#linuxcnc-devel Logs

Sep 24 2022

01:30 AM mrec: andypugh: are you there?
01:37 AM mrec: so far I have modified linuxcnc to travel a particular angle (but not synchronized yet, so the movement looks a bit jerky)
01:37 AM mrec: I'm still studying the part of the sourcecode, EMCMOT_MOTION_FREE is not synchronized among the axis
01:37 AM mrec: what is the purpose of EMCMOT_MOTION_COORD, coordinated movement?
01:37 AM mrec: but doesn't EMCMOT_MOTION_TELEOP also provide coordinated movement?
01:37 AM mrec: (I still need to look into COORD/TELEOP
04:46 AM andypugh: mrec: I really woudn’t have done it in the source code. It’s just begging to be a HAL compinent.
04:55 AM mrec: andypugh: I need to understand the components which are involved first
04:55 AM mrec: I don't have a full overview yet
04:56 AM andypugh: Not really, you could send jog-counts from the MPG into a HAL component and get sin and cos jog counts out the other end to send to X and Y.
04:57 AM andypugh: Unless you also want incremental GUI jogs to work at angles too?
04:58 AM mrec: well that's another story for another time, first the handwheel
04:58 AM mrec: which pins should I use for X/Y?
04:59 AM mrec: currently it uses EMCMOT_MOTION_FREE for jogging
04:59 AM mrec: which of course ends up in some jerky movement
05:01 AM mrec: I just don't have enough insight yet, but I think it's not too difficult to get it right
05:04 AM mrec: sin/cos is not a problem, take the counts multiply them with the scaling factor that's it; compound travel could be wired up with A/B on the lathe; but I'm not sure if this has some other side effects.
05:04 AM mrec: Using the Y axis in lathe mode (for some other purpose does have side effects)
05:05 AM mrec: I used it for the tailstock once
05:33 AM mrec: andypugh: still there?
05:36 AM andypugh: Yes, still here.
05:36 AM mrec: which pins would you recommend for jogging (after translating the coordinates)?
05:37 AM andypugh: The way you seem tobe doing this is so different from the way that I have been considering that I don’t think I can help much.
05:38 AM mrec: what is your consideration?
05:39 AM andypugh: Well, I would just take the jog counts from spare outputs of the DRO, pass them through a HAL component and link them to the jog-counts of the X and Z axes.
05:39 AM mrec: that doesn't work since jogging is done in EMCMOT_MOTION_FREE mode
05:39 AM andypugh: (Sorry, I mean MPG not DRO)
05:39 AM mrec: axes are controlled independently
05:40 AM andypugh: ?
05:40 AM mrec: when simulating the compound I basically have an angle and a length
05:40 AM andypugh: On my lathe I can jog X and Z from the handwheels at the same time.
05:41 AM mrec: yes but they are not linked
05:41 AM mrec: they will move independently
05:41 AM andypugh: Yes, and?
05:41 AM mrec: when simulating a compound the movement should be straight and not jerky
05:42 AM mrec: eg. like travels with G0 and G1
05:43 AM andypugh: You mean that you want constant speed movement, not related to how fast you are turning the jog wheel?
05:43 AM andypugh: In that case, why not MDI a G1?
05:44 AM mrec: because I want to control that via handwheel, that's quicker than G1 for prototyping
05:44 AM andypugh: But if you want to handwheel-jog at an angle, then just have a HAL component jog X and J for you.
05:44 AM andypugh: (X and Z)
05:44 AM mrec: https://snipboard.io/okWcGQ.jpg
05:45 AM mrec: that's what I have at the moment
05:45 AM mrec: it depends on how quick I turn the wheel
05:45 AM andypugh: Then the maths is wrong, somewhere.
05:46 AM mrec: no the math is easy and correct it's 45°
05:46 AM mrec: just the motors have different settings
05:46 AM mrec: acceleration...
05:47 AM mrec: again X/Z are traveling independently they are not linked
05:48 AM andypugh: As an experiment, have you tried linking axis.x.jog counts _and_ axis.z.jog counts to the same MPG channel? Does that give a straight line?
