ICC-1 breadboard friendly breakout board

[oliv3r]

This patch adds a new ICC-1 module, which is the real name as can be obtained from the FCC registration documents, Trådfri being the whole 'system' name. It uses the dimensions from the data-sheet which corrects some dimensional errors in the previous module. Further more it converts the supplied cad file into a KiCAD 3D part.

Secondly this patch introduces a carrier board for the ICC-1 poly module in a breadboard friendly manor. It addresses most comments from issue number #12.

I had a look at your TRÅDRI dev board, and I had a few comments and suggestions that I think would allow the board to be more versatile and more generally useful:

• [x] Redesign the board in such a way that allows it to be plugged into a breadboard if the headers are soldered facing down instead of facing upward.
• [ ] Remove the reset button. If you need a reset button for a project, it is easy to add on a breadboard. Removing the button should allow you to shrink the size of the board considerably, allowing it to more easily fit on a breadboard. One can always decide to not populate the button, the board won't get any smaller anyhow.
• [ ] Remove the power LED. It is easy to add a power LED on a breadboard, and having a LED on this board will throw off power measurements. One can always decide to not populate the LED.
• [x] Remove the bypass caps. All of the capacitors that are required by the EFR32 datasheet are already integrated into the module.
• [x] In order to be as conservative as possible, remove all of the copper fill directly under the module. Copper fill on the other side of the PCB might be fine as long as it is kept far away from the antenna, but many modules like this explicitly forbid copper fill anywhere under them.
• [x] Leave the antenna hanging off the end of the board
• [x] Keep the debug headers. Those are often useful, and having to always hook them up manually on a breadboard would be a pain. also, one can decide not to populate those either
• [x] Shrink the board to be as small as possible, to help make usage with a breadboard more easy.

Further more, I used the area available as breakout area and made the board so that it is 'breakable' into sections. The debug headers can go by cutting along the via's. The breadboard section can go by cutting along the via's.

Closes issue #11 and #12

Signed-off-by: Olliver Schinagl oliver@schinagl.nl

[oliv3r]

Any thoughts on this so far? @darconeous how do you like it so far? Does it address your issues?

[darconeous]

Neat!

Not exactly what I had in mind, but it fits the bill! It sure does allow for more breadboard space.

I personally would have made the board rectangular (with the module in-between the breadboard pins) to improve the production yield during manufacturing, but I think your design is great if that isn't a concern.

The image you posted makes it unclear to me if the antenna has PCB behind it or not.

[oliv3r]

@darconeous I've added a picture of the backside, which allows you to see the mouse bites more clearly. There's mouse bites at the top and via's just above the jtag connectors that could work as guides/mouse bites to more easily cut off the bottom bit.

I did it for two reasons. It allows pick-and place assembly (not sure for what :p) but it adds fiducials. More importantly, it adds room for the PCB fab to print their production data, without getting in the way. Oh and it helps with panelization :)

That said, I don't think there's much point to optimize mass production as to the size. If we produce these at all, i'd be supprised if more then 20 get distributed. 20 being the magical number as we can fit 2 boards in a 10x10 space easily. MAYBE even 3 boards.

As for the dimension of the board, there was not much width to go with. The bare ICC-1 module takes up a lot of the breadboard. My guesstimate was that it would leave 3 holes max. So 1 row at one side, 2 at the other, which made it far less usefull. Finally I wanted to be able to power it with the power rails on the side of the breadboard.

With those constraints a T-Shaped board seems quite obvious. Put it on a breadboard, put a breadboard PSU on one end, this board on the other and go.

The proto typing area was a bonus, as that room was unused, so I just filled it up with 2.54 mm spaced proto bits.

[oliv3r]

@basilfx any other feedback you want to add or any rework you need for this to get merged?

[oliv3r]

@basilfx ping?

[CableCatDK]

Hi Oliv3r

I am now using your KiCAD footprint of the ZigBee module for my project. First i design my own, which was a little wrong. Then I used basilfx's, which i also a little wrong. You version look fine. Not tested my me yet.

