--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jxW-1cKUzpG0rD8Z_KTXHmvKbFhv%3DkaTZK3mufenY%3DS%2BA%40mail.gmail.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BKNHNzep_VzwtO%2Be0Hu54yW-BRmK%2BOdbzgNBkRvSwOuz9t5Rg%40mail.gmail.com.
Thank you Jason.
Always a bit envious about in how few and simple words you manage to describe complex things so comprehensibly!
_Mark
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/6827f023-bcb7-4d25-8cf8-0bcc2f665a67%40googlegroups.com.
Hi Steve
see the PR Description, section "Instructions for Use" here:
https://github.com/openpnp/openpnp/pull/859
The examples there are from my Liteplacer with Smoothieware setup
(except for the pump where I have a hysteresis solution, so the
given pump on/off commands are fictitious).
If you have the TinyG instead, I guess you must translate the
G-Code slightly.
If you're successful, please share :-)
_Mark
If you already have one or more ReferenceNozzle
s set up, close OpenPNP, edit your machine.xml and replace all
<nozzle class="org.openpnp.machine.reference.ReferenceNozzle"
with
<nozzle class="org.openpnp.machine.reference.ContactProbeNozzle"
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/8c42ce57-5039-4fe1-871e-fcd527fb25a7%40googlegroups.com.
Most likely just a syntax error. If you open the file in a syntax highlighting editor like Sublime it should make it clear what's wrong. If not, post the machine.xml here.Jason
To unsubscribe from this group and stop receiving emails from it, send an email to ope...@googlegroups.com.
But isn't this your video? That's even nicer!!
https://www.youtube.com/watch?v=DO8lbh_2isU
https://groups.google.com/d/msg/openpnp/S7pDpZGIjmk/Soa8hZSqAAAJ
Without a probe, I'm afraid you can't use the ContactProbeNozzle.
_Mark
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/11eec0a7-05ba-4668-af12-0c79a9da14fc%40googlegroups.com.
But isn't this your video? That's even nicer!!
https://www.youtube.com/watch?v=DO8lbh_2isU
https://groups.google.com/d/msg/openpnp/S7pDpZGIjmk/Soa8hZSqAAAJ
Without a probe, I'm afraid you can't use the ContactProbeNozzle.
_Mark
Am 27.11.2019 um 20:47 schrieb Mike M.:
No - I do not have it!--
On Wednesday, 27 November 2019 20:44:57 UTC+1, ma...@makr.zone wrote:Nope. Here's the file:
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ope...@googlegroups.com.
Hi Mike,
> I thot this is vacuum Z sensing upgrade ?
But the ContactProbeNozzle is not about that. This nozzle will
probe every pick and every place, so it must be
very fast. It must use some sort of mechanical contact sensor.
Vacuum "sniffing" is very slow and won't work, if you have the
part on the nozzle i.e. in a place action.
_Mark
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/10a0eeaf-d7a1-4139-b2d3-590507952b58%40googlegroups.com.
Hi Jim
Agree completely, but Z height probing (as opposed to contact
probing) is a different, more general use case, one that
is planned to be added later.
Jason has already introduced the Z-probe Actuator on the head for
that:
Different ways to probe: laser, vacuum sniffing, stereoscopic (and also the nozzle with contact probe) have been discussed. The idea is to allow probing generically wherever a Z value must be captured. So we probably need another one of these blue-framed buttons, something like this:
It is not so easy to do this generically. The probe takes many
forms and may have its own X, Y offset from the OpenPNP coordinate
system (i.e. from the Camera). It could be a completely separate
device (in case of a laser) or it could be the down-looking camera
(stereoscopic) or it could be a nozzle (with vacuum sniffing or
with contact probe) and perhaps only one nozzle out of multiple
can do it and/or only one nozzle tip is calibrated for sniffing.
So aiming and deploying the probe is wildly
different and may involve X, Y moves (in case of the stereoscopic
method, the camera would have to quickly move laterally and take
two shots), it may even involve changing the nozzle tip for
consistent vacuum sniffing.
The idea is to both support the cheap way with stuff already on
the machine (vacuum sniffing, stereoscopic) and the enhanced way
with precision laser ranging etc. through he same common interface.
So we can build many useful solutions on top of that common
interface.
_Mark
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/4e4c6060-435e-4e45-b45d-43733bf8bf1b%40googlegroups.com.
@Mike M.
> Than this upgrade to Open PnP2 is related to OpenPlacer
type of Z only with single nozzle / Samsung -With limit switch
On Z? Am I correct?
Note: it is the Liteplacer, not Openplacer.
There are multiple perspectives: from a purely "what new features
do I get right now?" standpoint, you are more or less right. If
you don't have the sensor for a ContactProbeNozzle, there are only
small benefits right now:
As a basis for further development, the upgrade is much more. See
the other Mail to Jim for a glimpse. There are many other ideas.
