Ignoramus31989 wrote:
Yes. On the machine itself, there are several buttons among many:

OK, thanks

1) E-Stop
2) Green START
3) Red STOP

So, I think that E-Stop should function, like you say, through the
estop input of PPMC and cut off all motion power.


Your plan to have the EStop button shut down motion power is good. You
mentioned software problems, but another common source of uncontrolled
motion is the loss of an encoder signal. Some of the fancier servo amps
can detect this and shut down (especially if they have a tach input as
well as the encoder input), but many will pound max drive to the axis as
the integral error builds up with no encoder to correct it. I have seen
broken castings and cracked ball nuts result from this problem.

Good Luck,
BobH

Just wanted to do a little sanity check.

The question is about wiring power for all motive power (XYZ and
spindle) on this Bridgeport CNC mill.

The mill will be wired for single phase, with spindle motor on a VFD.

My current plan is as follows.

1. The CNC PC would be fed with incoming power without ANY regard for
any switches. When I need to shut it down I would type "sudo shutdown". I
will use this PC for many other shop purposes.

2. Have a switch (original switch on the door) be the first control of
incoming power to the mill (besides the PC, see 1)

3. Have all incoming power go through two fuses.

4. Jon's PPMC control box will be powered by 220v (more like 243 VAC
in reality). So any one fuse that blows, would cut the power to PPMC
control box.

5. Servo power supply, as well as power to the spindle VFD, will be
provided through a contactor.

6. Regular start and stop buttons would simply send signals to
software via PPMC.

7. The motive power contactor is actuated by a solid state relay built
into PPMC, but its control line will also go through a relay
controlled by estops, so that pressing ESTOP automatically cuts all
power to any moving equipment. I may need an extra relay in the estop
circuit, so that ESTOP not only cuts off motive power, but also sends
a signal to the control box and thus to the software that estop
occurred.

What I like about this approach is that it is relatively idiot proof
(estop means stop regardless of any software mistakes), and yet by
preserving power going into PPMC, I can keep track of my position
etc. It also keeps the PC going.

The minus is that in case of estop, the motors will coast to stop
instead of braking. But it is more simple.

Does this make any sense?

i

Ignoramus31989 wrote:

Just wanted to do a little sanity check.

The question is about wiring power for all motive power (XYZ and
spindle) on this Bridgeport CNC mill.

The mill will be wired for single phase, with spindle motor on a VFD.

My current plan is as follows.

1. The CNC PC would be fed with incoming power without ANY regard for
any switches. When I need to shut it down I would type "sudo shutdown". I
will use this PC for many other shop purposes.

2. Have a switch (original switch on the door) be the first control of
incoming power to the mill (besides the PC, see 1)

3. Have all incoming power go through two fuses.

4. Jon's PPMC control box will be powered by 220v (more like 243 VAC
in reality). So any one fuse that blows, would cut the power to PPMC
control box.

5. Servo power supply, as well as power to the spindle VFD, will be
provided through a contactor.

6. Regular start and stop buttons would simply send signals to
software via PPMC.

7. The motive power contactor is actuated by a solid state relay built
into PPMC, but its control line will also go through a relay
controlled by estops, so that pressing ESTOP automatically cuts all
power to any moving equipment. I may need an extra relay in the estop
circuit, so that ESTOP not only cuts off motive power, but also sends
a signal to the control box and thus to the software that estop
occurred.

What I like about this approach is that it is relatively idiot proof
(estop means stop regardless of any software mistakes), and yet by
preserving power going into PPMC, I can keep track of my position
etc. It also keeps the PC going.

The minus is that in case of estop, the motors will coast to stop
instead of braking. But it is more simple.

Does this make any sense?

i

Latching E-Stop pushbutton with DPST NC contacts, one contact set
interrupts power to the motion power contactor, and the other set
interrupts the estop signal to the control. No relays required befony
the motion power contactor.

On 2010-06-25, Pete C. wrote:

Ignoramus31989 wrote:

Just wanted to do a little sanity check.

The question is about wiring power for all motive power (XYZ and
spindle) on this Bridgeport CNC mill.

The mill will be wired for single phase, with spindle motor on a VFD.

My current plan is as follows.

1. The CNC PC would be fed with incoming power without ANY regard for
any switches. When I need to shut it down I would type "sudo shutdown". I
will use this PC for many other shop purposes.

2. Have a switch (original switch on the door) be the first control of
incoming power to the mill (besides the PC, see 1)

3. Have all incoming power go through two fuses.

4. Jon's PPMC control box will be powered by 220v (more like 243 VAC
in reality). So any one fuse that blows, would cut the power to PPMC
control box.

5. Servo power supply, as well as power to the spindle VFD, will be
provided through a contactor.

6. Regular start and stop buttons would simply send signals to
software via PPMC.

7. The motive power contactor is actuated by a solid state relay built
into PPMC, but its control line will also go through a relay
controlled by estops, so that pressing ESTOP automatically cuts all
power to any moving equipment. I may need an extra relay in the estop
circuit, so that ESTOP not only cuts off motive power, but also sends
a signal to the control box and thus to the software that estop
occurred.

What I like about this approach is that it is relatively idiot proof
(estop means stop regardless of any software mistakes), and yet by
preserving power going into PPMC, I can keep track of my position
etc. It also keeps the PC going.

The minus is that in case of estop, the motors will coast to stop
instead of braking. But it is more simple.

Does this make any sense?

i

Latching E-Stop pushbutton with DPST NC contacts, one contact set
interrupts power to the motion power contactor, and the other set
interrupts the estop signal to the control. No relays required befony
the motion power contactor.

