How nozzle concentricity deviation has influence on placement position

SMALLSMT Forum � Vision Placer software � Vision Placer software V2 & V1 � How nozzle concentricity deviation has influence on placement position

Topic settings

Print view

Jump to forum

Forum SMALLSMT Forum Hardware questions VISION PLACER Hardware questions SMT VISOR Hardware questions HR-300 Reflow Oven Found on the internet Vision Placer software Vision Placer software V2 & V1 Vision Placer V2 Bug List Tips & Tricks DotMaker Software PnPCreator Common area SMT related questions User talk Marketplace Electronics Kendryte K210 RISC-V Controller

#1 | How nozzle concentricity deviation has influence on placement position datum

Sun Jan 17, 2021 11:54 pm (Last edited: Fri Jan 22, 2021 8:34 pm)

I have not tested the new version 2.060 yet. Returned to version 2.057. There are fewer questions with her.
Found the reason for the displacement of the center of the nozzle relative to the center of the lower camera.
This happens on all nozzles and on two heads. On large nozzle, starting with CN140, this is noticeably less.
To see this effect, you need to perform the following sequence of actions:
1. Transfer head 1 (with CN40 nozzle) to the lower camera 1;
2. Center it on the camera using the controls for the X and Y axes;
Nozzle Rotate 1.png - Bild entfernt (keine Rechte)
3. Set the speed template XY100, Z50, A1600, F100. It can be different, but on this template the head is rotated 90 degrees, which is convenient and noticeable.
4. Press the head turn button a few times.
5. The result I get is the same as in the photo. The deviation is 1-1.2 mm.
Nozzle Rotate 2.png - Bild entfernt (keine Rechte)
Nozzle Rotate 3.png - Bild entfernt (keine Rechte)
Nozzle Rotate 5.png - Bild entfernt (keine Rechte)
Nozzle Rotate 4.png - Bild entfernt (keine Rechte)

link to video with nozzle:

link to video with component:

It turns out that the nozzle rotates not along a point, but in a circle with a diameter of 1-1.2 mm.
There is no problem with the nozzles. For this check, before taking the nozzle from the nozzle changer, I turned head 90 degrees. And as soon as this was done, then if you move the nozzle to the center of the camera, the nozzle will already be offset from the center of the camera.

Head rotate photo on camera 1
Head Rotate 1.png - Bild entfernt (keine Rechte)
Head Rotate 2.png - Bild entfernt (keine Rechte)
Head Rotate 3.png - Bild entfernt (keine Rechte)

I assume that the problem occurs of misalignment at the point where the nozzle is attached to the head. See photo:
2021-01-17_11-20-35.png - Bild entfernt (keine Rechte)

I believe that this effect will affect the accuracy of the selection of components from the feeder, the accuracy of the placement of components on the board, and absolutely on all calibrations machine .

Thus, the calibrations of the upper and lower cameras, of the second nozzle, relative to the first nozzle, must be performed taking into account a certain average value of the position of the first nozzle. Those. during the calibration process, turn the nozzle 90 degrees and watch its deviations relative to the element that is being calibrated.

Michael, please explain this fact. Is it possible to do something about it?

Also, I ask other users of the machines to check this with themselves and write about the results.

reply

#2 | RE: Bug Nozzle datum

Mon Jan 18, 2021 6:48 am (Last edited: Mon Jan 18, 2021 6:52 am)

You are describing nozzle concentricity deviation.
This is normal and has no effect during part placement.
Nozzle concentricity deviation has a effect during part pickup but during camera alignment this gets compensated. After Alignment no nozzle rotation applied.

It depends on nozzle mostly and if you want to calibrate camera center need to use a nozzle with low concentricity deviation or correct center after placement deviation measured.
Original Samsung/Hanwa nozzle have less deviation.
The cheaper clones not selected for this parameter.

Some improvements could be done by adding some grease to nozzle shaft. If deviation depends on nozzle seat worn out nozzle seat need to get replaced.

