Honestly I'd like to keep the PCB as simple as possible - it'd be Open Source so easy enough for anyone to get a new PCB spun with (for instance) LoRa on it if they wanted. This one would just be to get people started.
Adding a rechargeable battery solution is again really difficult - this thing will run for a year on a CR2032 - and probably at least 10 years on one of those non-rechargeable Lithium AA batteries - so I wonder how big a deal it really is.
Honestly I'd like to keep the PCB as simple as possible -
I agree, a PCB that just works.
But can a PCB just work and allow for something else with a layered design of building blocks.
ie whatever Gordon is comfortable is the best, and whatever happens can happen if more comfortable with whatever is created.
As long as all a lot of IO is brought out and the footprint can be easily uses in Eagle then people can make their own boards.
Although there's a lot to be said for just making a PCB that had the Arduino footprint? There might be enough spare space to get 1x Arduino and 2x DIP PCBs in the area of one Arduino PCB.
Arduino footprint has many merits : I am evaluating on the nRF52-DK.
However, it could be "nice" for a tight PCB with the io bought out into an easy expansion.
It could be nice to have a smaller Arduino io footprint if possible: the shield is quite big and a tight PCB is definable as small if this is what one wants.
I have to look thru the different versions of a standard Arduino footprint to see if there is enough on a smaller footprint ie tighter PCB
Yep: a pcb with full footprint with some thought to allow for a reduced pcb by reuse of a well thought out placement of components in relation to the risers on the board.
Any cubebit.js ideas can buy (some of) the stuff from espruino and tailer to own wishes.
The pic is posted before somewhere, but shows that when one has something in the hand to look at, one can better feel what is right.
I make a cardboard mock up for myself of the size of a cubebit.js and maybe makes some schematic/pcb with the only intension being to see if things look right.
Of course, the pdb should give value in comparison to the nRF52-dk which has a known price with pros/cons in order to achieve "bottom dollar" for espruino
PCB should give value in comparison to the nRF52-dk
It would give the value of having Espruino pre-installed, and by actually helping to support Espruino development.
I can't compete with boards like nRF52-DK - I believe these are sold basically at cost in order to get more nRF52s into products.
yes: the need is support Espruino.
I looked at the nRF52-DK: it is not allowed to really operate as a "consumer product".
So there is a markup for the value of being a "consumer product" for an Espruino product.
I have spent all day holding the top green board (mostly cos I have problems) but the size of a shield seems right
It is a bit bigger than the ORIGINAL ESPRUINO 54mm x 41mm (2.2in x 1.6in - half business card)
If I am write in Gordon wants the PCB as the same size as a shield with arduino pinouts
And if the area that is not with connector pins is a prototype area, I can allways break it of to make it square for a cubebit.js :)
If I remember correctly, the arduino culture was that a device should fit in a cigarette box.
As I like to feel something in my hand, I have of course put some shield in a cigarette box and looked and felt if the "size is right".
If Gordon goes for the shield size: ie a shield pcb with connectors for arduino, I am in.
Next step is leave as much copper on the board for anything that the community would like to SMD/prototype to so that the copper that would otherwise go down the drain to the environment is put to good usage
As I have been digging up the garden due to my bad drains, "drains" are in my mind at the moment.
I also was looking at a Arduino Shield prototype board and thinking of making female on the bottom of my puck.js that could connect with males on the prototype board and thinking a PCB with connectors to a puck.js and with smd to the, as yet not released, raytac nRF52840 would be fun allowing for future options while still having a nRF52832 product.
ie back to the nRF52 comparison where there are certain differences between the products that should probably be put in a comparison chart.
In releation to the above arduino shield drawing.
Power, programmers far left, with clear area and pin holes so if one has another form factor, just cut it out and reuse where one needs power and programmers.
Far right, smd patterns for both processor and pin holes to allow for cutting out to whatever one needs.
In the middle, smd patterns for accel and mag and pin holes for the puck.js
Almost forgot, anything else there is room for with pin holes for something else there may be room/usage for
In the meantime, I just use a prototype board with RTC and SD Card
Titbits from raytac about how to program from nRF52-dk
A nice guy is doing a new batch of this shield "soon"
I hope that he will leave a puck.js footprint on the board to save copper going down the drain
Hopefully the 8 main 0.1" holes on Puck.js are on a 0.1" grid, so should fit on any 0.1" prototyping board?
