In version 058 from 12/30/2020, very useful functions have been added when aligning components.
But, at the same time, there was appeared an unpleasant error related to the fact that the dimensions of the component specified in the parameters are not taken into account when aligning it using the camera (Video: https://youtu.be/ZfKWT1gJHhg). This bug in version 058 dated 01/05/2020 is partially fixed, but not completely (Video: https://youtu.be/jEYeDc_sggQ).
In addition, in version 058 dated 01/05/2020, the return of the machine to "0" after the installation of the selected component disappeared (Video: https://youtu.be/jEYeDc_sggQ).
Also, starting from version 056 or 055, that the vacuum has stopped auto turning off when performing the Nozzle To Throw Position operation, which is very inconvenient when resetting a component after adjusting the alignment.

The first video contrast setting looks really low did you test it before? The error could only get solved by stopping machine and adjusting well.



The day before, using version 057, 20 boards were mounted with 100 of these components. None of the settings were changed when the version was updated.
All components with the thise settings were placed without any questions.
Moreover, the boards were plced with version 057 precisely because of the described error in version 058, although the component parameters were configured using version 058 dated 12/30/2020.
This error also manifests itself not only on the component shown in the video.

Thank you Michael.

I am reporting.

1. The error with the size of components, at first glance, has been eliminated.

2. The vacuum is automatically turned off when the component is dropped on Throw Position. If we could also bring the state of the vacuum
checkbox on the screen into concordance (it does not turn off when the vacuum is automatically turned off), it would be great.

3. Numerical data in the alignment window of the Visual Original is still smeared (have an incorrect font size ?).

4. There is still no return to "0" of machine after mounting the selected component.

I can test it in more detail within a few days when will mount boards.

Zitat von SergOv im Beitrag #5

---

3. Numerical data in the alignment window of the Visual Original is still smeared (have an incorrect font size ?).

---

I noticed 2 problems.
1. If Visual Origin is enabled in Basic Settings and if press the Reset button, then the Top Camera always does not reach the point a little. After the alignment takes place. Moreover, the coordinates in the Basic settings are not updated. The displacement error is along the X axis = 0.032 mm, Y = 0.0741 mm.
Link to video:

Zitat von mbruch im Beitrag #9

---

First point you have simple analog cameras.
The precision for mark points depends on camera resolution. So machine has 0.015mm resolution and reaches so small deviation i think that’s not bad.

---

Zitat von mbruch im Beitrag #10

---

By the way if you increase mark point diameter you can improve precision too!

---

Zitat von mbruch im Beitrag #10

---

I forgot is there only box size reduced or real vision output too?
Because box drawing is cosmetics but if visual alignment picture cropped too it could be a problem.
I saw vision output is cropped too.

---

Michael, I chose your machine.
And the characteristics of competitors are of little interest to me. If they are interested, only as self-education.

I am a perfectionist at work. And I want to understand how and what works. Where is what accuracy depends on what. Understand how you can improve the build result. I need the machine to always work equally well for me. But so far I cannot achieve this. Each time I collect new boards in a different way. I cannot achieve the same result. This is very frustrating and time-consuming. Some installation flaws are corrected by the reflow oven. And some she just can't fix.

My main components are 0603 and LQFP-64 and LQFP-100. And in my opinion, the machine does not cope with them. I suppose that it does not cope, not because it cannot recognize, but because the initial conditions for the placement themselves slightly change. Or impulses are lost. Or the float math gives an error. Or the coordinates of the lower camera go away. I dont know.

What works for other users is not reflected in my machine in any way.
Now the temperature in the room is stable with the air conditioner running + 24 degrees, in the room - 18..20 degrees. On the machine the temperature in the aisles is 20..22 degrees.
The speed is used just below average for the X and Y axes and low for the Z axis. For LQFP chips, the pitch is 0.5 mm, the chip speed is used.
And to be honest, sarcastic and promotional responses will not help my machine run well. I am asking for help, not ridicule or advertising.
Some photos taken yesterday
3.1.png - Bild entfernt (keine Rechte)
3.2.png - Bild entfernt (keine Rechte)
3.3.png - Bild entfernt (keine Rechte)
3.4.png - Bild entfernt (keine Rechte)
3.5.png - Bild entfernt (keine Rechte)
3.6.png - Bild entfernt (keine Rechte)
3.7.png - Bild entfernt (keine Rechte)
and today
1.1.png - Bild entfernt (keine Rechte)
1.2.png - Bild entfernt (keine Rechte)
1.3.png - Bild entfernt (keine Rechte)
1.4.png - Bild entfernt (keine Rechte)
1.5.png - Bild entfernt (keine Rechte)
2.1.png - Bild entfernt (keine Rechte)
2.2.png - Bild entfernt (keine Rechte)
2.3.png - Bild entfernt (keine Rechte)
2.4.png - Bild entfernt (keine Rechte)
2.5.png - Bild entfernt (keine Rechte)

Photo of the nozzle on the 1st head for some time. The transition is always from mechanical zero.
1.png - Bild entfernt (keine Rechte)
2.png - Bild entfernt (keine Rechte)
3.png - Bild entfernt (keine Rechte)
4.png - Bild entfernt (keine Rechte)
5.png - Bild entfernt (keine Rechte)

Tomorrow I will try to assemble these boards on an older version of the program. Because before, component placement was better.