Tony Smith wrote:
>>> Does this need to be a random-access unit, or are you just testing 
>>> durability of the keypad?
>>>       
>> Bob, the 'fingers' have to simulate a person using the device
>>
>> A PIC will take serial data from a PC and activate the keys
>>
>> After a chat with the other principals last night, they 
>> suggested I keep it simple. Which, given my work load, I'm 
>> inclined to go along with
>>
>> What we're doing will be demonstrated to interested parties 
>> and should be kept as compact as possible. So that ruled out 
>> air and XYZ actuators and all the other great contributions
>>
>> I've also been told that there are some (slightly important) 
>> timing specs involved and that pretty much means solenoids
>>     
>
>
> Simulating a 'real' keypress is harder than it looks.  I remember seeing a
> documentary once where they were building a robot piano player (as you do).
> Cams, air, solenoids etc were all tried, they ended up with a combination of
> air & solenoids.
>
> Someone earlier had a good idea, use car door lock actuators.  These are
> quite cheap (~$25 a set of 4), although I got 12 packs off eBay once for
> $50.  $1 each!
>
> Suspend them above the keypad in a 4x4 grid, add dowels to the end to reach
> the keypad, and a piece of perspex with guide holes drilled in it above the
> keypad.  The 'fingers' will be on a angle depending on how you arrange the
> actuators, and how far above.  (Using levers overcomes this.)
>
> This is simpler than motors, cams & levers, but you can't vary the force of
> each keypress, and timing could be a problem.
>
> Failing that, small children work cheap, but have poor long-term
> reliability.
>
> Tony
>
>   
To get a "force" just put a spring on the end of your solenoid something 
like a ball point pen spring would do. Compress the spring by X amount = 
Y force. Should be close enough for government work.
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist