wireless shield usable for digispark (and other arduinos)

[gogol]

Just thinking about a project, where I would like to have wifi connection on the digispark.
When I checked current possibilities, I thought, that it might be a good idea, to create a wireless shield just with the module used on the digix.

Available shields are almost as expensive as the whole digix, so a shield with only the wireless module should be possible for 50 to 60% of the digix price, I guess.

Just as a thought...

regards

  gogol

[MichaelMeissner]

Does the digispark even have enough memory to run the typical code you need with wifi?  I recall people saying that on Arduino Uno's, adding the wifi shield used up most of their SRAM (Uno has 2k of ram, Digispark has 0.5k).

[dougal]

I was wondering the same thing. I think with the limitations of Digispark, the WiFi addon would have to be able to do a lot of the heavy lifting on its own, and provide a really simple interface and API to move data back and forth. I'd love for someone to prove me wrong, though.

[MichaelMeissner]

And if the wifi does all of the heavy lifting, presumably the processor doing all of the protocol work has much more memory and speed than the spark does, and it might be better to do any embedded programming on that chip.

[digistump]

A wifi shield (with the same module as the DigiX) will be offered for the Digispark Pro as it will have 16kb which will be more than enough to hold the library and do something with it - it also will have a hardware UART and using it will still leave pins for attaching any original Digispark shields.

[PeterF]

Sorry if I missed it somewhere, but is there an approximate timeline on the Digispark Pro? Next six months? Next year? Not after a guaranteed release date, just wondering when we'll here some more specifics about size, layout, capabilities, etc.

Thanks,
Pete

[digistump]

We're shooting for a Kickstarter for the Digispark Pro shortly after the DigiXs are all shipped - so full specs, demos, photos, etc in late October/Early November. Possibly sooner, possibly more towards year end (depends a lot on the DigiX production and the prototypes for the Pros) - but not 6 months or a year or anything like that. Since we'll be making them in house we'll be targeting starting to ship them (albeit at a limited number per day) shortly after the Kickstarter ends, so hopefully before year end as well.

[gogol]

That sounds terrific!
excellent!

regards

  gogol

[dougal]

Any chance you can give us past supporters an early heads-up right before the Kickstarter goes live? I want to have a good chance at being an early bird. 

[digistump]

Will do for sure!

[gogol]

And if the wifi does all of the heavy lifting, presumably the processor doing all of the protocol work has much more memory and speed than the spark does, and it might be better to do any embedded programming on that chip.

Hello Michael,

that is the problem of the most WiFi solutions: either its only a NIC (like all the very cheap tiny USB adapters), in that case you need a powerful CPU to handle the protocol stack. The other solutions are embedded devices, which handle the stack, do all (or most) configuration work (via web-browsers), but they have no GPIO pins (otherwise you would be totally right).

There is one interesting thread going on about devices called "HAME A1" or "HAME A2", which are available as mobile ethernet, wifi and G3/G4 mobile internet-router. They are available about $20-$35 and they seem to have some GPIO...  https://forum.openwrt.org/viewtopic.php?id=37002&p=4  In the "A2" version, the enclosed 5200 Ah LiPo would be also the perfect base for mobile digispark solutions!

If you need it smaller, than the Transcend WIFI SD-card may be an option: https://forum.openwrt.org/viewtopic.php?id=45820, Here you will get in addition to an WIFI enabled linux system 16 to 32GB of memory for the same money, adafruit calls for the CC3000 wifi shield!

Up to today I found no cheap wifi solution, which has 5 to 10 GPIO pins.  However the two examples above are showing, that there are devices on the market, which are cheap, small and low power consuming.   

Arduino started with the yun a standard sized wireless arduino board, also following the shared cpu concept.  There is for sure an market for much smaller devices with similar possibilities. Wifi has several advantages over NFC (which has for sure its own use cases).

So I think, that an digispark pro with wifi shield, which will be smaller and cheaper than the yun, would be the perfect solution for many small projects, where you wish to control some devices over WLAN.

Regards

  gogol

[digistump]

Digispark Pro + Wifi target price is $30 for both. We can never beat things like the HAME since we just can't buy 1million+ units at a time, but $30 makes it the cheapest arduino compatible complete solution as far as I know and the wifi module does its own stack so the Pro is free to do other things.

[MichaelMeissner]

Well there is always the Rasberry Pi model A ($25) + usb wifi dongle ($12), which is roughly the cost of the CC3000.


I tend to think that SRAM size on the Spark/Gemma/Trinket/etc. will limit what you can do with wifi, even if most of the heavy lifting is moved off chip.  But if it solves your problem, go for it.

[gogol]

Digispark Pro + Wifi target price is $30 for both. We can never beat things like the HAME since we just can't buy 1million+ units at a time, but $30 makes it the cheapest arduino compatible complete solution as far as I know and the wifi module does its own stack so the Pro is free to do other things.

The HAME and all the clones are (at least now) not comparable, as they are not sold as a programmable platform. Its all based on reverse engineering, and it looks like, the PCB and the components are just changing from revision to revision. 

So an arduino-compatible small device has an market, especially in cases, where size matters as well.

Regards

  Gogol