Archive of the former Yahoo!Groups mailing list: Homebrew PCBs

previous by date index next by date
previous in topic topic list next in topic

Subject: Re: [Homebrew_PCBs] Homebrew Tool Changer

From: mpdickens <md30022@...>
Date: 2003-12-03

--- bsjoelund <tecnoconsult@...> wrote:
> Hi,
>
> I�m looking for small/miniature tool changer for PCB
> drilling.
> Any links - pictures - ideas ??

Hello Bengt,

This is robotics and robotics is as much an art as a
science. Further, robotics can provide humbling
experiences as well as be exceptionally rewarding
(Both experiences the same time...). For drill bits,
excluding the ability to position the head of the
drill (The head of the drill is called a chuck in the
America. I don't know what they call it elsewhere...)
at the correct latitude/logitude/altitude, you going
to need quantity-4 servo motors and quanity-1
sensor.(I don't think you can change a drill bit with
less than four motors and one sensor). Three of the
motors ∗must∗ have a high tork and all motors must
step in fine steps.

Some drills have keyless chucks and some have keyed
chucks. This example assumes a keyed chuck:

Motor A: Pushes the key up to and into the chuck.
Motor B: Adjusts the key to fit the gears of the chuck
Motor C: Rotates key to tighten/loosen the chuck
Motor D: Rotates what is known as a "lazy susan"

A lazy susan is a round table that in this case has
holes in it that the drill bits sit in (Standing up)
The drill bits are changed over the lazy susan.
Further, the lazy susan rotates.

It would work like this:

The drill head has no drill bit in it. The drill head
now moves over to the lazy susan. The head is lowered
down to the lazy susan directly over the desired drill
bill. Motor A pushes the key up to and into the chuch
while Motor B is ∗slowly∗ rotating the key. Once the
gears on the key catch with the gears on the chuck
(This is a predetermined distance plus or minus a
small distance. The gear catch is also monitored by a
senor that measures tork. Further, They make chips
that measure tork), Motor A stops and Motor B
continues to turn the key to tighten the chuck (This
is a predetermined number of rotations for each drill
bit and size. Also, tightening is monitored by
another tork sensor). The sensors are "fail safes" for
the both distance and tork which is predetermined,
but ∗estimated∗. Once the prodecure is complete your
good to go. Reverse the processes to take the bit out
of the drill head/chuck.

What you want to do requires very fine motor steps
combined with software that does not glitch during the
process. This process definately is an embedded system
thing and should be automomous: In other words, the
drill files should be handed to the embedded system
and the computer that hands the embedded system the
drill file should stay out of the process (Which
greatly simplifies the software. Further, this makes
debugging MUCH MUCH MUCH easier...)

If I were going to do this, I would look at:

A.) Existing products and how they work mechanically.
B.) Tape libraries for extra idea's and inspiration.

Figure everything out that you can before purchasing
anything, spinning the first wrench or writing any
code. Also, draw schematics of how it's going to be
buildt and how it will work. Flow chart the software.
Keep damn good notes while
building/programing/testing.
Keep rescipts from your purchases in case you buy the
wrong part and want to return it. Do business with
small to medium size companies while prototyping. Cut
a deal with them that you can return anything you
don't need if you have not used it and it's not
special order (Anybody who will not cut this type of
deal, don't do business with them. Don't fall for the
15% restock BS fee. There are too many companies out
there that will work with you the right way).

Good luck and best regards


Marvin Dickens
Alpharetta, Georgia USA

=====
Registered Linux User No. 80253
If you use linux, get counted at:
http://www.linuxcounter.org

__________________________________
Do you Yahoo!?
Free Pop-Up Blocker - Get it now
http://companion.yahoo.com/