No love for the PIC10 series? Perhaps too many PIC already in the review would make it look like a MicroCHIP(TM) press release. I do see some justification in that the 10F series is more of the twenty five cents class not the $1 class. Those things are strangely charming, I had a lot of fun with the now-discontinued 10F2xx series. Adjusted for inflation the 10F3xx are now cheaper than the 10F2xx were, which is cool. Its strange to drink a can of diet coke knowing it costs 3 times as much as the controller you're troubleshooting.<p>Another novelty of small/cheap controllers is when I was playing with the defunct 10F2xx family I didn't own a meter sensitive enough to measure the current draw at 32 khz clock. I do now, but it was an expensive meter! In the old days cheap EE would wire up a capacitor and run the ckt off the cap and time how long the voltage drop took and use that to create an imaginary resistance and use the measured V and theoretical R to calculate the theoretical current. Of course my cheap meter drew more current than the microcontroller which leads to weird games like sample voltage for a second. Also leakage currents in most caps is larger than the uC current draw oh so much fun fun fun. One trick is to drive a square wave generator thru a diode to a low leakage ceramic disk cap, assuming your reset circuitry is faster than the square wave and you trust the DC response of your scope. Another thing I tried was creating a transfer function where a 100 watt lightbulb and a tiny solar panel generate say, 1 mA, so a 99.9% optical attenuator in front of the cell means 1 uA, right? Of course sleep current is a tiny fraction of that, so you stack a SECOND optical attenuator on the solar cell, say merely 99% blocking, and that should be 10 nA which is about half sleep current, right? You can also play games when your CPU costs 30 cents like hooking up 100 in parallel for $30 to measure 100x average current. Low current is fun if you're ever bored.