I will try to reuse my estop buttons, I was not sure if they are DP.

i

On a second thoughts, I was reading about PPMC on the linuxcnc
website. It would seem that PPMC includes this functionality already:
there is a relay on it that is switched off when estop is activated.

What this means is that I can wire the motive power contactor through
that relay in ppmc and it will do what I want: when estop is
activated, the motive power contactor drops out.

i

Ignoramus31989 wrote:

On 2010-06-25, Pete C. wrote:

Ignoramus31989 wrote:

Just wanted to do a little sanity check.

The question is about wiring power for all motive power (XYZ and
spindle) on this Bridgeport CNC mill.

The mill will be wired for single phase, with spindle motor on a VFD.

My current plan is as follows.

1. The CNC PC would be fed with incoming power without ANY regard for
any switches. When I need to shut it down I would type "sudo shutdown". I
will use this PC for many other shop purposes.

2. Have a switch (original switch on the door) be the first control of
incoming power to the mill (besides the PC, see 1)

3. Have all incoming power go through two fuses.

4. Jon's PPMC control box will be powered by 220v (more like 243 VAC
in reality). So any one fuse that blows, would cut the power to PPMC
control box.

5. Servo power supply, as well as power to the spindle VFD, will be
provided through a contactor.

6. Regular start and stop buttons would simply send signals to
software via PPMC.

7. The motive power contactor is actuated by a solid state relay built
into PPMC, but its control line will also go through a relay
controlled by estops, so that pressing ESTOP automatically cuts all
power to any moving equipment. I may need an extra relay in the estop
circuit, so that ESTOP not only cuts off motive power, but also sends
a signal to the control box and thus to the software that estop
occurred.

What I like about this approach is that it is relatively idiot proof
(estop means stop regardless of any software mistakes), and yet by
preserving power going into PPMC, I can keep track of my position
etc. It also keeps the PC going.

The minus is that in case of estop, the motors will coast to stop
instead of braking. But it is more simple.

Does this make any sense?

i

Latching E-Stop pushbutton with DPST NC contacts, one contact set
interrupts power to the motion power contactor, and the other set
interrupts the estop signal to the control. No relays required befony
the motion power contactor.

I will try to reuse my estop buttons, I was not sure if they are DP.

i

Most are modular and you can get different contact blocks for them. New
ones aren't too expensive either if you hunt around a bit.

Ignoramus31989 wrote:

On a second thoughts, I was reading about PPMC on the linuxcnc
website. It would seem that PPMC includes this functionality already:
there is a relay on it that is switched off when estop is activated.

What this means is that I can wire the motive power contactor through
that relay in ppmc and it will do what I want: when estop is
activated, the motive power contactor drops out.

i

I don't know how that relay is controlled, but the NC contacts of the
latching E-Stop button opening the circuit to the motion power contactor
is about the most direct and failsafe setup you can get.

"Ignoramus31989" wrote in message
...
On a second thoughts, I was reading about PPMC on the linuxcnc
website. It would seem that PPMC includes this functionality already:
there is a relay on it that is switched off when estop is activated.

What this means is that I can wire the motive power contactor through
that relay in ppmc and it will do what I want: when estop is
activated, the motive power contactor drops out.

i

You got a pretty good plan there iggy. I'd suggest you also have a machine
stop button. This stops motion and spindle through software (servo inhibit
and VFD brake etc.). This is the normal everyday stop button. I also like a
spindle stop button. Buttons are cheap. this assumes inputs aren't a
limiting issue in EMC. I know Mach 3 really limits you on inputs for this
sort of stuff.

karl

Karl Townsend wrote:

"Ignoramus31989" wrote in message
...
On a second thoughts, I was reading about PPMC on the linuxcnc
website. It would seem that PPMC includes this functionality already:
there is a relay on it that is switched off when estop is activated.

What this means is that I can wire the motive power contactor through
that relay in ppmc and it will do what I want: when estop is
activated, the motive power contactor drops out.

i

You got a pretty good plan there iggy. I'd suggest you also have a machine
stop button. This stops motion and spindle through software (servo inhibit
and VFD brake etc.). This is the normal everyday stop button. I also like a
spindle stop button. Buttons are cheap. this assumes inputs aren't a
limiting issue in EMC. I know Mach 3 really limits you on inputs for this
sort of stuff.

karl

Correction, Mach3 does not limit you on inputs for this sort of stuff.
Many implementations of Mach3 operate with few inputs, but that is not a
limitation of the program. There are a number of different ways to get
plenty of inputs to Mach3, ranging from using multiple LPT ports, to USB
connected motion engines, to USB or serial connected I/O boards, to PCI
based I/O boards. Pretty much just as flexible as EMC/EMC2, and in
either case you need to add hardware to get the inputs to the PC.

On 2010-06-25, Karl Townsend wrote:

"Ignoramus31989" wrote in message
...
On a second thoughts, I was reading about PPMC on the linuxcnc
website. It would seem that PPMC includes this functionality already:
there is a relay on it that is switched off when estop is activated.

What this means is that I can wire the motive power contactor through
that relay in ppmc and it will do what I want: when estop is
activated, the motive power contactor drops out.

i

You got a pretty good plan there iggy. I'd suggest you also have a machine
stop button. This stops motion and spindle through software (servo inhibit
and VFD brake etc.). This is the normal everyday stop button. I also like a
spindle stop button. Buttons are cheap. this assumes inputs aren't a
limiting issue in EMC. I know Mach 3 really limits you on inputs for this
sort of stuff.

Yes. On the machine itself, there are several buttons among many:

OK, thanks

1) E-Stop
2) Green START
3) Red STOP

So, I think that E-Stop should function, like you say, through the
estop input of PPMC and cut off all motion power.

