"Ignoramus11878" wrote in message
...

On 2013-10-17, RogerN wrote:
"jon_banquer" wrote in message
...

On Wednesday, October 16, 2013 5:26:48 PM UTC-7, RogerN wrote:
I'm considering building an AK-47. I guess the way to go is to buy a
parts

RogerN

How many more guns do you need? Why not build something that will really
challenge and build your CNC machining skills? I'm talking about parts
that
have curvy/swoopy surfaces. How about building a model turbine or
something like that? I'll help you create toolpaths for your parts with
curvy surfaces. Why not make the jump from prismatic parts to curvy parts
and leave most of the machinist wannabe's in this group in the dust.
Idiots
like Wieber, Larry Jackass, Tom Gardner,Danny boy, etc. couldn't create
CAM program for curvy/swoopy parts if their lives depended on it.


The AK-47 build is to get a popular semi-auto, with good parts
availability,
in 7.62 X 39, the ammo seems cheaper than .223 right now. larger &
heavier
bullets for less money, that's my interest in an AK-47. The AK build was
something to be done by drilling a few holes on my manual mill and
perhaps
hammering some rivets.

But on the CNC parts, I share your passion. I'm very interested in 3D, 3
axis tool paths, I'm interested in making molds and patterns.

So far I'm still without a convenient shop, it's about 5 miles out of the
way. I contacted the bank about getting money for a building, they
didn't
sound too interested, they were supposed to call me back but didn't.
After
buying my new car, my credit rating dropped to around 765, I doubt that's
the problem. I plan to ask another bank next time I take some time off.

Back on topic, what you are talking about is the reason I'm interested in
CNC and have spent money on a CNC mill and lathe. I know my machines are
old technology but I can update them to EMC2 and have massive amounts of
memory, billions of lines of G-code, no problem! :-)

RogerN



Roger, it is a fun project, but I am wondering, you presented so many
ideas, which one was the one you actually implemented?

Currently I have a PC with Linux loaded on the Solid State Hard Drive and
have it working with my Mesa board, I have a couple of encoders I can turn
by hand, it counts and the analog outputs respond to the error. To do
includes mounting the breakout boards and connector, wiring the connector to
the breakout board, plugging it all in and then adjusting the PID parameters
for best response. My board has 6 axis of encoder and analog outputs so I
can do my X,Y, & Z axis, plus the spindle, plus add a 4th axis later, and
still have a spare. So I could have it running in a few hours but a few
more hours to get the spindle control and encoded for tapping.

If I get a building here I should be able to get it going in a little spare
time. My lathe conversion went well, the drives responded to the analog
cards like I hoped, it was easy to tune since the drives are already
adjusted to the tachometers on the motors. Velocity feed forward gives the
basic velocity command and the PID gains adjust to minimize following error,
the lathe adjusts to 1 encoder count, 1/12,700th of an inch, pretty quick
and smooth with maybe 15 minutes spend on adjusting the gain. Getting good
spindle response using encoder feedback took a bit longer and still leaves
something to be desired, but it automatically adjusts to the S speed in any
gear that the speed can be reached.

RogerN