Does anyone have any idea how something like this pin clock works? That (and variants thereof) seem like they'd make great Arduino projects.
Mikal
Hi Mikal,
I guess are 14 small solenoids for each digit (two solenoids per segment). Each pair of solenoids drives a small rigid bar that is around 2pins wide and around 8 pins long. Activating the solenoids moves the segment bar which pushes the pins to indicate the segment.
That's my guess anyway. 
One might be able to accomplish it with 7 solenoids per digit if they could be made to be mechanically latched and unlatched. That's still a heck of a lot of solenoids! It would be very interesting to be able to reverse engineer how they do this. But once learned I would try and apply the method to something for interesting then just a clock.
It could be done with one motor. Imagine that each segment is controlled by a simple cam arrangement. Rotating the cam brings the segments in and out to form the 0 to 9 sequence, 0 to 5 sequence, or null to 2 sequence. Each cam section is geared appropriately for each digit. Spinning the motor will generate the desired 12 hour sequence on the pins.
Sigh. I was naively hoping for individual pin control, but you all make very convincing arguments why that almost certainly is not the case. Doesn't it bug you that about 75% of the pins are nonfunctional?
@mem: I don't know much about solenoids, but I thought that it was generally inadvisable to keep a solenoid in the "deployed" state for long periods. Perhaps the solenoid just pushes the pins out and then immediately retracts? How do you think the pins get "pulled back"?
@retrolefty: I agree that if I had this technology, I certainly wouldn't make a clock.
Who's going to spend the $80 to reverse engineer? 
Thanks, all!
Mikal
How do you think the pins get "pulled back"?
A latching mechanism similar to push on/push off switch would mean power need only be applied briefly when the time changed. only segments that changed would need to be toggled.
That said, I think macegr may have nailed it.
As far the the pin array goes, I have seen those used as standalone assembly used to duplicate the shape of an item pressed into the array of pins. There are all spring loaded so the pins return to the their original position when no force is applied to them. Yes can you imagine what you could do if you could control each pin, you would have a mechanical graphical display panel, maybe not with the fastest update time know to man, but still very very cool
Yea, someone should jump on getting one of these and take it apart. It would be like being ten years old again with a new screwdriver in hand ;D
Lefty
"I don't know much about solenoids, but I thought that it was generally inadvisable to keep a solenoid in the "deployed" state for long periods. "
It depends on the solenoid design. They can be designed and made for continous operation or not. Another method, even if not designed for continous operation, is to supply full rated current only for inital pull-in and then cut back on the current once there, as holding force is much higher then pulling force.
Lefty
I received one of these clocks as a gift a while back. First off, there are no pins in this clock; it's designed to look like there are though. I'll try to describe it.
Behind the clear cover there is a main background with bumps that look like pin heads. The board and pins are the same color grey plastic, since it is just one molded piece of plastic.
Also on the main background there are areas with holes for the "pins" that do move, laid out in a pattern like a 7-segment LED display. Each segment moves in and out (matching those fixed bumps or raised) of the holes to display that segment's "pins."
I say "pins" because it's really just another piece of that grey plastic but shaped like the segment and with taller bumps that appear to be pins.
I haven't had it running in quite a while since it's noisy enough to bug my wife, so this is from memory. On the minute, each segment gets reset to the correct position in sequence (with much clatter, leading to the shutting off mentioned above). It does it this way because the segments aren't ever held in place, they only get toggled in or out. If you lay the clock on its back, the raised segments will fall back down immediately so the display is blank. You can lay it on its face and the opposite will happen.
I have never taken it apart to see what kind of solenoid or motor arrangement is in there, but I believe each digit is generated by pulling/pushing on the segments in order, not all at once.