The START should be a signal to EMC to start (resume) the process. The
STOP button should pause the milling program. The START button should
start, or resume, the program.

PPMC has 16 inputs. Of them, one is a E-Stop. So I have 15 inputs
left.

The first candidates for inputs are limit switches (6), Start/Stop
(2). This leaves 7 more digital inputs. Ideally, I would like to make
my X-Y-Z hand controls to work. For that, I would need at least 6
inputs.

I hope that if I need more buttons to work, I could expand this PPMC
by adding one more I/O cards.

I will make a separate post about limit switches.

i

"Ignoramus31989" wrote in message
...
On 2010-06-25, Karl Townsend wrote:

"Ignoramus31989" wrote in message
...
On a second thoughts, I was reading about PPMC on the linuxcnc
website. It would seem that PPMC includes this functionality already:
there is a relay on it that is switched off when estop is activated.

What this means is that I can wire the motive power contactor through
that relay in ppmc and it will do what I want: when estop is
activated, the motive power contactor drops out.

i

You got a pretty good plan there iggy. I'd suggest you also have a
machine
stop button. This stops motion and spindle through software (servo
inhibit
and VFD brake etc.). This is the normal everyday stop button. I also like
a
spindle stop button. Buttons are cheap. this assumes inputs aren't a
limiting issue in EMC. I know Mach 3 really limits you on inputs for this
sort of stuff.

Yes. On the machine itself, there are several buttons among many:

OK, thanks

1) E-Stop
2) Green START
3) Red STOP

So, I think that E-Stop should function, like you say, through the
estop input of PPMC and cut off all motion power.

The START should be a signal to EMC to start (resume) the process. The
STOP button should pause the milling program. The START button should
start, or resume, the program.

PPMC has 16 inputs. Of them, one is a E-Stop. So I have 15 inputs
left.

The first candidates for inputs are limit switches (6), Start/Stop
(2). This leaves 7 more digital inputs. Ideally, I would like to make
my X-Y-Z hand controls to work. For that, I would need at least 6
inputs.

I hope that if I need more buttons to work, I could expand this PPMC
by adding one more I/O cards.

I will make a separate post about limit switches.

i

Buttons I really like on my controls

1. Cycle start. that's your green button
2. Feed hold. red button - halts Gcode till you press cycle start again.
3. Single step switch. when on each cycle start runs one line of Gcode
4. optional stop switch - put an option stop M code in your program and it
will stop executing till you press cycle start.
5. spindle start. need to start spindle for edge find most often.
6. spindle stop.
7. machine stop. Put beside the estop and use this for nearly all stops.
8. Pot for spindle speed. change speed while running program
9. Pot for axis feed rate. change feed while running program.
10, Rotary Mode switch - three inputs. using base2, you can have handwheel
for all axis (four), jog, run, MDI (you can put single step on the mode
switch - line 3)
11. handweel increment switch. On for coarse, off for fine
12. six limit switches
13. three home switches
14 Nine Jog buttons in three rows of three. perimeter is X+,X-, Y+,Y-,
Z+,Z-, A+,A- center is rapid jog.

This is the sort of stuff you find on a Professional control. I'm spoilt -
I HATE using a keyboard as a substitute.

I'm sure you won't do all these, just stuff to think about.

Karl

On 2010-06-25, BobH wrote:
Ignoramus31989 wrote:
Yes. On the machine itself, there are several buttons among many:

OK, thanks

1) E-Stop
2) Green START
3) Red STOP

So, I think that E-Stop should function, like you say, through the
estop input of PPMC and cut off all motion power.


Your plan to have the EStop button shut down motion power is good. You
mentioned software problems, but another common source of uncontrolled
motion is the loss of an encoder signal. Some of the fancier servo amps
can detect this and shut down (especially if they have a tach input as
well as the encoder input), but many will pound max drive to the axis as
the integral error builds up with no encoder to correct it. I have seen
broken castings and cracked ball nuts result from this problem.

Yeah, scary.

Well, with tachometer feedback and limiting signal, I can limit speed,
at least.

Maybe EMC has some kind of logic that if there is no motion detected
when it should be detected, it would do some kind of software shutdown?

i

On 2010-06-26, Karl Townsend wrote:

"Ignoramus31989" wrote in message
...
On 2010-06-25, Karl Townsend wrote:

"Ignoramus31989" wrote in message
...
On a second thoughts, I was reading about PPMC on the linuxcnc
website. It would seem that PPMC includes this functionality already:
there is a relay on it that is switched off when estop is activated.

What this means is that I can wire the motive power contactor through
that relay in ppmc and it will do what I want: when estop is
activated, the motive power contactor drops out.

i

You got a pretty good plan there iggy. I'd suggest you also have a
machine
stop button. This stops motion and spindle through software (servo
inhibit
and VFD brake etc.). This is the normal everyday stop button. I also like
a
spindle stop button. Buttons are cheap. this assumes inputs aren't a
limiting issue in EMC. I know Mach 3 really limits you on inputs for this
sort of stuff.

Yes. On the machine itself, there are several buttons among many:

OK, thanks

1) E-Stop
2) Green START
3) Red STOP

So, I think that E-Stop should function, like you say, through the
estop input of PPMC and cut off all motion power.

The START should be a signal to EMC to start (resume) the process. The
STOP button should pause the milling program. The START button should
start, or resume, the program.

PPMC has 16 inputs. Of them, one is a E-Stop. So I have 15 inputs
left.