You have to read the PR discussion including some previous PRs to
understand.
> I will make a short video to show how this SMC vacuum sensor works - I hope I can show that it is fast enough for confirming part is on the nozzle and to Sense the pickup Z high !
Yes, this is valuable. Thanks.
> Final question - how than this new feature suppose to work with two heads or Quadro head ?
The ContactProbeNozzle? Same way it does with one. Why have you
removed the hall sensors? I'm sure it would have worked nicely
with those two.
_Mark
> to be sure we should check the vacuum again to confirm it. ... You will get info the part is On, raise the nozzle and to be sure we should check the vacuum again to confirm it.
That is already improved in OpenPNP 2.0, because it now lifts the nozzle, before checking. Unless you have a superfast Z, this already adds some time for the vacuum value to deteriorate if no part was picked. So you don't need to enter a long extra dwell time for that ... but you still can if you have superfast Z as with your pneumatic "shooters" ;-).
This not only detects when the covering tape was still on, but
also when no part was in the pocket and the tip smooched the
pocket bottom. Like after a tape feed error perhaps or with an
empty cut strip or tray.
_Mark
Yes, Marek this might be right for your powerful machine with professional vacuum system. But most hobby machines have weak "aquarium" vacuum pumps that will be completely depleted if the nozzle tip valve is left open so long with no part on it. Especially if you have a multi-nozzle machine, that may pick another part before going to the bottom camera. Even worse: when the vacuum dies down, the other nozzle(s) will lose their (potentially expensive) parts too!
_Mark
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/d4c7b1ca-c431-446a-a6a6-f6f5849c0816%40googlegroups.com.
For 2) Frankly, Mike, I think this is way too slow to use for
each and every pick.
While I made it and still think it is useful, you shouldn't
overestimate the benefit of the ContactProbeNozzle. I hope it will
be useful for the passives that I hope to be able to place without
bottom vision. Then it is nice you don't have to setup the part
height precisely for these numerous feeders.
But for all the parts that use bottom vision, you need to setup part heights quite precisely. Otherwise you get focal plane/parallax errors. So the ContactProbeNozzle has not much benefit there (yet).
You should be patient for the real Z probing.
_Mark
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/73b6dedd-28e4-43ad-a46c-2ff0cc35bf2d%40googlegroups.com.
This seems to be working fine for me. Can you check your OpenPnP.log and see if there is an error there when you try this?Jason
On Thu, Nov 28, 2019 at 1:46 PM Mike M. <mike...@gmail.com> wrote:
@Marek - good news - so I am not alone with this.....@Mark/ JasonMy add feeder "+" does not work after this update - OpenPnP 2.0 Breaking ChangeI attach screenshot here;How can I resolve this?Can you give me some hints? sujestion what shall I try ?ThanksMikeOn Thursday, 28 November 2019 18:18:05 UTC+1, Marek T. wrote:--Yes, I know you were forced with your machine setup to work as you work... But I think that this method really makes sense - and I'm gonna to test it soon.
W dniu czwartek, 28 listopada 2019 18:07:56 UTC+1 użytkownik cncmachineguy napisał:Thanks for the credit Marek, but I can't say I was thinking that. My commercial machine worked that way so I just kept it going like that. My commercial machine only took seconds at most to load 6 nozzles and grab a pic of each. It could decide faster than I could if it had a part.
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ope...@googlegroups.com.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/850cd4ff-ace4-4cd2-9653-35de83041948%40googlegroups.com.
In the .openpnp2 folder there should be an OpenPnP.log file (no number). Try to open the feeder dialog, then see if there is any errors at the end of that file, or just send me the whole file.Jason
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/850cd4ff-ace4-4cd2-9653-35de83041948%40googlegroups.com.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/dafddfb3-3231-4262-995a-94a3fec77e55%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/dafddfb3-3231-4262-995a-94a3fec77e55%40googlegroups.com.
Hi Mike
the problem is that currently, probing is done both on pick() and
place() it can't be disabled for the place() and you can't use
your method with the part on the nozzle, right? :-)
For this to work, I would have to add checkboxes to the
configuration. I might do that, but right now I don't have the
time, unfortunately.
_Mark
{True:G38.2 Z-4 F1200 ; Probe down max. 4mm for contact with picked/placed part }
{True:M400 ; Wait until machine has stopped }
{True:M114.2 ; Report current realtime position, as M114 does not report rotation }
{False:G91 G0 Z1 G90; Retract 1mm }
{False:M400 ; Wait until machine has stopped }
{False:M114.2 ; Report current realtime position, as M114 does not report rotation }
Hi Steve
see the PR Description, section "Instructions for Use" here:
https://github.com/openpnp/openpnp/pull/859
The examples there are from my Liteplacer with Smoothieware setup (except for the pump where I have a hysteresis solution, so the given pump on/off commands are fictitious).
