开云体育

Art,
Is this something you could add a toggle for? Radius vs. diametral
lathe input?

Does any one else have a opinion about this?

Thx,
Pete

I am typing the programs in by hand, no CAM help. Every other CNC
lathe (Fanuc, Mazak, Citizen) I've run has been Diameter input

and

have difficulty thinking in radius. Would it help if I changed

the

steps/rev on the driver instead of on the "motor tuning" page?

No - your axis should move the distance it is commanded, not half.

Just carry on writing it the way you are then divide only the X

values

by 2, gets a little more difficult with IK arcs, but not too
difficult.

You need to change your post post processor to Radius output, it
should move .1 for a .1 output, which would take .2 off the

diameter.

If it was diametric it would move .05 for .1 off the diameter.

--
Steve Blackmore

Art, Robert

Thanks, now I have one more thing to add to the list of things to buy.

John Guenther
'Ye Olde Pen Maker'
Sterling, Virginia

Steve:

We're still on version 1.003, but it is a new 1.003. It is online now.

1) Adder will no longer climb.
2) Separate debounce for Spindle index.
3) Dwell in millisecoinds bug fixed. It will switch immediately to or from
milliseconds mode when selected or changed.

This version will stop the climbing of the adder, and I don't really think
you need any debounce. I just ran several tests, using a step/dir controlled
spindle. The spindle is virtual, I really controlled the spindle with a pot.
I tryed vary'ing the speed of the spindle during a G32 to see the various
effects.

1) Keep an eye on the "Time derivitive during the thread. Is it stable? Does
it change by much?
2) Is your Current Spindle count about equal to your normal spindle count
dutring a thread. (Do all this in the air, no need to actually thread.)
3) Try this by starting your spindle with a M3 command, and then after the
spindle is up to speed, in the threading diags, just do something like
G32Z5F.2
Is this similar to your testing proceedure, or are you running a file? If
so, send it to me please. I think we're close to a breakthrough here. Like
most of the code of Mach2 (over 70,000 lines I think), I gradually forget
the specifics of the algorithms, but I have studied very deeply tonight so I
understand most of the interactions now. (I need a refresher course every
once in a while.)
Your numbers on this test will be interesting and may point the way. I'm
most interested in the numbers (approx) during the cut and the time
correction derivitive is very important.

Thanks, Good luck,
Art
www.artofcnc.ca

On Fri, 05 Sep 2003 17:25:35 -0300, you wrote:

Well, the adder values are correct for the sensed speeds, but the sensed
speeds are still screwed up. I will redo this section, hopefully tonight,
and see what can be done. Separate debounce setting for index is obviously a
must and will be added immediately. I will try to simulate this problem and
see what happens. I'll set up a step/dir control on my spindle output to see
if I can make these same values happen..

I'll let you know, Thanks for the tests, I think we're slowly getting
there...

Thanks Art, I'm thinking I should spend more time getting PWM linear
<G>.
--
Steve Blackmore

Step/Dir spindle sends out a scaled constant step/dir pulse to control the
speed of the spindle. It is ramped up and down just like a motor. The
thought was I could use a Gecko320 and a 1-2HP servo motor as a Lathe
spindle motor. Not only could I get a very low speed, but I could do 3d
images on the lathe by moving the spindle as a controlled axis for 3d work.
I will eventually get there some day. (I even have the Gecko320 waiting for
it.)

Art
www.artofcnc.ca

John,

It is a method for controlling the speed of a spindle through Mach2. My
partner Jim Cullins is working a special PC board to use with Mach2.

Bob Campbell

Art,

Not meaning to sound stupid, but what is a step and direction spindle? I
may have missed something is a message over the past couple of weeks or so.

John Guenther
'Ye Olde Pen Maker'
Sterling, Virginia

Thanks John:

It may indeed.

Art
www.artofcnc.ca

--- In mach1mach2cnc@..., Art <fenerty@a...> wrote:
<snip>
I will redo this section, hopefully tonight,

and see what can be done. Separate debounce setting for index is

obviously a

must and will be added immediately. I will try to simulate this

problem and

see what happens. I'll set up a step/dir control on my spindle

output to see

if I can make these same values happen..

<snip>

I will have a step/dir spindle on Monday (with luck) - you may have
got success by then but if not then another set of results might
help!

John Prentice

Isak:

Counts per rev wil come as soon as Threading is being done as well as it
can be on single pulse. I don't want to complicate the debugging. This code
is really rocket science code as far as complexity , but I hope soon to have
it solved, (Thanks to all of Steves tests) and then I will add multi count
threading and things will get better from there.