The first candidates for inputs are limit switches (6), Start/Stop
(2). This leaves 7 more digital inputs. Ideally, I would like to make
my X-Y-Z hand controls to work. For that, I would need at least 6
inputs.

I hope that if I need more buttons to work, I could expand this PPMC
by adding one more I/O cards.

I will make a separate post about limit switches.

i

Buttons I really like on my controls

1. Cycle start. that's your green button

i have it

2. Feed hold. red button - halts Gcode till you press cycle start
again.

I have it too

3. Single step switch. when on each cycle start runs one line of
Gcode

I have it too

4. optional stop switch - put an option stop M code in your program and it
will stop executing till you press cycle start.

no idea

5. spindle start. need to start spindle for edge find most often.

got it

6. spindle stop.

got it

7. machine stop. Put beside the estop and use this for nearly all stops.

gt it

8. Pot for spindle speed. change speed while running program

I have a pleumatic speed changer

9. Pot for axis feed rate. change feed while running program.

nope

10, Rotary Mode switch - three inputs. using base2, you can have handwheel
for all axis (four), jog, run, MDI (you can put single step on the mode
switch - line 3)
11. handweel increment switch. On for coarse, off for fine

I have something along the lines

12. six limit switches

got only five

13. three home switches

no idea what it is

14 Nine Jog buttons in three rows of three. perimeter is X+,X-, Y+,Y-,
Z+,Z-, A+,A- center is rapid jog.

This is the sort of stuff you find on a Professional control. I'm spoilt -
I HATE using a keyboard as a substitute.

I'm sure you won't do all these, just stuff to think about.

Yep, thanks Karl.

i

Ignoramus31989 wrote:

On 2010-06-26, Karl Townsend wrote:

"Ignoramus31989" wrote in message
...
On 2010-06-25, Karl Townsend wrote:

"Ignoramus31989" wrote in message
...
On a second thoughts, I was reading about PPMC on the linuxcnc
website. It would seem that PPMC includes this functionality already:
there is a relay on it that is switched off when estop is activated.

What this means is that I can wire the motive power contactor through
that relay in ppmc and it will do what I want: when estop is
activated, the motive power contactor drops out.

i

You got a pretty good plan there iggy. I'd suggest you also have a
machine
stop button. This stops motion and spindle through software (servo
inhibit
and VFD brake etc.). This is the normal everyday stop button. I also like
a
spindle stop button. Buttons are cheap. this assumes inputs aren't a
limiting issue in EMC. I know Mach 3 really limits you on inputs for this
sort of stuff.

Yes. On the machine itself, there are several buttons among many:

OK, thanks

1) E-Stop
2) Green START
3) Red STOP

So, I think that E-Stop should function, like you say, through the
estop input of PPMC and cut off all motion power.

The START should be a signal to EMC to start (resume) the process. The
STOP button should pause the milling program. The START button should
start, or resume, the program.

PPMC has 16 inputs. Of them, one is a E-Stop. So I have 15 inputs
left.

The first candidates for inputs are limit switches (6), Start/Stop
(2). This leaves 7 more digital inputs. Ideally, I would like to make
my X-Y-Z hand controls to work. For that, I would need at least 6
inputs.

I hope that if I need more buttons to work, I could expand this PPMC
by adding one more I/O cards.

I will make a separate post about limit switches.

i

Buttons I really like on my controls

1. Cycle start. that's your green button

i have it

2. Feed hold. red button - halts Gcode till you press cycle start
again.

I have it too

3. Single step switch. when on each cycle start runs one line of
Gcode

I have it too

4. optional stop switch - put an option stop M code in your program and it
will stop executing till you press cycle start.

no idea

5. spindle start. need to start spindle for edge find most often.

got it

6. spindle stop.

got it

7. machine stop. Put beside the estop and use this for nearly all stops.

gt it

8. Pot for spindle speed. change speed while running program

I have a pleumatic speed changer

9. Pot for axis feed rate. change feed while running program.

nope

Most "real" machines don't have this either. What they do typically have
is a rotary switch providing feed rate override capability in 10% steps
+/- some decent range. This allows the machine operator to tweak the
feed rates to match the real world cutting conditions, slightly dulling
tools, etc. vs. the ideal conditions the person who generated the G-code
wrote it for.


10, Rotary Mode switch - three inputs. using base2, you can have handwheel
for all axis (four), jog, run, MDI (you can put single step on the mode
switch - line 3)
11. handweel increment switch. On for coarse, off for fine

I have something along the lines

Handwheel increment rotary switch - 1.000, 0.100, 0.010, 0.001
Handwheel axis rotary switch - X,Y,Z,A


12. six limit switches

got only five

That is sufficient generally. The control doesn't really have to know
which direction the overtravel was, as long as you can hold the
momentary limit override toggle switch and manually jog the axis back
off the limit. This is how most real machines operate. Many have three
limit LEDs next to the limit override toggle switch to show which axis
is on a limit as it can sometimes be difficult to tell if more than one
axis is at the end of travel.


13. three home switches

no idea what it is

Most real machines don't have any idea on this either. The home switch
is mostly limited to home shop machines, I gave the homing procedure
that many / most of the real machines use up a thread or two. That home
procedure gives very precise homing, very fast homing, and immunity to
light chip buildup on the limit switch rollers or trip ramps.


14 Nine Jog buttons in three rows of three. perimeter is X+,X-, Y+,Y-,
Z+,Z-, A+,A- center is rapid jog.

This is the sort of stuff you find on a Professional control. I'm spoilt -
I HATE using a keyboard as a substitute.