My project: https://www.thingiverse.com/thing:3655354

Do you need to be credited for using the footprint?

[oliv3r]

I'm still waiting for basilfx to merge it :) I have not ordered the pcb yet myself (too busy) and so I haven't verified it's validity. I took the measurements from the datasheets from the FCC database. I first tried to manually measure it and already noticed discrepancies with basilfx's module, which is why I re-drew it.

Having said all that, I'm not sure what the license of the repository is, but credit is usually appreciated ;)

Out of curiosity, how did you figure out the size was wrong? And how 'perfect' does it fit? :)

[CableCatDK]

The footprint basilfx has made is about 1.1mm too long, and was almost usable. I made a correction, which is only 0.1-0.2 mm too long. Your footprint is a little smaller than mine. So expect it to fit perfectly. I have ordered PCBs with it now.

Did you make the footprint from scratch? or did you reuse some of basilfx work?

[oliv3r]

I started initially with basilfx's version, but when measuring the ICC-1 module with a caliper, I found a lot of things to not align well. The pins and the package. As I was very into it back then, I by chance found the FCC documentation, which lists the exact dimensions, so I started from scratch, as there where a few other details off.

While I haven't tested it on a PCB, i did check it on paper by printing it, and it looked fine there.

If all works out and fits on your PCB's, can you drop a note here, so I can poke @basilfx to merge this PR?

[CableCatDK]

I have now tested your footprint by making 30 of them. The height of the pads fits exactly. But you need to move the pads 0.5 mm to the center. That is, the left side need to move 0.5 mm to the right, and the right side 0.5mm to the left. Your mistake makes it easy to handsoder, but not good for pick-and-place mass production.

[oliv3r]

Not sure if it is a mistake, or if this intended as the purpose is to handsolder for this issue's project :) If we extend it 0.5 mm to the center, will both use-cases be helped? Also, it could be that indeed, that this was a mistake; it's been quite a while ;) So I'll extend them, allowing both work flows. Maybe have 2 variants, icc-1 and icc-1w (wide) which is useful for hand-soldering ...

[CableCatDK]

Extending the pads sounds fine. You should summit you work to the general KiCAD library, or what is called. I am new to KiCAD.

[weakfl]

So I'll extend them, allowing both work flows. Maybe have 2 variants, icc-1 and icc-1w (wide) which is useful for hand-soldering

@oliv3r it would be great if you could push a final design. I'd love to get some PCBs made for testing…

[oliv3r]

I only last week looked at the foto's from @CableCatDK and saw exactly what he meant. So I am in a better place to extend the pads. I'll try to get this sorted in the next few days.

[CableCatDK]

I only last week looked at the foto's from @CableCatDK

@oliv3r Which photo are you referring to?

[oliv3r]

I only last week looked at the foto's from @CableCatDK

@oliv3r Which photo are you referring to?

https://thingiverse-production-new.s3.amazonaws.com/assets/ea/e7/4d/06/f3/E2D6EAA4-0377-4F91-B6E0-32FBA3AE5063.jpeg

[CableCatDK]

@oliv3r that is photo of my own footprint. Which was my first try.

[oliv3r]

Ah right, but from the 'width' point of view, are the pads now as your picture, too wide, or contrary, too narrow and they need to be elongated? You mentioned moving the pads 0.5 towards the center, e.g. the pads are 'too short'. Your image would suggest the same, that the pads where too short. Maybe you should update the picture in your thingieverse repo with v2 :D

[CableCatDK]

@oliv3r In the link below you see a picture of the current version. Which is based on your footprint, but the pads are moved 0,5 mm to the center. https://thingiverse-production-new.s3.amazonaws.com/assets/93/59/e3/54/96/70AEB275-758A-43BE-B18D-1381F4E5B709.jpeg

[oliv3r]

did you move the pads in the picture already? If so, that explains my confusion :)

[CableCatDK]

@oliv3r Yes, I have moved the pads in the last picture.