We've been doing a lot of hacking around with BLE lately, so I thought I'd start publishing some of it. Sorry in advance about the scarcity of details. 
 
Recently we took apart a Belkin Conserve Insight energy monitor (model # F7C005) and determined that the unit used a simple SPI interface to an Analog Device's ADE7753 Metering IC. After sniffing the SPI comms of the original unit, it was pretty easy to tweak some BLE code we had lying around, and get it working with a Nordic EVK. The only real trick was interfacing the power ... the small (ugly) board in the upper right is a small 5V to 3V LDO, which is used to power the EVK. NOTE that we can not (nor should you try) to interface this set up to a computer for debugging. The 5V power supply in the main housing is non-isolated! 
 
The comparative wattage results from our setup to another F7C005 and a Kill-A-Watt were within 1 - 2W over a 1KW range. 
 
 
 
Image of the cable termination (SPI+Power) on the PCB. 
 
 
Pinout. 
 
 
The unit does a little level shifting (the ADE7753 is a 5V part, and the uC in the display unit is 3V). 
 
 

Posted by: Lab Coat

We finally fixed a nagging power consumption issue related to the TWI (nee I2C) on early rev Si of the nRF51822. You can see the PAN for all of the gory details. Suffice it to say, the TWI wasn't shutting down correctly, leaving us with an unexpected +1mA current draw. We poured over some code that Nordic uses in the nRF Smart Remote 2, and adapted it for our needs. Problem solved. 
 
We'll post some "TWI to XXX sensor" examples shortly (IR Temp, Barometric pressure, Humidity, Ambient Temp, etc). 

Posted by: Lab Coat

So we've noticed an issue with the iOS 6.0 simulator … when running some iOS test code, and while trying to connect - disconnect from the target hardware (nRF51822 EVK), the iOS 6.0 simulator will not release the BLE connection for about 60 seconds. 
 
Annoying, but not fatal. 
 
Problem is, I was trying to debug some new nRF51822 firmware yesterday and in the end couldn't tell if it was the BLE firmware or the simulator. Long story short, when running the same iOS code in the iOS 5.1 simulator, we get the expected / correct operation (that being when the iOS code disconnects, the simulator immediately disconnects from the hardware).  
 
My setup is 2011 13.3" 1.8GHz i5 MBA running X.7.5 using Xcode 4.5 with iOS 5.1 and iOS 6.0 simulators installed. 

Posted by: Lab Coat

Here is one of the nRF51822 EVK's in a demo setup … the goal of this hardware stack is to run the EVK from a single "AA" style battery. The lower of the 2 boards is a booster (actually 2) based on the Microchip MCP1640. It takes 1.5VDC of the single battery and boosts it up to a little over 2VDC. Current draw ( … on this board … ) when the nRF51822 is "off" is around 20uA; when on it's an average around 2mA. Really like these parts! 
 
We used this setup to demo a thermistor temperature probe and a high pressure stainless steel pressure sensor for an HVAC application. 
 
 

Posted by: Lab Coat

A big THANK YOU! to M2STech and NIC for getting me these 0402 engineering sample kits. 
 
Lots of resistors! (100 x 150pc) 
 
 
 
Lots of capacitors! (40 x 50pcs) 
 
 

Posted by: Lab Coat