Yes, and they can be readily implemented on a home control as well, be
it EMC/EMC2 or Mach3 based. While Karl may not have looked at Mach3
enough to realize it, both EMC/EMC2 and Mach3 provide for a pretty much
unlimited number of inputs and outputs with the appropriate hardware.
Both EMC/EMC2 and Mach3 share the same I/O limitations if you configure
them in the lowest cost configuration doing all I/O through a single LPT
port.

Ignoramus31989 wrote:

Maybe EMC has some kind of logic that if there is no motion detected
when it should be detected, it would do some kind of software shutdown?

I would suspect after a certain amount of following error an alarm would be raised.

Wes

On 2010-06-26, Pete C. wrote:
Ignoramus31989 wrote:
On 2010-06-26, Karl Townsend wrote:
9. Pot for axis feed rate. change feed while running program.

nope

Most "real" machines don't have this either. What they do typically have
is a rotary switch providing feed rate override capability in 10% steps
+/- some decent range. This allows the machine operator to tweak the
feed rates to match the real world cutting conditions, slightly dulling
tools, etc. vs. the ideal conditions the person who generated the G-code
wrote it for.

I can do that from keyboard, it seems more industrial in nature.

12. six limit switches

got only five

That is sufficient generally. The control doesn't really have to know
which direction the overtravel was, as long as you can hold the
momentary limit override toggle switch and manually jog the axis back
off the limit. This is how most real machines operate. Many have three
limit LEDs next to the limit override toggle switch to show which axis
is on a limit as it can sometimes be difficult to tell if more than one
axis is at the end of travel.

But this means that, for homing with limit switches, theq limit
switches should not automatically inhibit drives via the inhibit
function, right? At least on my Y axis with only one limit switch?
And so, EMC should be responsible for proper behavior when limits are
reached. Right?

i

"Ignoramus11285" wrote in message
...
On 2010-06-26, Pete C. wrote:
Ignoramus31989 wrote:
On 2010-06-26, Karl Townsend wrote:
9. Pot for axis feed rate. change feed while running program.

nope

Most "real" machines don't have this either. What they do typically have
is a rotary switch providing feed rate override capability in 10% steps
+/- some decent range. This allows the machine operator to tweak the
feed rates to match the real world cutting conditions, slightly dulling
tools, etc. vs. the ideal conditions the person who generated the G-code
wrote it for.

I can do that from keyboard, it seems more industrial in nature.

12. six limit switches

got only five

That is sufficient generally. The control doesn't really have to know
which direction the overtravel was, as long as you can hold the
momentary limit override toggle switch and manually jog the axis back
off the limit. This is how most real machines operate. Many have three
limit LEDs next to the limit override toggle switch to show which axis
is on a limit as it can sometimes be difficult to tell if more than one
axis is at the end of travel.

But this means that, for homing with limit switches, theq limit
switches should not automatically inhibit drives via the inhibit
function, right? At least on my Y axis with only one limit switch?
And so, EMC should be responsible for proper behavior when limits are
reached. Right?

i

My lathe that I use with EMC has 2 limit switches and a separate home
switch, the home gets made shortly before one ends limit switch.

RogerN

Ignoramus11285 wrote:

On 2010-06-26, Pete C. wrote:
Ignoramus31989 wrote:
On 2010-06-26, Karl Townsend wrote:
9. Pot for axis feed rate. change feed while running program.

nope

Most "real" machines don't have this either. What they do typically have
is a rotary switch providing feed rate override capability in 10% steps
+/- some decent range. This allows the machine operator to tweak the
feed rates to match the real world cutting conditions, slightly dulling
tools, etc. vs. the ideal conditions the person who generated the G-code
wrote it for.

I can do that from keyboard, it seems more industrial in nature.

Yes, as a HSM guy, you don't need quite the same programmer vs. operator
setup.


12. six limit switches

got only five

That is sufficient generally. The control doesn't really have to know
which direction the overtravel was, as long as you can hold the
momentary limit override toggle switch and manually jog the axis back
off the limit. This is how most real machines operate. Many have three
limit LEDs next to the limit override toggle switch to show which axis
is on a limit as it can sometimes be difficult to tell if more than one
axis is at the end of travel.

But this means that, for homing with limit switches, theq limit
switches should not automatically inhibit drives via the inhibit
function, right? At least on my Y axis with only one limit switch?
And so, EMC should be responsible for proper behavior when limits are
reached. Right?

Yes, and no. You can actually have it both ways, since you're a
programmer. Your limits can connect directly to the servo inhibits and
you can have your "limit override" momentary toggle switch actually
disconnect and override the inhibit signals to the servos (use a 4 pole
switch). That way you still have true hard limits even if EMC goes nuts
for some reason, and the real ability to override the hard limits to
manually jog the axis back off the limits. The three LEDs limit LEDs
should be able to be directly driven from the inhibit signals to the
servos.

My lathe that I use with EMC has 2 limit switches and a separate home
switch, the home gets made shortly before one ends limit switch.

RogerN


I've got four machines running with separate home switches. This is much
easier to set up but uses more inputs (I'm limited to only 128 inputs). My
CNC knee mill, similar to yours, uses homing off the limit switch. I got it
working nicely after a fair bit of effort. One solution I did was run the
inhibit through the control, not hardwired. At least on this machine it
would only work well with very slow speeds and accelerations - means homing
takes more than a minute vs. a few seconds on the other machines. Annoying
if you are developiong something else and rebooting the control constantly.
This may be a "feature" of Camsoft and you won't see trouble in EMC. (The
Galil board, behind Camsoft, has homing off home switches built into it)

I'd ask an EMC guru if they are homing off the limits and how they handle
the inhibit when doing this.

