I got around to taking the clock apart and measuring power draw. The results were encouraging.
Using a 4.55V alkaline battery pack (nimh will be lower voltage), the clock uses a pretty impressively low 20 uA (microamps) with the back light off.
With the back light on, power consumption increase 11X to about 235uA! This isn't real surprising. The majority of the power consumption of even laptops is the back lighting for the screen.
However, even 235 microamps is still a very tiny draw. The batteries I'm using are rated for 2200 milliamp hours. That is the equiavalent of 2,200,000 microamp hours. So, even if the back light was on 24/7, the batteries should theoretically last 390 days! But, that is not the true case, in reality the current draw will be even lower, lets say 128 microamps (average of 20uA and 235uA). This almost doubles the battery life to 716 days. Still, this isn't realistic because nimh batteries are known to 'self discharge' on their own.
So, the solar panel must compensate for the draw of the clock plus the self discharging cells. I'll note you can buy low self discharge cells (aka precharged). Mine are not precharged cells, so I'll have to deal with this.
It was also
mentioned on the blog post that my solar panel voltage may not be high enough to sufficiently charge the batteries. The commenter says you need at least 2 volts above the battery pack voltage to get a charging to happen. Testing will have to be done to ensure that the solar panel is in fact charging the batteries.