I think your deviation is not 1mm it�s more like 0.1 to 0.3mm if really bad nozzle.

reply

#3 | RE: Bug Nozzle datum

Mon Jan 18, 2021 10:26 am (Last edited: Mon Jan 18, 2021 10:27 am)

Zitat von mbruch im Beitrag #2
You are describing nozzle concentricity deviation.
This is normal and has no effect during part placement.
Nozzle concentricity deviation has a effect during part pickup but during camera alignment this gets compensated. After Alignment no nozzle rotation applied.

I agree, but only partially.
1. The fact that this affects the pickup of the component from the feeder is unambiguous. And this explains why I saw the offset on pickup from the feeder when the top camera was set to the center of the component.
2. Regarding the alignment of the component on the camera, this effect will not appear if the component does not rotate before it is aimed at the lower camera from the feeder.
3. If the component is rotated 90, 180 or 270 degrees before being moved to the bottom camera, then I think this effect will affect the alignment.

Zitat von mbruch im Beitrag #2

It depends on nozzle mostly and if you want to calibrate camera center need to use a nozzle with low concentricity deviation or correct center after placement deviation measured.
Original Samsung/Hanwa nozzle have less deviation.
The cheaper clones not selected for this parameter.

What is the real deviation for original nozzle?
And how can you tell the original from the clone?

Zitat von mbruch im Beitrag #2

Some improvements could be done by adding some grease to nozzle shaft. If deviation depends on nozzle seat worn out nozzle seat need to get replaced.

Please show in a photo or video where and how to apply the lubricant. And what kind of lubricant should you use?
What is the resource of the nozzle?

Zitat von mbruch im Beitrag #2

I think your deviation is not 1mm it�s more like 0.1 to 0.3mm if really bad nozzle.

Typo. Deviation 0.1-0.12 mm.
But I think that's a lot.

reply

#4 | RE: Bug Nozzle datum

Tue Jan 19, 2021 9:37 pm (Last edited: Tue Jan 19, 2021 9:39 pm)

Zitat
3. If the component is rotated 90, 180 or 270 degrees before being moved to the bottom camera, then I think this effect will affect the alignment.

It couldn't influence because the alignment use steps to adjust.
After each step it measures deviation and adjust again.

Zitat
Typo. Deviation 0.1-0.12 mm.
But I think that's a lot.

You're wrong that's not much!

Zitat
What is the real deviation for original nozzle?
And how can you tell the original from the clone?

If we buy we can't find any difference in marking i think it has only better selection for concentricity deviation.
Original nozzle has deviation < 0.1mm but 0.1mm is already good remember there is a spring inside has influence too.

Zitat von u.m. im Beitrag #3

Please show in a photo or video where and how to apply the lubricant. And what kind of lubricant should you use?
What is the resource of the nozzle?

You can use same lubricant as used for guide lines and I told already apply to shaft it is the part placed inside nozzle seat!

reply

#5 | RE: Bug Nozzle datum

Tue Jan 19, 2021 10:56 pm

Zitat von mbruch im Beitrag #4

It couldn\'t influence because the alignment use steps to adjust.
After each step it measures deviation and adjust again.

I will try to clarify. On the camera, of course the component will align in a few or one step. But the fact is that the displacement coefficient would be different if the component came to the camera without rotation. Those. the nozzle follows the camera with the coordinates of X_cam and Y_cam, and instead comes to coordinates X'_cam and Y'_cam . Where X'_cam = X_cam + X_nozzle_error and Y'_cam = X_cam + X_nozzle_error. The component is aligned on the camera where there is a nozzle offset error in its corrected coordinates.
Perhaps I do not understand correctly. Can you explain the alignment process on the camera?

reply

#6 | RE: Bug Nozzle datum

Wed Jan 20, 2021 6:41 pm (Last edited: Wed Jan 20, 2021 7:22 pm)

Short version is the part center in vision area gets detected by moving part in center of area.
The visual offset calculated to modify placement position.
Placement angle need to get adjusted during alignment, there is no additional offset applied during placement.