Karl

"Pete C." wrote:

Ignoramus11285 wrote:

On 2010-06-26, Pete C. wrote:
Ignoramus31989 wrote:
On 2010-06-26, Karl Townsend wrote:
9. Pot for axis feed rate. change feed while running program.

nope

Most "real" machines don't have this either. What they do typically have
is a rotary switch providing feed rate override capability in 10% steps
+/- some decent range. This allows the machine operator to tweak the
feed rates to match the real world cutting conditions, slightly dulling
tools, etc. vs. the ideal conditions the person who generated the G-code
wrote it for.

I can do that from keyboard, it seems more industrial in nature.

Yes, as a HSM guy, you don't need quite the same programmer vs. operator
setup.


12. six limit switches

got only five

That is sufficient generally. The control doesn't really have to know
which direction the overtravel was, as long as you can hold the
momentary limit override toggle switch and manually jog the axis back
off the limit. This is how most real machines operate. Many have three
limit LEDs next to the limit override toggle switch to show which axis
is on a limit as it can sometimes be difficult to tell if more than one
axis is at the end of travel.

But this means that, for homing with limit switches, theq limit
switches should not automatically inhibit drives via the inhibit
function, right? At least on my Y axis with only one limit switch?
And so, EMC should be responsible for proper behavior when limits are
reached. Right?

Yes, and no. You can actually have it both ways, since you're a
programmer. Your limits can connect directly to the servo inhibits and
you can have your "limit override" momentary toggle switch actually
disconnect and override the inhibit signals to the servos (use a 4 pole
switch). That way you still have true hard limits even if EMC goes nuts
for some reason, and the real ability to override the hard limits to
manually jog the axis back off the limits. The three LEDs limit LEDs
should be able to be directly driven from the inhibit signals to the
servos.

Forgot to add: If you need to move off limits under EMC control, as in
for a homing process, you can use an output from EMC to operate a relay
(4 pole) in series with the limit override toggle switch, to allow the
control to override limits during homing.

On 2010-06-27, Pete C. wrote:

"Pete C." wrote:

Ignoramus11285 wrote:

On 2010-06-26, Pete C. wrote:
Ignoramus31989 wrote:
On 2010-06-26, Karl Townsend wrote:
9. Pot for axis feed rate. change feed while running program.

nope

Most "real" machines don't have this either. What they do typically have
is a rotary switch providing feed rate override capability in 10% steps
+/- some decent range. This allows the machine operator to tweak the
feed rates to match the real world cutting conditions, slightly dulling
tools, etc. vs. the ideal conditions the person who generated the G-code
wrote it for.

I can do that from keyboard, it seems more industrial in nature.

Yes, as a HSM guy, you don't need quite the same programmer vs. operator
setup.


12. six limit switches

got only five

That is sufficient generally. The control doesn't really have to know
which direction the overtravel was, as long as you can hold the
momentary limit override toggle switch and manually jog the axis back
off the limit. This is how most real machines operate. Many have three
limit LEDs next to the limit override toggle switch to show which axis
is on a limit as it can sometimes be difficult to tell if more than one
axis is at the end of travel.

But this means that, for homing with limit switches, theq limit
switches should not automatically inhibit drives via the inhibit
function, right? At least on my Y axis with only one limit switch?
And so, EMC should be responsible for proper behavior when limits are
reached. Right?

Yes, and no. You can actually have it both ways, since you're a
programmer. Your limits can connect directly to the servo inhibits and
you can have your "limit override" momentary toggle switch actually
disconnect and override the inhibit signals to the servos (use a 4 pole
switch). That way you still have true hard limits even if EMC goes nuts
for some reason, and the real ability to override the hard limits to
manually jog the axis back off the limits. The three LEDs limit LEDs
should be able to be directly driven from the inhibit signals to the
servos.

Forgot to add: If you need to move off limits under EMC control, as in
for a homing process, you can use an output from EMC to operate a relay
(4 pole) in series with the limit override toggle switch, to allow the
control to override limits during homing.

That's smart, looks like a way to go.

I think that hitting a limit, as such, is not a disaster if it is
handled right. So I would not want a limit to do a full estop.

I am really not sure what I want at this point, what is the ideal
estop behavior and what EMC can do with estop signals. Maybe I will
ask on the EMC forums.

i

On 2010-06-27, Karl Townsend wrote:

My lathe that I use with EMC has 2 limit switches and a separate home
switch, the home gets made shortly before one ends limit switch.

RogerN


I've got four machines running with separate home switches. This is much
easier to set up but uses more inputs (I'm limited to only 128 inputs). My
CNC knee mill, similar to yours, uses homing off the limit switch. I got it
working nicely after a fair bit of effort. One solution I did was run the
inhibit through the control, not hardwired. At least on this machine it
would only work well with very slow speeds and accelerations - means homing
takes more than a minute vs. a few seconds on the other machines. Annoying
if you are developiong something else and rebooting the control constantly.
This may be a "feature" of Camsoft and you won't see trouble in EMC. (The
Galil board, behind Camsoft, has homing off home switches built into it)

I'd ask an EMC guru if they are homing off the limits and how they handle
the inhibit when doing this.

I will head right to EMC forums now

On Sun, 27 Jun 2010 11:30:44 -0500, Ignoramus12901
wrote:

On 2010-06-27, Pete C. wrote:

"Pete C." wrote:

Ignoramus11285 wrote:

On 2010-06-26, Pete C. wrote:
Ignoramus31989 wrote:
On 2010-06-26, Karl Townsend wrote:
9. Pot for axis feed rate. change feed while running program.

nope

Most "real" machines don't have this either. What they do typically have
is a rotary switch providing feed rate override capability in 10% steps
+/- some decent range. This allows the machine operator to tweak the
feed rates to match the real world cutting conditions, slightly dulling
tools, etc. vs. the ideal conditions the person who generated the G-code
wrote it for.

