LumenPnP Populating 0402

[@miguelflores-acton8581]

These are awesome, so nice especially if standard assembly services don’t have the parts you want

[@Factory400]

This is the key to make this system useful for my low volume commercial work. My existing system is huge, heavy, complex, and very difficult/expensive to support.

[@ferologics]

absolutely amazing, i can’t wait to see what comes of this project in the future!

[@DadofScience]

Almost as fast as me by hand! 😏 But seriously, awesome tech, Stephen & Co. Makes me think we could make a small fortune developing and selling a silent vacuum pump 🤔

[@Echotronic.network]

Very nice, i am a happy LumenPnP owner, can you provide the vision pipeline used in this demo ?

[@opulo]

Yes! For each type of pipeline we have all the most recent settings available on our docs page, easy to copy paste into OpenPnP: docs.opulo.io/openpnp/vision-pipeline-adjustment/1-introduction/

[@Echotronic.network]

@@opulo Thank you so much !

[@mitchellhw2006]

I stick to 0603 as the minimum size (if forced past 0805) but great too see the LumenPNP can snatch up those evil little grains of 0402 dust.

[@GodotDev-mx4yr]

Could you please develop on which tool/method you personally recommend to use for the solder paste deposition right before that PNP operation ? Do you do it with common solder paste through a stencil with something like a solder paste printer (If so, do I need to first apply the solder paste, heat it, then cool it, so that each footprint have solid solder on them, then position the SMD component with the PNP, then reheat to bind the solder to the components ? Or can I PNP on the solder paste directly ?)

[@opulo]

First, solder paste is squeegeed on the board with a stencil. Then it's mounted on the LumenPnP and all the parts are populated into the paste. Lastly, the board is put in a reflow oven where the paste heats up and solidifies, connecting the parts to the board. We have a blog post about the equipment we recommend for the process!: www.opulo.io/blogs/articles/setting-up-a-mid-scale-smt-line

[@malandr1]

@@opulo Thanks a lot ! 🙏

[@stormbreaker932]

Impressive! I remember people saying on the discord that 0402 was difficult with the stock nozzle. Has there been some changes hardware/software wise that have enabled this to be more reliable? 0402 is essential for my use case 😅

[@opulo]

Yes, we're always making lots of firmware, configuration, and hardware changes! The v1.0.3 release of the feeder firmware in particular was a tremendous improvement.

[@michaelpxd6325]

Is this the Lumen v3 or v4? Asking because of the speed (for reference)

[@opulo]

When this video was shot, it was done on a V3.1, so you should expect the V4 to be faster.

[@rodguinot3024]

Which version of OpenPNP and which OS did it run on?

[@opulo]

We used our recommended version of OpenPnP, which is located on our docs page (2023-04-05 release). This demo runs a workaround in OpenPnP that allows for concurrent feeding and movement. There's work in a separate fork with the same functionality that has not yet been merged into mainline OpenPnP. The OS was Linux. We specifically use Ubuntu v20.04 LTS

[@Sulzi86]

Why does the camera window say that the component is a 0603 part?

[@opulo]

The board that was used is designed for 0603. That board is the one that we include with every LumenPnP sale for testing the machine out after they get it assembled and tuned.