I have a soldering iron ready, just have to find a good female connector for the bottom of the puck.js
I only have one puck.js so I have to be carefull.
Also a minor issue in obtaining contact to "interesting io" on the top of the puck.js (solder blobs which require thought)
Well, if I make a mistake, I know where to order for more ...
Did anyone say shaky hands?
As Gordon is working hard on the puck.js and Espruino support and I am trying to show the puck.js as an addition on Arduino boards I thought I would write the following as an analysis.
nRF52840 - NEW! Multi-protocol Bluetooth 5, Bluetooth low energy, ANT, 802.15.4, 2.4GHz RF SoC Contra nRF52832 Multi-protocol Bluetooth low energy/ANT/2.4GHz SoC
New is nice, but the efforts to create support for a new device, nRF52840 would probably not be a "good idea"
Concentration on nRF52832 is the best for puck.js and possible future PCB break out.
I drop thoughts of nRF52840 : next year maybe.
This year? I hope for enough intressted people for the nRF52832 PCB but the puck.js is filling my needs at the moment, PCB would also be real great if possible, but not at the cost of time for puck.js
Does the forum support movement of a thread to another forum, so this very interesting thread is not under puck.js but "something else" as it isn't really puck.js?
Addition to Gordons comment: possible small difference in power consumption.
Some analysis of the difference at some time but just not now :)
It's worth noting that nRF52832 does do Bluetooth 5 - as far as I can tell the only radio difference is a slightly higher power output, which allows longer range. Otherwise it's the same.
Yes, I have to move threads - I've just done this.
And strange how few pins are exposed:
And who is using espruino:
Edit: from Gordons comment below
http://forum.espruino.com/comments/13453118/ : keeping Espruino going with backers and keep things simple by just buying Espruino: which is good enough for me.
"Nobody needs to fight for what is right"
You might see some other threads on redbear/espruino if you search.
I'd love to have an "Original" espruino in the nano2 form :-).
Or even smaller: but there might be problems with some approval EMC or what not
Oh wow, that is small.
The idea of using a small/basic ble esprunino in a plug an play manner would be great, BYOB - Bring Your Own Bluetooth if you like.
Also 3.7V compatible :-)
Yes, I'd seen those boards - they look lovely! Sadly a bit on the pricey side when I looked (and nRF51-only at the moment).
I was considering a slightly larger DIL footprint that'd allow me to stick the module in the middle without having to do anything too costly (or fragile :).
3.7v is actually a really tricky one...
I could add a voltage regulator, but they draw some power just by being there - and also make it difficult to run it off a flat 3v. Another option might be to just stick a diode on the board to drop the voltage enough.
But even if I do that, the IO isn't 5v tolerant at all - which could really cause some problems.
Perhaps this: https://www.u-blox.com/de/product/b200-nina-b1
Edit: as a starting Point.
Having made a cardboard cube to find out if the size feels right, one could also consider what to do with the boring base, add wheels and some light sensor(s) and http://ozobot.com/ as an add-on "when/if the time is right" either by oneself or by peripheral suppliers to a proven espruino sw base.
Make sure the cube has some means to be pinned into for some form of bigger bot so the kids have something a bit bigger at some time
I mean, this is just plain ugly
Even after a bit of presentation work, there are still "big problems" and in the end, one says to one's self "there must be a better way"
What would make one version of a cube case, look at the espruino logo: the handle of the coffe cup could also have a function, what about a set of rgb's nicely down it?
Or just a Big Red Button on the top of the handle and use the hollow of the cup for cool things like a couple of buttons and a 5x5 led display add on or what about a 5x5x5 RGB Led stack.
The hollow of the cup could also just be a place to put one's puck.js so one can find it again while making integrations between the cubebit.js and the puck.js
Put the cubebit.js on another add on, an induction charger where the internal coil supplies power to the rechargeable batteries as another add-on
Peripheral business with suppliers of well thought out add-ons who can also take some of the pain of assembly and test of possible DIY espurino kits