I can do that from keyboard, it seems more industrial in nature.

Yes, as a HSM guy, you don't need quite the same programmer vs. operator
setup.


12. six limit switches

got only five

That is sufficient generally. The control doesn't really have to know
which direction the overtravel was, as long as you can hold the
momentary limit override toggle switch and manually jog the axis back
off the limit. This is how most real machines operate. Many have three
limit LEDs next to the limit override toggle switch to show which axis
is on a limit as it can sometimes be difficult to tell if more than one
axis is at the end of travel.

But this means that, for homing with limit switches, theq limit
switches should not automatically inhibit drives via the inhibit
function, right? At least on my Y axis with only one limit switch?
And so, EMC should be responsible for proper behavior when limits are
reached. Right?

Yes, and no. You can actually have it both ways, since you're a
programmer. Your limits can connect directly to the servo inhibits and
you can have your "limit override" momentary toggle switch actually
disconnect and override the inhibit signals to the servos (use a 4 pole
switch). That way you still have true hard limits even if EMC goes nuts
for some reason, and the real ability to override the hard limits to
manually jog the axis back off the limits. The three LEDs limit LEDs
should be able to be directly driven from the inhibit signals to the
servos.

Forgot to add: If you need to move off limits under EMC control, as in
for a homing process, you can use an output from EMC to operate a relay
(4 pole) in series with the limit override toggle switch, to allow the
control to override limits during homing.

That's smart, looks like a way to go.

I think that hitting a limit, as such, is not a disaster if it is
handled right. So I would not want a limit to do a full estop.

I am really not sure what I want at this point, what is the ideal
estop behavior and what EMC can do with estop signals. Maybe I will
ask on the EMC forums.

i

Typically...there are two switches on each axis end. The first one is
for a warning/control change..the other is a home switch

When you run into the first switch..it signals the control that it
NEEDS to slow down and now..and after it does..it gently approaches the
home switch and stops where its supposed to, rather than slamming into
the hard stops as it overtravels beyond the home switch while
desperately trying to slow down.

Just a heads up.


Gunner

One could not be a successful Leftwinger without realizing that,
in contrast to the popular conception supported by newspapers
and mothers of Leftwingers, a goodly number of Leftwingers are
not only narrow-minded and dull, but also just stupid.
Gunner Asch

In article ,
Ignoramus31989 wrote:

Just wanted to do a little sanity check.

The question is about wiring power for all motive power (XYZ and
spindle) on this Bridgeport CNC mill.

The mill will be wired for single phase, with spindle motor on a VFD.

My current plan is as follows.

1. The CNC PC would be fed with incoming power without ANY regard for
any switches. When I need to shut it down I would type "sudo shutdown". I
will use this PC for many other shop purposes.

2. Have a switch (original switch on the door) be the first control of
incoming power to the mill (besides the PC, see 1)

3. Have all incoming power go through two fuses.

4. Jon's PPMC control box will be powered by 220v (more like 243 VAC
in reality). So any one fuse that blows, would cut the power to PPMC
control box.

5. Servo power supply, as well as power to the spindle VFD, will be
provided through a contactor.

6. Regular start and stop buttons would simply send signals to
software via PPMC.

7. The motive power contactor is actuated by a solid state relay built
into PPMC, but its control line will also go through a relay
controlled by estops, so that pressing ESTOP automatically cuts all
power to any moving equipment. I may need an extra relay in the estop
circuit, so that ESTOP not only cuts off motive power, but also sends
a signal to the control box and thus to the software that estop
occurred.

What I like about this approach is that it is relatively idiot proof
(estop means stop regardless of any software mistakes), and yet by
preserving power going into PPMC, I can keep track of my position
etc. It also keeps the PC going.

The minus is that in case of estop, the motors will coast to stop
instead of braking. But it is more simple.

Does this make any sense?

In addition to all the good points made by others, I would add one thing:
Ensure that if prime power is lost and later regained, all motion will stop and
stay stopped until you give permission.

I would also go through your design and ask the question for each and every
component: what happens if this fails (stuck on, stuck off, or stuck open)? I
particularly liked the part about busted encoders breaking machines as the
servos manfully tried to achieve the unachievable.

Joe Gwinn

Ignoramus12901 wrote:

On 2010-06-27, Pete C. wrote:

"Pete C." wrote:

Ignoramus11285 wrote:

On 2010-06-26, Pete C. wrote:
Ignoramus31989 wrote:
On 2010-06-26, Karl Townsend wrote:
9. Pot for axis feed rate. change feed while running program.

nope

Most "real" machines don't have this either. What they do typically have
is a rotary switch providing feed rate override capability in 10% steps
+/- some decent range. This allows the machine operator to tweak the
feed rates to match the real world cutting conditions, slightly dulling
tools, etc. vs. the ideal conditions the person who generated the G-code
wrote it for.

I can do that from keyboard, it seems more industrial in nature.

Yes, as a HSM guy, you don't need quite the same programmer vs. operator
setup.


12. six limit switches

got only five

That is sufficient generally. The control doesn't really have to know
which direction the overtravel was, as long as you can hold the
momentary limit override toggle switch and manually jog the axis back
off the limit. This is how most real machines operate. Many have three
limit LEDs next to the limit override toggle switch to show which axis
is on a limit as it can sometimes be difficult to tell if more than one
axis is at the end of travel.

