odd/nonexistant homing behavior

More
01 Jun 2017 19:51 #93984 by JR1050
See attached hal/.ini files

suggested settings produce endless following errors. Im not even sure how one can get a following error on a stepper , it has no feed back!! I can see how the following error can happen with an encoder attached to a stepper though


File Attachment:

File Name: knights.hal
File Size:10 KB

File Attachment:

File Name: knights_20...6-01.ini
File Size:4 KB
Attachments:

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

More
01 Jun 2017 20:01 - 01 Jun 2017 20:02 #93985 by JR1050
"You are paying too much attention to the INI and not enough to the HAL files.

The INI is fine. It is how the bits fit together in HAL that needs to be tweaked."

can you be more specific about what needs more tweaking. Truthfully, most of what is in these hal files are from proven machines i have already completed. The difference being they were on a standard pc and this is on a rspi3.. Again, this worked before the encoder, new bit file .....
Last edit: 01 Jun 2017 20:02 by JR1050.

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

More
01 Jun 2017 20:06 #93986 by andypugh
Following errors with steppers happen when the step-generator component says that it has made the wrong number of steps. (which can happen for a number of reasons).
However, I have to say that things all look as I would expect.

Which axis / joint is the following error reported on?

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

More
01 Jun 2017 20:08 #93987 by andypugh
You can change

net motor.00.pos-fb => joint.0.motor-pos-fb

to

net motor.00.pos-cmd => joint.0.motor-pos-fb

To make following-error impossible. But it isn't the way to leave it.

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

More
01 Jun 2017 20:14 #93988 by andypugh

Again, this worked before the encoder, new bit file .....


I wonder if the stepgen function in that bitfile is actually broken?

You should be able to check it at the command-line

halrun
loadrt hostmot2
loadrt hm2_rpspi config=.....
loadrt threads
addf hm2_7i90.0.read thread1
addf hm2_7i90.0.write thread1
start
setp hm2_7i90.0.stepgen.00.position-cmd 100
show pin hm2_7i90.0.stepgen.00.position-fb
show pin hm2_7i90.0.stepgen.00.position-fb
(repeat setting position and monitoring feedback)

(use tab-completion to get the pin-names right, I guessed them)

(loadusr halmeter might give a more convenient way to watch the feedback, or halscope for that matter)

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

More
02 Jun 2017 12:07 #93998 by JR1050
Andy, Ill give this a try. Consider these 3 scenarios


1- original setup. No encoder, stepper was on the X axis. original bit file without an encoder. All functions worked, stepper axis would home.

2-set up above. New bit file with encoder. Encoder is on joint-0, but not used on the stepper motor, stepper is on Joint_0. Encoder and its functions work. X axis will home. Lots of stepper axis following errors.

3- set up at the start of this thread. Encoder is on the first joint(joint_0) at encoder_00, stepper motor is on joint_1(Y),
Encoder and all its functions works, no following errors on the stepper axis, Y (joint_1) will not search for home. It will jog in both directions. Where ever the motor is on start up, that is where lcnc is calling home.Lots of other goofy errors. Id like to use this arrangement, seems to make more sense, it seems if the stepgen is broken, it wouldn't jog?

Thanks

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

More
02 Jun 2017 12:13 #93999 by andypugh

2-set up above. New bit file with encoder. Encoder is on joint-0, but not used on the stepper motor, stepper is on Joint_0.


I really don't know what you mean by "encoder is on joint-0" in this setup.

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

More
02 Jun 2017 12:24 #94000 by JR1050
Im using the encoder that would be associated with Joint_0 (x) on a shaft to sense angular position.From Hal

####encoder feedback for chain position sense####
setp hm2_7i90.0.encoder.00.counter-mode 1
setp hm2_7i90.0.encoder.00.filter 1
setp hm2_7i90.0.encoder.00.scale 1
setp jog.0.index-position 1223

net enc_count_reset jog.0.enc-count-reset => hm2_7i90.0.encoder.00.reset
net enc_position hm2_7i90.0.encoder.00.position => jog.0.enc-position

The encoder is not associated mechanically in any way to the linear stepper axis.

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

More
02 Jun 2017 12:31 #94001 by andypugh
So you simply mean that it is encoder.00 ?
The index numbers have no significance at all, you can connect any encoder to any joint in HAL, and if you don't connect it, then it won't be considered at all in f-error calculations.

(My lathe uses resolver.05 for the X axis and resolver.00 for the spindle, for example and on my mill encoder.00 is the A axis)

If it jogs then the stepgen isn't broken. But can you monitor the stepgen.nn.feedback as you jog to make sure that is making sense?

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

More
02 Jun 2017 12:42 #94003 by JR1050
Well, it seems like there is some thing amiss. By your above explanation, everything should work in scenario 2&3. It errors out in scenario 2 and wont home in scenario 3.
On my big mill(the Brown and sharp) encoder 00 is on the A axis also, it tracks the position of A
But.... encoder 00 is used in scenario 2 and it is not physically connected to the stepper axis. This encoder tracks a separate
for a lack of better term, rotary axis.

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

Time to create page: 0.084 seconds
Powered by Kunena Forum