24 April 2016

2.77 Seek and Geek #8: Car Mirror Adjustment Mechanism

So my passenger-side mirror got busted on my car. Upon replacing it ( for only $25.00 on Amazon Prime and 10 minutes to repair), I decided to take it apart for my Seek and Geek! Making the most out of a busted mirror!

Here it is in its temporary-fix Gaff-taped glory. 

I used a chunk of a sheet or mirrored acrylic as a temporary fix so I could legally still drive (and practically still see behind me!). 
Putting the "Why?!" back in DIY (DIWhy?!?!?!)

The whole mirror assembly comes off with three self-tapping screws for plastic. There is a connector with three conductors, presumably for Motor1/Motor2-, Motor1+ and Motor2+. Using these three, you can spin both motors forward and reverse, but only one at a time. Foam gaskets are glued to the inside to dampen noise and catch any water/dust/roadstuff that may make its way in there. 

The glass is (was) connected to this inner face. There is a single snap-fit ball joint in the middle, about which the mirror pitches and rolls. Two additional smaller snap-fit ball joints, one on the bottom, and another on the right, are connected internally to the outputs of the motor assembly.

A better view of the middle ball joint, and the two ball-joints that connect to the motor outputs. The balls are at the ends of linear actuators. The motor assembly pushes or pulls the bottom ball of the mirror plate to aim it up or down, and pushed/pulls on the right ball to aim it left-right. 

A closer view of the motor assembly, with the big middle ball joint which acts as a pivot point for the mirror plate, and the two linear actuators, each with a ball end. Grease is applied liberally as lubricant. 

Upon taking the motor apart, you can see two small motors with identical transmissions that do a few things. 

First, a worm drive right out of the motor enables it to apply plenty of torque onto a long idler gear. This idler gear then attaches to the actuator output gear, which is threaded onto a screw. As the long idler gear spins, the big output gear also spins, and the ball output moves up or down along the screw. The idler gear is long enough to continually engage throughout the length of travel. 

(Idler gear removed). Here is the output threaded down to its lowest point. 

And here is the output stage threaded up to its highest point, about 0.75" of travel. 

Here is the plastic screw on which the output shaft spun. 

Taking a look underneath the Five Elastically Averaged Flexures used as the leadscrew "nut"! Smart! There are no limit switches or feedback control on the mechanism, so it isn't a servo, just an open loop motor. With these flexures, when the shaft output (the mirror angle) reaches its limit, the flexures move out of the way and it can click down or up along the plastic screw without causing damage to any device. BRILLIANT!

19 April 2016

2.77 PUPS 5678 Pupdate (April 19)

ALL MCM PARTS HAVE BEEN ORDERED! Should Arrive by Wednesday, and I can build the structure around that and test them in time to have the whole machine working by next week. I need to pick up 2 more servos on Wednesday. 

And now, for the work that got me to this point:



Part 1 was uploaded on April 18, this is Part 2.


18 April 2016

2.77 April 18 PUPdate 5,6,7,8












Gonna place an order VERY soon! I also have plenty of 0.25" MDF stocked, and will be using that lasercut, along with wooden half-balls, for the robot. Benefit to MDF is the ease of lasercutting and its low density. The biggest risk in using MDF is its low stiffness, as it only has a E of 4GPa, but it can probably work with some additional Aluminum parts. Making sure the clamping mechanism doesn't yield the machine will be one of the most important parts of the design moving forward.

Check out part 2 of this PUPS: http://yameb.blogspot.com/2016/04/pups-5678-pupdate-april-19.html