Re: [CAD_CAM_EDM_DRO] Re: robotic workcell design

HI Rick;

A few more comments>

skykotech wrote:

>I split this discussion off from the laser safety one because that
>one is certainly off topic. I think this universal gantry is
>perfectly on topic with this group.
>
>I also had the idea of using aluminum extrusion and getting the parts
>pre-cut from a place like 80/20. I have found it is very very easy
>to cut aluminum on a regular wood cutting miter saw if you have a
>good Freud carbide tipped saw blade. Mine has like 120 teeth and is
>specifically meant for cutting non-ferous metals. I picked up the
>porter cable miter saw at Costco for about $280 and the 12" Freud
>blade at home depot for around $50. It cuts a 2 inch x 2 inch block
>of aluminum like butter. Enco and other places sell a type of solid
>lubricant (green colored) that looks like a bar of soap. I wipe this
>on the sides of the blade after a few cuts.
>

Cutting aluminum extrusions is not that bad. Things can be improved by
buying a blade made for that activitiy, most better tool supply shops
have the blades. The problem with extrusions is that they are
expensive so you eat up some of your budget. There may also be an
issue with getting a solid structure. It is hoever a good approach
for the home shop. If you are willing to shop around you will find a
huge variety of extrusions on the market so a good portion of the
machine could be fabricated that way.

>
>But I digress, since 80/20 and others will pre-cut your extrusions to
>exacting specifications for a nominal fee. Techno-Isel makes a nice
>T-slot table surface, but I am not sure how much that is. Perhaps
>the table surface could just be several pieces of 80/20 bolted
>together, or a ground aluminum plate available at most metal
>suppliers. A total ballpark on the extrusion prices...$500?
>

It looks like you ar well on your way to finding the extrusion
vendors. For the base of the table I'd go with the plate and use it
as a platoform to bolt other things to. I'd also make it rather thick
to prevent deflection and to provide a solid purchase for threaded
holes. Of course how thick depends on a number of things such as the
machine having a stationary gantry versus a moving gantry.

One does have to be careful about which Aluminum plate is chosen for the
application as some can be dead soft.

As far as the extrusions go and the price paid, I think you will have to
work with the vendors. As stated above the stuff is not cheap and if
you have then cut and fit a kit it will be even more expensive.
Expensive to the point where having some of the parts fabricated out of
steel might be cheaper. Either way the vendor would have to be
willing AND able to handle the spike in sales such an article would
generate.

>
>If the linear rails/carriages can be had for another $500, the
>ballscrews and ballnuts for another $300, and leaving $700 for servo
>motors, drives, and power supply (would really have to scrimp
>here...gecko g320 might be out of the question).
>

Well on the initial machine you could always design for steppers letting
people know that a future iteration will be servo based. To be honest
I don't really see the point in building any thing new with a step and
direction servo motor drive. If you want the advantage of servo motor
drives that implies having position feed back to the CPU running the
control algorithms. I'd stilll want to shoot for lower prices if at
all possible. Many people put up stiff resistance to anything >999 in
the home shop. There is usually good reason for that as even a small
shop ends up needing many machines. It is also in my estimation why
so many people on the net fabricate their own hardware as much as
possible. Sure linear bearings and rails offer great technology, but
box ways are very cheap in comparison even if you buy ground stock.
Don't forget that the drives / power supply & etc. need a box to go
into. The incidentals do add up too.

Please don't miss understand, I'm not saying to disregard linear rails
and such. Just that one needs to determine from the point of economics
where to put those devices. If the price increases to much you will
soom find that you are competeing with machines for sale on the net and
else where. I'm not sure you really want to do that.

Check out Pico-Systems on the net, he has an almost ideal I/O board for
building a machine in this manner. It is called the Universal
Stepper Controller and there is a version for servos also. Now I
have not used this product yet, but it certainly has the right feature
set. So this isn't an endorsement, just a pointer to something I've
found to be very interesting.

>
>I think this would make an excellent article, and should make the
>editors at Servo drool.
>

Now if you could only get them to underwrite the cost of the machine ;)

Servo is an interesting magazine and may even be worth a subscription in
the future. At the moment though their editorial slant seems to be
confused. If they don't see the light I'm sure there are other
magazines that might pickup on such an article. Of course the bigger
the magazines distibution the better for you.

Dave

>
>Rick
>
>
>
>