Nozzle 2 is offset!

[Eric Norton]

Hi All,

 I have a strange situation... When Calibrate precise camera ↔ nozzle N2 offsets the Z value is -0.900 as shown:

I tried to remove the -0.900 and set it to 0.000 which is what I want it to be to no avail. I calibrated both Nozzle Z as accurately as possible but it still changes it back to -0.900. I also made sure the dynamic and safe z were set to zero for N1 and N2. Where else do I look to override this? Any help is appreciated.

Thanks,

Eric Norton President, Velocitronics Motion Systems, Inc. 352-325-2994 | www.vmstech.net | er...@vmstech.net  3 Hemlock Loop Way, Ocala FL 34472

Create your own email signature

[mark maker]

This is not set in the "precise camera ↔ nozzle N2 offsets" step but in an earlier one called "Nozzle N2 offsets for the primary fiducial" where you had to jog N2 down to the primary fiducial precisely and then press Accept.

Like shown here in the video, albeit for a single nozzle machine:
https://youtu.be/md68n_J7uto?t=110

If you did that carefully for both N1 and N2, you should not be bothered by this Z offset, it makes the two nozzles correspond in Z, i.e., it compensates a small mechanical inequality between your nozzles' Z, i.e. it makes sure a certain Z means the same for both nozzles.

I strongly recommend to let it stand.

But if you still want to set it back to zero, then don't forget to press Accept, and it should be gone for good. Otherwise please report.

_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/CAEfPpXzKzhDYb%2BfHPaVGhpy70sVdeB7892wHH0b1gPYxw2yc2g%40mail.gmail.com.

[Eric Norton]

Hi Mark,

 Well that's the thing I wouldn't be bothered by the offset if it picked up the small paper circle and did the test successfully. I went back and re-did the N1 and N2 z offset steps thinking that would solve the problem but the N2 nozzle is still raised up -0.900 from the surface when it goes to pick it up. 

I'm not sure why it is doing this as those previous steps you mentioned would most certainly clear this issue up. Is there any other steps in issues and solutions that adjusts the z axis offsets?

Thanks,

Eric

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/0376c6f7-150a-4486-ad8c-23f4e45d0a41%40makr.zone.

[mark maker]

Hi Eric

this does in deed sound strange...

> I went back and re-did the N1 and N2 z offset steps thinking that would solve the problem

What happened then? Did it confirm the -0.9mm? Can you send the log of that? (OpenPnP saves the last 100 logs)

If you go to the nozzle and set the Z offset to zero and press Apply, will it take it?

Or does it "magically" reappear? Can you pin down when?

_Mark

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAEfPpXzrb4nevvEz_QLzG4p6HWq6aM9aWcLPva-K%2BU7E-8faPA%40mail.gmail.com.

[Eric Norton]

Hi Mark,

I have attached a log and am not sure if it is the one you are asking for. If it is not, let me know where and which one you want.

If you go to the nozzle and set the Z offset to zero and press Apply, will it take it? Yes it accepts it when I click Apply

Or does it "magically" reappear? Can you pin down when? The -0.900 re-appears when I run the Calibrate precise camera ↔ nozzle N2 offsets. The very first time when the test runs it DOES pick up the small paper circle but then the second and other following cycles in the test the -0.900 offset looks to be reapplied and does not pick up the small paper circle. Hope this helps.

Thanks,

Eric

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/60e69f7e-eefd-a5ee-9dce-3670ad16bd76%40makr.zone.

[mark maker]

Hi Eric

> I have attached a log

It appears that when you "re-did the N1 and N2 z offset steps", you got only -0.1mm for N2:

2022-09-11 19:42:57.558 VisionSolutions INFO: Set nozzle N2 head offsets to (33.431684, -74.000922, -0.100000, 0.000000 mm) (previously (33.117077, -73.578623, 0.000000, 0.000000 mm))

So that would explain it not working before, with -0.9mm. The manual nozzle tip capturing needs to be quite precise!

> The -0.900 re-appears when I run the Calibrate precise camera ↔ nozzle N2 offsets. T

I have a suspicion.

The old offsets are recorded when you (or the system) performs Find Issues & Solutions. The offsets are restored, when the solution is Reopened, or when it fails. Normally that's OK, because I&S solutions are sequenced for this to make sense (the solution would not yet be available, so you are forced to press Find Issues & Solutions first).

But now redid the manual offsets and/or you set the Z offset to zero while the I&S list was already manifested. The stored Z offset became stale.

