Re: DeskNC limitation

Tim,

First of all please read the message titled Hoof 'n Mouth. It will
explain alot. I'm eating alot of 'Humble Pie" right now. My TI
calculator died on me and I replaced it with an HP. I was misreading
the display because it is in a different format. What can I say
but "Back to school".

> Couple of questions:
>
> Did you ever figure out how to get DeskNC to act like it is going

around the circle on the A axis or does it still go to increasing
negative numbers when you jog backwards from zero? The release I
played with seemed as if it was handling the A axis as just another
linear axis.

DeskNC handles degrees in decimal format and it handles them just
like some of the commercial machines do. It has a way of keeping
track of it's position in relation to the A0 position in absolute
mode. So it knows which direction to rotate and how many rotations to
go back to the A0. This is actually a handy feature when your doing
a job in which you have to keep track of the accumulated degrees of
rotation. This is true forth axis positioning. This software appears
to have the capability to machine a varible lead screw with no
problem; i.e. the screw in a plastic injection molding machine. The
rotary table on the Matsuura 760V (Yasnak MX3 control) at the shop
I'm employed at works exactly the same way in absolute mode. (We have
a Fadal on the other hand that can't handle anything over 360
degrees. After one rotation you have to start at zero again).

I had to figure out how many pulses I needed per 1 degree. Then I
entered this into the setup information for the machine under pulses
per unit.

The rotary table alone (Troyke 12") has a 15 second accuracy. This
table has a 90:1 ratio. 90 turns on the handwheel for 360 degrees
rotation. One turn = 4 degrees rotation. It was decided to use a 5:1
toothed timing belt reduction to get within the tables accuracy. That
equates to 2000 half step pulses for 4 degrees rotation with the 5:1
toothed belt/pully reduction. 500 pulses being equal to 1 degree of
rotation. 1 pulse = 7 & 20/100 seconds. Well within our tables
accuracy of 15 seconds.

> How exactly are you using this rotary table? I know you told me

some, but I am not exactly clear. If I remember correctly you are
wanting to drive it totally independent of the other axis's on the
machine. Sort of like an automated dividing head. If this is correct
so that all you want to do is rapid to the next position and then
just hold until you command another rapid, it would not be very
difficult to write a little Q-basic program to drive it.

>
> Tim
> [Denver, CO]

We are going to use the table for for indexing mostly. Some of the
work we do involves a lot of odd angles. Also, some of the people
that will use this have expressed concern about positioning accuracy
for gear cutting. Most of them however, are not computer literate. We
are not sure where we are going with this yet. This rotary table
conversion was to try out the available products to see how it might
work for us. There has been some mention of putting together a CNC
waterjet but it's just talk for now. I'm breaking ground with this
modification as no one I work with has tried it before. I work for a
small family owned shop employing around 10 manual and CNC machinist
total at this time. I'm lucky to work for people who trust my
judgment and were willing to foot the bill to let me learn about this.

We wanted something that would eliminate some of the human error that
occurs from time to time. No body is perfect 100% of the time. (It's
like they say, never buy a vehicle that was built on a Monday or a
Friday... just kidding ;-). Since all of our CNCs use G-code it only
made sence to find a piece of software that that ran it for the table.
I just wish I hadn't made a fool of myself misreading my calculator.


Donald Brock
Baton Rouge, LA