05:49 AM andypugh: If I want an angled cut, I just programme an angle into the relevant turning or facing cycle on my macro tab….
05:50 AM mrec: are your X/Z motors the same, with the same configuration?
05:51 AM andypugh: https://forum.linuxcnc.org/41-guis/26550-lathe-macros?start=240#247196
05:52 AM andypugh: No, completely different. Z is a 1kW servo about the size of a small dog and the X is a 600W one that fits inside the aprom casting.
05:52 AM mrec: In free mode, the free mode trajectory planners are disabled. That results in each joint stopping as fast as its accel (decel) limit
05:52 AM mrec: allows. The stop is not coordinated.
05:52 AM mrec: did you wire your pendant with the joint or with the axis?
05:52 AM andypugh: mrec: Are you jogging by counts?
05:52 AM mrec: yes
05:53 AM andypugh: I don’t actually have a pendant, I have two handwheels on the apron. But, yes, the counts go to both jount and axis jog pins.
05:54 AM andypugh: Forget free/coordinated, onece the machine is homed it jogs on the axis-counts inputs.
05:54 AM andypugh: So, what have you actually done to get your angled jog?
05:54 AM mrec: I did not home it and it used joints
05:55 AM mrec: I have added sin/cos calculation to handle_jjogwheels
05:55 AM andypugh: Yes, that’s right. Before you home it it jogs on the joint pins, and after on the axis pins.
05:56 AM mrec: and I'm putting the result into the X/Y joint. the result is expected so far.
05:56 AM andypugh: Can you jiust try what I suggested, link the same jog-counts MPG output to both the X and Z axis jog-counts inputs and see if you get a straight line?
05:57 AM andypugh: I _really_ don;t think that this belongs in the source code.
05:57 AM mrec: it doesn't matter I'm just studying it.
05:58 AM mrec: my acceleration settings differ X/Z that's why the output is expected the stop is always at the correct position (45°)
05:58 AM andypugh: Maybe you need to limit the count rate to be something that both joints can follow?
05:59 AM mrec: no again in free mode they just move independently
05:59 AM mrec: without the modifications I can just jog X and while X is moving I can also jog Z (and Z might be complete even before X is completed)
06:00 AM mrec: there needs to be some trajectory planning in free mode there is no linked one
06:01 AM andypugh: axis.L.jog-vel-mode IN BIT
06:01 AM andypugh: Try it.
06:01 AM andypugh: I think you are on the wrong track. But carry on ignoring me if you want.
06:05 AM mrec: velocity mode is something I need to have a look at, I could certainly be on the wrong track but it's just part of studying the code for me
06:43 AM mrec: too bad that he left
06:44 AM mrec: https://i.snipboard.io/Ayign4.jpg (that's the result when pushing 2 arrow keys at the same time, of course I did it multiple times)
06:44 AM mrec: but that just shows the acceleration
08:26 AM linuxcnc-build: build #2451 of 1660.rip-buster-python3 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1660.rip-buster-python3/builds/2451 blamelist: Hans Unzner <hansunzner@gmail.com>
08:28 AM pere: wish hansu were here. :)
08:30 AM linuxcnc-build: build #2853 of 1640.rip-buster-rtpreempt-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1640.rip-buster-rtpreempt-amd64/builds/2853 blamelist: Hans Unzner <hansunzner@gmail.com>
08:32 AM linuxcnc-build: build #2612 of 1650.rip-buster-rtpreempt-rpi4 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1650.rip-buster-rtpreempt-rpi4/builds/2612 blamelist: Hans Unzner <hansunzner@gmail.com>
08:32 AM linuxcnc-build: build #9253 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/9253 blamelist: Hans Unzner <hansunzner@gmail.com>
05:21 PM -!- #linuxcnc-devel mode set to +v by ChanServ
05:21 PM -!- #linuxcnc-devel mode set to +v by ChanServ
05:42 PM -!- #linuxcnc-devel mode set to +v by ChanServ