How good is BLE for keyboards? Anyone trying this yet?

[mrflow3r]

Ok, I get the standard/classic BT modules consume too much power to be used as keyboard controllers.
I think I have read 12 hours is usually what we get with the combination of a Teensy + Bluefruit EZ Key.

Well, it's 2015 and I see a number of BLE modules in the market. One being Rigado's BMD-200 [1].
On their specification page [2], it says it consumes 5.5mA @ -20dBm for TX, and 2.4mA @ 16Mhz for the CPU.
Off the top of my head, Bluefruit EZ Key alone uses 25+mA. So these numbers look pretty good.

I ordered a BMD-200, and made a breakout board[3], so I will try to find out the numbers in a few days (hopefully, if I can get my head wrapped around BLE).
That said, is this even the right tech to use? I heard proprietary radio protocol gets *months* of use time (the ones we find in Logitech kbds...).
So perhaps that's what we should aim for, rather than these general purpose BLE stuff?

[1] https://www.rigado.com/product/bmd-200/
[2] https://www.rigado.com/product/bmd-200/#55de7bc5f1311-1
[3] https://oshpark.com/shared_projects/f1aGKcII

[twiddle]

I've heard some good things about the Nordic nRF51822 BLE solution. I'm still doing a fair bit of work on my actual firmware implementation before I get to looking at BLE, but I've got a devkit ready for when I'm at that point.
BLE is a *very* different protocol compared to Classic BT in terms of how it consumes power, so I think you'll find the results are quite positive.
Here's a very trivial, untested power consumption estimation for the NRf chipset:


As you can see, even with very high connection frequency, we still get about 9 days battery life from a 220mAh battery. If we combine that with USB charging functionality, which isn't *too* difficult to implement from my preliminary research, that seems quite acceptable to me.

The primary disadvantage of the proprietary protocols is a lack of universal support. I use tablets and iPads often enough that I think being compatible with those is a definite plus, and you aren't likely to find a Shockburst or other proprietary protocol that works with that sort of device.

[mrflow3r]

I've heard some good things about the Nordic nRF51822 BLE solution. I'm still doing a fair bit of work on my actual firmware implementation before I get to looking at BLE, but I've got a devkit ready for when I'm at that point.
BLE is a *very* different protocol compared to Classic BT in terms of how it consumes power, so I think you'll find the results are quite positive.
Here's a very trivial, untested power consumption estimation for the NRf chipset:

Show Image

As you can see, even with very high connection frequency, we still get about 9 days battery life from a 220mAh battery. If we combine that with USB charging functionality, which isn't *too* difficult to implement from my preliminary research, that seems quite acceptable to me.

The primary disadvantage of the proprietary protocols is a lack of universal support. I use tablets and iPads often enough that I think being compatible with those is a definite plus, and you aren't likely to find a Shockburst or other proprietary protocol that works with that sort of device.

Thanks twiddle for ur response! It looks like BLE is a good choice for the time being. *9 DAYS* of juice on paltry 220mA battery! I may still look for Rf development, but BLE could just do the job looking at your  numbers. Btw what's the estimation tool you are using?

[twiddle]

I was just using the official software from Nordic themselves, theres an estimation function included.