X_Place + Vis_offsetX / Y_Place + Vis_offsetY

Vis_offset is the deviation of part from visual center.

I understand your idea but nozzle center deviation can not get compensated for high precision you need good nozzle.
The big nozzles like CN400 mostly have very low center deviation.
And for nozzle over camera center alignment i rotate nozzle and adjust to median center.

reply

#7 | RE: Bug Nozzle datum

Thu Jan 21, 2021 10:41 am

I drew pictures.
I wanted to clarify if I understand the error that is occurring correctly?
And it turns out that the accuracy of the machine is more dependent on the centering of the nozzle with the head?
Name.png - Bild entfernt (keine Rechte)
Step1.png - Bild entfernt (keine Rechte)
Step2.png - Bild entfernt (keine Rechte)
Step3.png - Bild entfernt (keine Rechte)
Step4.png - Bild entfernt (keine Rechte)
Step5.png - Bild entfernt (keine Rechte)

Zitat von mbruch im Beitrag #6
And for nozzle over camera center alignment i rotate nozzle and adjust to median center.

The last time I did it. But the documentation does not say about it.
But so far, with these changes, I have not collected boards.

reply

#8 | RE: Bug Nozzle datum

Thu Jan 21, 2021 1:24 pm (Last edited: Thu Jan 21, 2021 1:29 pm)

Zitat
And it turns out that the accuracy of the machine is more dependent on the centering of the nozzle with the head?

I think it's logically right to know the nozzle center deviation has influence.
But as I wrote before for high precision placements we need a low concentricy deviation.

I think it should be possible to use the CP45 without springs. https://www.daftonsmt.com/samsung-cp45-n...-j7055267c.html
TN030, TN040, TN065 for example

But second thing i repeat again it's not necessary to place small chip components on such precision.

Most important is that your big nozzles have low concentricy deviation.

reply

#9 | RE: Bug Nozzle datum

Thu Jan 21, 2021 4:32 pm

The spring has a useful property, it works as a protection if you make a mistake with the height. And without a spring, it's somehow dangerous. The head may not return to its place if it cannot push through.

reply

Bug in triode optical recognition method? � � SW V2.058 end of backward compatibility

Related topics

Replies

Last post

No way to place large nozzles components first.
Created in forum Vision Placer V2 Bug List by hjalamr 1 Sat Aug 28, 2021 9:13 am
mbruch • Views: 231
Change the calibration "Mark Camera To Nozzle 1"
Created in forum Vision Placer software V2 & V1 by u.m. 6 Sat Sep 26, 2020 10:28 pm
u.m. • Views: 443
"Library has nozzle"
Created in forum Vision Placer V2 Bug List by Bienja 4 Fri Dec 04, 2020 10:06 pm
mbruch • Views: 388
Wrong placement speeds with 2 heads
Created in forum Vision Placer software V2 & V1 by Louis 1 Fri Mar 06, 2020 6:16 pm
mbruch • Views: 359
2 heads with nozzle changer
Created in forum Vision Placer software V2 & V1 by hjalamr 1 Mon Feb 03, 2020 7:40 pm
mbruch • Views: 472
How to calibrate mark camera to head1 position distance
Created in forum Tips & Tricks by mbruch 0 Sun Nov 17, 2019 11:44 am
mbruch • Views: 506
[V2.046] A2 losing nozzle after pickup from nozzle changer
Created in forum Vision Placer software V2 & V1 by gj 2 Sun Nov 17, 2019 12:08 pm
mbruch • Views: 539
Parking position after placement job
Created in forum Vision Placer software V2 & V1 by Louis 14 Mon Oct 28, 2019 12:59 pm
mbruch • Views: 554
Problem with placement stopping permanently in the middle of a run
Created in forum Vision Placer software V2 & V1 by MtnMan 11 Thu Aug 10, 2017 7:58 am
mbruch • Views: 1209

SMALLSMT Forum

Please enter a reason for warning.