If you have the TinyG instead, I guess you must translate the G-Code slightly.
If you're successful, please share :-)
_Mark
Am 26.11.2019 um 15:00 schrieb 'Steven Doyle' via OpenPnP:
Hi all,Has anyone setup the ContactProbeNozzle for a liteplacer machine ? I would like to know what the GCODE is for the probing actuator and what POSITION_REPORT_REGEX should be.
Thanks !
Steve
Yep, that's a downer:
https://github.com/synthetos/TinyG/wiki/Probing-Description-and-Operation
I guess it is that quirk that is biting you "G4P0 does not wait for end of movement":
https://github.com/openpnp/openpnp/wiki/TinyG#quirks
Try these things:
1. Make sure you have the workaround by Jason in place. Also use
it in the wait after probing.
"It is possible to use M2 for end of move detection, example using lazy Json - https://justpaste.it/1d3ht"
2. If this does not help, use the standard F20 they seem
to use with the TinyG. There are error reports
on the web with other feedrates. However F20 will probably
be too slow for "always" probing in PnP.
So something like this, I don't know if it is valid syntax with the double curly brackets:
{True:G38.2 Z-4 F20 ; Probe down max. 4mm for contact with picked/placed part } {True:
{gc:"g91x0y0z0m2"}; Wait until machine has stopped }
{True:
{
sr:n}; Wait until machine has stopped }
{True:M114.2 ; Report current realtime position, as M114 does not report rotation }
{False:G91 G0 Z1 G90; Retract 1mm }
{True:
{gc:"g91x0y0z0m2"}; Wait until machine has stopped }
{True:
{
sr:n}; Wait until machine has stopped }
{False:M114.2 ; Report current realtime position, as M114 does not report rotation }
> I'm beginning to think that TinyG is more trouble than it worth.
You might be right. I guess there are good reasons, Jason issued a warning here:
That's why I went with Smoothieware from the start.
_Mark
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/52d2065a-eba9-4133-9768-90123f01f757%40googlegroups.com.
To unsubscribe from this group and stop receiving emails from it, send an email to ope...@googlegroups.com.
One thing that came to mind: you could try to cheat and look into
the Liteplacer software that does contact probing too:
https://github.com/jkuusama/LitePlacer-DEV/blob/master/LitePlacer/CNC.cs#L583
> You've got a liteplacer I think, could you post your smoothie config file please.
I should finally put some stuff on the blog. For now
I'll send it PM to you, as I'm not yet happy with the
documentation.
_Mark
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/2ac85978-3440-4cff-a7e5-679fc612e2c4%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/2ac85978-3440-4cff-a7e5-679fc612e2c4%40googlegroups.com.
Feeder location correction - I have had this idea but lack of time... :-(.
Fly over vision, vision on the fly... yes, in Openpnp 3.0 :-).
I don't check vac directly after place. I did it priorly but it time costs, vac valve must be on and need wait long time to give it a chance to vac level could build up if part is not left on the board. So I've removed it from here and test it before the pick, so after the motion from place to pick location is done. During this move the vac level has time to rise if there is a part...
Feeder location correction - I have had this idea but lack of time... :-(.
Fly over vision, vision on the fly... yes, in Openpnp 3.0 :-).
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/5be29254-661f-4b9b-a7a7-7d3a1e115d0b%40googlegroups.com.
Hi Jim
> so i made it selectable in the v2 ReferencePnPJobProcessor
This should no longer be needed in the version that this thread is about (i.e. with PR #859 merged).
If you set your "Part Off Nozzle Vacuum Range" both to zero, it
won't be checked.
I made it two separate criteria:
Nozzle.isPartOnEnabled()
Nozzle.isPartOffEnabled()
https://github.com/openpnp/openpnp/pull/859/files#diff-a593adfb65af36a22ab0278f18e54f1a
_Mark
Hi Jim
Ah, I see. In deed the option could even make sense per
package.
_Mark
You've got scripting exceptions:
2019-12-12 09:51:04.336 VisionUtils WARNING:
javax.script.ScriptException: ReferenceError: "camera" is not
defined in Vision.PartAlignment.Before.js at line number 9
at
jdk.nashorn.api.scripting.NashornScriptEngine.throwAsScriptException(NashornScriptEngine.java:470)
_Mark
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/e719bb6b-e84f-481d-b75e-6efff4e77c47%40googlegroups.com.
In deed the option could even make sense per package.
To unsubscribe from this group and stop receiving emails from it, send an email to ope...@googlegroups.com.
Sorry Mike I don't understand what you mean.
_Mark