A separate debounce for SPindle will be in the next version, though the
latest test shows that debounce is probabaly not the culprit and could be
removed.

Just a matter of time.
Thanks,
Art
www.artofcnc.ca

Steve:

Well, the adder values are correct for the sensed speeds, but the sensed
speeds are still screwed up. I will redo this section, hopefully tonight,
and see what can be done. Separate debounce setting for index is obviously a
must and will be added immediately. I will try to simulate this problem and
see what happens. I'll set up a step/dir control on my spindle output to see
if I can make these same values happen..

I'll let you know, Thanks for the tests, I think we're slowly getting
there...
Art
www.artofcnc.ca

Roger:

If the errors your getting are "Radius not equal" or similar, then its
probably just the IJ mode. This can be toggled on the config/state screen in
Mach2.
Mach2 will run anything Master5 will, but it is more complex to set up
initially. Let me know the exact errors , or send me a troublesome file, and
I'll let you know how to proceed.

Thanks, Sorry for the trouble.
Art
www.artofcnc.ca

Ron:

In Mach1 or Mach2 you can zoom by holding the Shift key as you move the
mouse. You cannot lock it into 2d mode yet though.

Art
www.artofcnc.ca

Thank you to all for help.
The advices I got were great help.

It was not easy but there were two separate thing to make it work:

First I did need were resistors between lines because of long cable. this is
something what is described in Rutex manuals.
Second the phasing of pulse lines were wery close to the limit where
direction of rotation is sensed, in clear words I was connecting wrong lines
as A and A- etc...

Rutex at least R9xxH models can use differential encoders directly, maybe if
same development cycle will continue then next year Rutex will support
absolute encoders. ;)

Now I have 5 pc of R991H in use with mitsubishi servos.

-Mika

On Fri, 05 Sep 2003 16:06:00 -0000, you wrote:

I am typing the programs in by hand, no CAM help. Every other CNC
lathe (Fanuc, Mazak, Citizen) I've run has been Diameter input and
have difficulty thinking in radius. Would it help if I changed the
steps/rev on the driver instead of on the "motor tuning" page?

No - your axis should move the distance it is commanded, not half.

Just carry on writing it the way you are then divide only the X values
by 2, gets a little more difficult with IK arcs, but not too
difficult.

You need to change your post post processor to Radius output, it
should move .1 for a .1 output, which would take .2 off the

diameter.

If it was diametric it would move .05 for .1 off the diameter.

--
Steve Blackmore

I am typing the programs in by hand, no CAM help. Every other CNC
lathe (Fanuc, Mazak, Citizen) I've run has been Diameter input and
have difficulty thinking in radius. Would it help if I changed the
steps/rev on the driver instead of on the "motor tuning" page?

Pete

You need to change your post post processor to Radius output, it
should move .1 for a .1 output, which would take .2 off the

diameter.

If it was diametric it would move .05 for .1 off the diameter.
--
Steve Blackmore

On Fri, 05 Sep 2003 13:47:03 -0000, you wrote:

I am retrofitting a Hardinge HC with Mach2 and am having trouble with
the radius / Dia programming. I have 2 identical drives and ball
screw setups, I have set the steps/rev to be half on the X ( dia)
axis. It moves .05 for a .10 command, like intended. But the arcs are
oval. Do you have to program this like a mill? or is there a way to
change the config?

You need to change your post post processor to Radius output, it
should move .1 for a .1 output, which would take .2 off the diameter.

If it was diametric it would move .05 for .1 off the diameter.
--
Steve Blackmore

Hello All,
I am retrofitting a Hardinge HC with Mach2 and am having trouble with
the radius / Dia programming. I have 2 identical drives and ball
screw setups, I have set the steps/rev to be half on the X ( dia)
axis. It moves .05 for a .10 command, like intended. But the arcs are
oval. Do you have to program this like a mill? or is there a way to
change the config?
Pete

Hi,
I found the debounce argument.
Thanks.

--- In mach1mach2cnc@..., "sisakl" <isak@h...> wrote:

Hi, Art.
First of all, thanks for fixing the backladh problem.

A few spindle/index questions:
Can the spindle index debounce be disabled ?
I got a slotted disc, and it seems that the width of the slot is

too

small for the program to capture.

Is it possible to add a counts per rev argument for the index

input ?

I think it could improve the threading operation.

Thanks, Isak.

Hi, Art.
First of all, thanks for fixing the backladh problem.

A few spindle/index questions:
Can the spindle index debounce be disabled ?
I got a slotted disc, and it seems that the width of the slot is too
small for the program to capture.

Is it possible to add a counts per rev argument for the index input ?
I think it could improve the threading operation.

Thanks, Isak.