Placement of integrated circuits

[JuKu]

> What is the resolution of your cameras?
640x480.

> Is it possible to install cameras with higher resolution?
- what would be the boundries for that?

Yes, and some users have done that. The software uses standard drivers, so Mechanical issues apart, about any camera should work.

> Would I get any advantages from a higher resolution in combination with the actual software?

Maybe. Vision accuracy (mm/pixel) improves. The basic cameras already have resolution at or below motor resolution but error sources accumulate. Also, better cameras would have less noise, which cold improve the visual fuction performance. Handling larger datastream might slow things down (depending on your computer). View area and focal length could be an issue, depending on camera. Even if there may be no difference in actual performance, user view would look nicer.

[thereza]

higher resolution results in lower frame rates and the visual anslysis messes up sometimes because you are looking at where the camera was not is if you are doing a lot of activity. the code supports it, but by default it just uses the lower resolution which seems to work fine.

[mawa]

In many occassions the image ist zoomed. So using a higher res cam could give an improved zoom result.

But due to the fact that the zoom is not done by the camera but is applied to the full scale image by software.

The image processing load rises by the product delta of the X Y image pixel count.
If you want short feedback e.g. recognizing and correcting the rotation angle of a component by the up cam you need a high frame and processing rate.

As fas as I have experienced by now the current resolution is sufficient for the current purpose as designed by Juha.

There is certainly an improvement potential and I hope that more liteplacer developers with image processing know how join in.

[thereza]

one thought is to add a filter which changes the zoom of the camera. this should be easy to do. though still not sure if there is a need.

[WayOutWest]

higher resolution results in lower frame rates and the visual anslysis messes up sometimes because you are looking at where the camera was not is if you are doing a lot of activity.

In many occassions the image ist zoomed. So using a higher res cam could give an improved zoom result.
But due to the fact that the zoom is not done by the camera but is applied to the full scale image by software.

All of this is correct. But I would like to note that all USB webcams can be told to only send back a particular rectangular region of the visual field or to downsample it before sending over USB -- it's part of the generic USB video device protocol that all USB webcams are required to support.

Now, whether or not it's possible to access that command through all the layers of software (driver, windows, .NET vm, vision library) is a totally different story... but the hardware does have this capability, so a higher-resolution camera can always be "at least as good as" a lower resolution one if you have the right software.

Also if you got the Andonstar camera make sure all your hubs support USB 2.0 High Speed (not just "USB 2.0"). The camera will actually pick a very lossy encoding (MJPEG, it seems, the blocky DCT artifacts are visible) if it can't establish a 480mbit/sec connection to he host.

[cheeef]

What is the current status regarding placement of IC's like TSSOP, TQFP, QFN or BGA?

Is there already an option to place such components? The machine looks really interesting
but without the option to place such components either by manual alignment or even automatic
placement it doesn't seem to be usable yet.

Thanks for the update.

[Spikee]

I have not tested tqfp / bga etc... but you can manually align the parts and jog the machine to the correct location and place them via the computer.

Yes it can automatically do it but the question is if the accuracy is good enough
For small pitch qfn and bga I would do it manual style. With some improvements qfn (and maybe bga) will be possible.

[dc37]

Is there any update on the status of TSSOP / smaller-pitch parts?

[mrandt]

Not sure what Juha's point of view is on this...

As a LitePlacer user, I believe the mechanics are accurate enough to place at least 0.4mm pitch IC; maybe even smaller - many BGAs should work as well.

The issue is missing functionality in control software: It should allow us to leverage the uplooking table scanner camera as part of the placement process.

After picking up an IC or other larger part, nozzle should move the part above the camera and bring it up to the correct Z-level (to avoid focus issues and perspective errors).

Next, the table camera should be used to detect the part, center it and measure its rotation.

When placing the part onto the PCB, XY (needle position) + A axis (rotation) should be automatically corrected based on the measured offsets.

This would not only be relevant for ICs or BGAs but also other larger parts which need to be well aligned and accurately placed. For most small parts and passives a few degrees of rotation and a bit of offset do not matter much - but for others straightening these out is vital.

Currently, some of the available software forks have slightly better functionality than Juha's base version, for example detection of a part's body based on template matching. This works well for QFN (no leads ICs) and a few others but uses the downfacing / flying vision camera. But be warned, there is quite a bit of diversity (to put it mildly) and of course Juha does not support those software versions.

I do believe it is just a matter of time until this will work as the hardware is in place - but I cannot tell how long that might take.

[martin123]

I just made an order for the Liteplacer and it would be great if this got implemented.
There is a lot of fragmentation in the software. Does any of the unsupported software handle IC placement?