Hello,

What is the recommended latest stable version to use for a basic VP2800HP machine? It looks like V2.065, V2.064,... changes don't apply to my setup.

Thank you & regards, gj

Thank you. I followed the manual, copied the 4 files from my existing installation, then started + connected. Looks good but when I load a previous project and try to place parts, I get "96:[ Line ] No feeder assigned!" error - regardless of strategy (camera simuate fails with same error). Also, it makes no difference if I place a single part or the whole project. All lines in the project have feeders assigned & I can open the part --> open feeder --> move (camera) to pickup... all working fine. Re-assigning the feeder number also doesn't help. Any hints?

Today I changed version 057 to 065b at 2800HP.
After the version was changed, Gxx feeders became known as Sхх.
But, after I made changes to the feeder coordinates, the name Gxx returned back.
Could this was the issue?

Thanks for the pointer but I don't see Sxx here. The feeders are all named as they were before the update. Also, all interaction with the feeders is normal - pick up parts & visual alignment all works fine. I also tried changing coordinates... no luck.

Should I try with a different SW version or is there another way to import previous feeder config into a new setup other than copying the feeder.fig there? Any other advice?

What group of feeders did this happen to?
In my current work feeders of all groups - W, N, E, G and vibrating, designated as W40, are used.
After changing the version, except for the described incident, nothing was broken.
But, it seems, in the PNP file, I changed the G group feeder in the corresponding line manually when correcting the file.
Looks like I got lucky ... )))))

I use version 2.057 as the most stable for me.
After several unsuccessful updates, I no longer want to try them.
Updates are mainly about new machines. And on versions of VP2800HP and below, there are constant bugs.

Yes, I also reverted to version 2.057.
When working during 20 hours on version 2.065c, the machine lost coordinates for two times, due to which the nozzle was damaged.
When working on version 2.057 during 110 hours, there was never a loss of coordinates.
In addition, optical alignment of components in version 2.065c performs worse than in version 2.057, unreasonably discarding 3 to 10 percent of components.

I will add to the previous post.

What I did for normal board assembly on VP2800HP:
1) Version software Vision Placer 2.057
2) Disabled the marker point alignment on the machine in Basic Settings. Always used mechanical 0 as a starting point.
3) Disabled alignment to Marker Points on the board. And for each new board I always adjusted the initial coordinates of the board (0 board).
4) The coordinates of the components on the board, I imported so that the starting point of the board is equal to the coordinate of the Marker Point.
5) I set the coordinates of the Lower Cameras taking into account the deviation of the nozzle from the center. As the mean of the deviation. To do this, I turned my head on the camera along with the nozzle and watched the deviation from the center.
6) The distance between the head 1 and the Top Camera, the same taking into account the deviation of the nozzle from the center. For this I used a marker point on the board. Determined the coordinates of the Marker Point by the Top Camera with optical alignment. Then he lowered the nozzle onto it and rotated it too.
7) Distance between head 1 and head 2, the same taking into account the deviation of the nozzle from the center. As in part 2.
8) Changed the accuracy of the machine. It is equal to 0.02 mm. By changing the parameters in the settings in the servo driver settings and program settings VP. I also set all the X and Y coordinates with an accuracy of 0.02 mm. Because the step of the machine became 0.02 mm.

After all these actions, basically there are only problems with installing LIFP cases. I often correct them under a microscope before soldering.

Little notes:
1) I also noticed that the image on the Lower Camera 1 and Camera 2 is slightly different for the same component. Because the setting for them is the same and the illumination is done differently.
2) I think the Top Camera is just awful for such good mechanics in the X and Y axes.
3) The mathematics of calculating the coordinates is single precision (float), which leads to errors when rounding the coordinates. At least in the program window. Therefore, I moved the car to a step of 0.02 mm.
4) It seems that when analyzing the image from the camera, Erosion and Dilation are not used. Which also affects the alignment of the component and its placement.

So far, I have only come to the conclusion that it is at least desirable to change the lower video camera from newer machines. As a maximum, write your own program.
Unfortunately, nothing can replace the top video camera.

Sorry for my english.

Typo, not LIFP case, but LQFP case.