You then had problems with detection:

2022-09-11 19:54:08.432 AbstractMachine TRACE: Exception caught, executing pending motion: java.lang.Exception: Subject not found.
    at org.openpnp.machine.reference.solutions.VisionSolutions.getPixelLocationShot(VisionSolutions.java:1495)
    at org.openpnp.machine.reference.solutions.VisionSolutions.getSubjectPixelLocation(VisionSolutions.java:1462)
    at org.openpnp.machine.reference.solutions.VisionSolutions.centerInOnSubjectLocation(VisionSolutions.java:1341)
    at org.openpnp.machine.reference.solutions.CalibrationSolutions.calibrateNozzleOffsets(CalibrationSolutions.java:1144)

As I said above, when the solution fails, it also restores the old nozzle offsets, hence the stale Z offset of -0.9mm was restored.

To work around this:

1. Set the Z offsets to zero, or better yet to the -0.1mm that you got when you re-did the manual offset capturing.
   
2. Press Apply.
3. Workaround: Press Find Issues & Solutions to store the fresh Z offset in the I&S list.
   
4. Only now redo the "precise camera ↔ nozzle N2 offsets".

(I will try and fix this behaviour in a subsequent version).

However, looking at your log, you have detection problems many times. So I believe there is something not right there too. Perhaps camera settling?

_Mark

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAEfPpXzTYaNV1w4VVh8-7COnDiR0sRkYVLezF0V9SntKgddejw%40mail.gmail.com.

[mark maker]

Hi Eric,

I have fixed the stale offsets problem in the testing version.

See details here (fixed together with another bug):

https://github.com/openpnp/openpnp/pull/1486

@Eric, you would help me a lot, if you tested this:

While the "precise camera ↔ nozzle N2 offsets" I&S solution was already generated, set the Z offset differently, Apply, then Accept the solution, then Reopen it, to see whether the stale Z offset was restored, or the good one.

_Mark

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/5171e9e5-8bed-5cc2-dab6-58d4f1d93187%40makr.zone.

[eric.no...@gmail.com]

Hi Mark,

 

I will try your workaround and report back if it was a success or not. By the way the reason why you see so many fails are because I am adjusting the vacuum speed valve to get the right suction for each axis. When I go to pick that little circle thing up it tends to stick to the nozzle tip because of too much vacuum. Here is a picture of the speed valve:

 

 

I also attached pictures of the machine with all the feeders attached finally.

 

Thanks,

 

Eric Norton Velocitronics Motion Systems,Inc.   Phone: (352) 325-2994  Email: er...@vmstech.net   3 Hemlock Loop Way Ocala, FL 34472   www.vmstech.net

 

I tried to remove the -0.900 and set it to 0.000 which is what I want it to be to no avail. I calibrated both Nozzle Z as accurately as possible but it still changes it back to -0.900. I also made sure the dynamic and safe z were set to zero for N1 and N2. Where else do I look to override this? Any help is appreciated.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/5171e9e5-8bed-5cc2-dab6-58d4f1d93187%40makr.zone.

[eric.no...@gmail.com]

Hi Mark,

 

Ok I will download the test version and test it as soon as I can. I have attached the picture of my machine just to give you an idea of what I have been up to with it and why it has taken so long. The feeders actually took more time than the rest of the machine.

 

Thanks,

 

Eric

 

From: ope...@googlegroups.com <ope...@googlegroups.com> On Behalf Of mark maker
Sent: Monday, September 12, 2022 2:10 PM
To: ope...@googlegroups.com
Subject: Re: [OpenPnP] Nozzle 2 is offset!

 

Hi Eric,

 

I tried to remove the -0.900 and set it to 0.000 which is what I want it to be to no avail. I calibrated both Nozzle Z as accurately as possible but it still changes it back to -0.900. I also made sure the dynamic and safe z were set to zero for N1 and N2. Where else do I look to override this? Any help is appreciated.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/9a3efc3a-f78d-7348-e2d0-46f18430257b%40makr.zone.

[Mike Menci]

Eric - which feeders are these? 

[Eric Norton]

Hi Mike,

 They are heavily modified 0816 feeders with custom board with RS-485 interface, servo PID control and Opulo remix tape take up. Major work was done to get it running.

Thanks,

Eric

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/fe6b2367-dc30-4eab-adf2-3c58ec9fb7e9n%40googlegroups.com.

[Mike Menci]

Thanks Eric - sorry to disturb your offset subject. Let us have some more details when you get some time and I&S things sorted out!

Good luck 

Mike

[Eric Norton]

Will do! I can't give out source code or board design because I plan on selling them when I get around to it. I may release the feeder mechanical solid works files though.

Thanks,

Eric

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/29a7141a-aacf-42e9-9214-3354c996d2c3n%40googlegroups.com.

[Mike Menci]

Well Noted Eric no problem. 

Cheers, Mike

[Eric Norton]

Hi Mark,

 I have tested the new testing version and all is good now. I can confirm that the good z offset replaces the stale value and I can successfully pick the circle piece for Nozzle 1 and 2. Thank you very much for your help with this. Hopefully I won't have any other issues :-).

Thanks,

Eric

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/9a3efc3a-f78d-7348-e2d0-46f18430257b%40makr.zone.

--