
Attachments:

Please Log in or Create an account to join the conversation.

More
30 Jul 2011 07:36 #12013 by cmorley
ok you found another bug. looks like 'enable' should be 'xenable'

Please Log in or Create an account to join the conversation.

More
30 Jul 2011 07:54 #12014 by cmorley
or maybe not - let me study this. can you post you pncconf file?

Please Log in or Create an account to join the conversation.

More
30 Jul 2011 07:56 #12015 by 1:1
hmmm,

are you answering my other thread ??

and in either case - which 'enable' are you referring to ?

and why xenable ? why not - yenable or ??
:S

Please Log in or Create an account to join the conversation.

More
30 Jul 2011 08:01 - 30 Jul 2011 08:02 #12016 by 1:1
cmorley wrote:

or maybe not - let me study this. can you post you pncconf file?


I think thats it - I've done a lot of switching about and copy and paste today so I cant be sure ...

>>UPDATE - FORUM WILL NOT LET ME UPLOAD IT - WONT ACCEPT FILE TYPE NOR FILE SIZE :(

Still though - what problem are we solving here ? It sounds like the other thread ?

>>>> www.linuxcnc.org/component/option,com_ku.../lang,english/#12012
Last edit: 30 Jul 2011 08:02 by 1:1.

Please Log in or Create an account to join the conversation.

More
30 Jul 2011 08:02 #12017 by cmorley
EMC does not have a single enable it has one for each axis. we just choose one axis to enable every thing. all config have xenable.
but something is not quite clear to me here so let me digest. ( its 1 am here )

Please Log in or Create an account to join the conversation.

More
30 Jul 2011 08:04 #12018 by cmorley
archive the pncconf file. or change the end from pncconf to txt

Please Log in or Create an account to join the conversation.

More
30 Jul 2011 08:11 #12019 by 1:1
Can we please stop using this thread about this issue ? I'm scared my estop query will get lost...

proper thread here:

www.linuxcnc.org/component/option,com_ku.../lang,english/#12012

Please Log in or Create an account to join the conversation.

More
30 Jul 2011 08:16 #12020 by 1:1
I've got 5mins here before I go ... Haven't archived on linux yet (is there a utility on the LiveCD install?) and have to switch cables everytime I change machines

>>update, computer crash

not tonight

Please Log in or Create an account to join the conversation.

More
30 Jul 2011 11:55 #12027 by 1:1
an attempt to get back on topic:

I've set up both an EStop input and output in pncconf today and got the following hal set up (its been formatted to my way of reading)
#Estop_______________________________________
setp    hm2_5i22.0.gpio.049.is_output true
net 	estop-out     hm2_5i22.0.gpio.049.out   <=  iocontrol.0.user-enable-out
net 	estop-ext     iocontrol.0.emc-enable-in <=  hm2_5i22.0.gpio.048.in

#Chargepump_________________________________
loadrt 	charge_pump
addf 	charge-pump 	 servo-thread
net 	estop-out       =>  charge-pump.enable
setp    hm2_5i22.0.gpio.051.is_output true
net 	charge-pump   hm2_5i22.0.gpio.051.out  <=  charge-pump.out

The way I read it (in the manual) initially was that you'd have an output signal on say the 5i22 gpio pin 49. 5v which you'd link to your machines physical estop system so if you E-stopped it in Axis (or EMC decided it was appropriate) then whatever protection you had outboard would also kick in (active brakes, relays on power whatever...). The flip side was that if you hit a physical estop button that the same machine protection would kick in but also advise EMC2 it had done so. Basically whichever 'side' caused the stop the other would be aware, by a Normally On loop closing and breaking the 'not estopped' voltage.

First Question: Does this sound logical - is it how its done ? I want to get into machinery that operates around humans, so I want to get very particular about the safety logic.

The way I actually see it is different however - I'm getting 5v on the input and 0v on the output - EMC/Axis is live from word go and attempts to hit the Estop in axis do nothing - but connecting the two pins I get estop - pretty much the opposite of what I expected. I dont want to simply hack around with inverting things until it works as I might end up with a "two wrongs = a right" situation of which one of those wrongs will cause a problem elsewhere...

Another aspect to this is that I see no charge-pump in any state either - there is a connection (literally and figuratively) - I just dont get it yet (I understand charge pump, just not why it isn't on)

any help appreciated,
Nick

Please Log in or Create an account to join the conversation.

Moderators: cmorley
Time to create page: 0.191 seconds
Powered by Kunena Forum