But this means that, for homing with limit switches, theq limit
switches should not automatically inhibit drives via the inhibit
function, right? At least on my Y axis with only one limit switch?
And so, EMC should be responsible for proper behavior when limits are
reached. Right?

Yes, and no. You can actually have it both ways, since you're a
programmer. Your limits can connect directly to the servo inhibits and
you can have your "limit override" momentary toggle switch actually
disconnect and override the inhibit signals to the servos (use a 4 pole
switch). That way you still have true hard limits even if EMC goes nuts
for some reason, and the real ability to override the hard limits to
manually jog the axis back off the limits. The three LEDs limit LEDs
should be able to be directly driven from the inhibit signals to the
servos.

Forgot to add: If you need to move off limits under EMC control, as in
for a homing process, you can use an output from EMC to operate a relay
(4 pole) in series with the limit override toggle switch, to allow the
control to override limits during homing.

That's smart, looks like a way to go.

I think that hitting a limit, as such, is not a disaster if it is
handled right. So I would not want a limit to do a full estop.

I am really not sure what I want at this point, what is the ideal
estop behavior and what EMC can do with estop signals. Maybe I will
ask on the EMC forums.

i

I would think you'd want to configure EMC so that a limit trip would
cause a normal stop. For another level of safety, you could control the
limit override relay from two EMC controlled outputs, setup so one has
to be high and one low to override the limits, so something like a crash
and port reset can't accidentally trip it since a port reset will send
both outputs to the same state.

Gunner Asch wrote:

On Sun, 27 Jun 2010 11:30:44 -0500, Ignoramus12901
wrote:

On 2010-06-27, Pete C. wrote:

"Pete C." wrote:

Ignoramus11285 wrote:

On 2010-06-26, Pete C. wrote:
Ignoramus31989 wrote:
On 2010-06-26, Karl Townsend wrote:
9. Pot for axis feed rate. change feed while running program.

nope

Most "real" machines don't have this either. What they do typically have
is a rotary switch providing feed rate override capability in 10% steps
+/- some decent range. This allows the machine operator to tweak the
feed rates to match the real world cutting conditions, slightly dulling
tools, etc. vs. the ideal conditions the person who generated the G-code
wrote it for.

I can do that from keyboard, it seems more industrial in nature.

Yes, as a HSM guy, you don't need quite the same programmer vs. operator
setup.


12. six limit switches

got only five

That is sufficient generally. The control doesn't really have to know
which direction the overtravel was, as long as you can hold the
momentary limit override toggle switch and manually jog the axis back
off the limit. This is how most real machines operate. Many have three
limit LEDs next to the limit override toggle switch to show which axis
is on a limit as it can sometimes be difficult to tell if more than one
axis is at the end of travel.

But this means that, for homing with limit switches, theq limit
switches should not automatically inhibit drives via the inhibit
function, right? At least on my Y axis with only one limit switch?
And so, EMC should be responsible for proper behavior when limits are
reached. Right?

Yes, and no. You can actually have it both ways, since you're a
programmer. Your limits can connect directly to the servo inhibits and
you can have your "limit override" momentary toggle switch actually
disconnect and override the inhibit signals to the servos (use a 4 pole
switch). That way you still have true hard limits even if EMC goes nuts
for some reason, and the real ability to override the hard limits to
manually jog the axis back off the limits. The three LEDs limit LEDs
should be able to be directly driven from the inhibit signals to the
servos.

Forgot to add: If you need to move off limits under EMC control, as in
for a homing process, you can use an output from EMC to operate a relay
(4 pole) in series with the limit override toggle switch, to allow the
control to override limits during homing.

That's smart, looks like a way to go.

I think that hitting a limit, as such, is not a disaster if it is
handled right. So I would not want a limit to do a full estop.

I am really not sure what I want at this point, what is the ideal
estop behavior and what EMC can do with estop signals. Maybe I will
ask on the EMC forums.

i

Typically...there are two switches on each axis end. The first one is
for a warning/control change..the other is a home switch

When you run into the first switch..it signals the control that it
NEEDS to slow down and now..and after it does..it gently approaches the
home switch and stops where its supposed to, rather than slamming into
the hard stops as it overtravels beyond the home switch while
desperately trying to slow down.

I've not seen decel switches on both ends of an axis, perhaps that's
normal on some of the faster machines. The machines I've seen have a
decel switch only on the homing end of the axis, and the homes and
limits are both roller limit switches and trip rams that so that the
limit trip point is somewhat before the physical axis limit so the axis
can safely travel a little past the limit trip if it was moving fast. In
one of the previous threads I gave the detailed homing procedure using
the encoder index and why it's used.

On Sun, 27 Jun 2010 20:39:36 -0500, "Pete C."
wrote:


Typically...there are two switches on each axis end. The first one is
for a warning/control change..the other is a home switch

When you run into the first switch..it signals the control that it
NEEDS to slow down and now..and after it does..it gently approaches the
home switch and stops where its supposed to, rather than slamming into
the hard stops as it overtravels beyond the home switch while
desperately trying to slow down.

I've not seen decel switches on both ends of an axis, perhaps that's
normal on some of the faster machines. The machines I've seen have a
decel switch only on the homing end of the axis, and the homes and
limits are both roller limit switches and trip rams that so that the
limit trip point is somewhat before the physical axis limit so the axis
can safely travel a little past the limit trip if it was moving fast. In
one of the previous threads I gave the detailed homing procedure using
the encoder index and why it's used.

Did I say both ends...hum...I think I actually did, didnt I?

You are correct, two switches on one end of each axis.

Now some machines, such as the OmniTurn, have no switches and simply use
encoder counts to know where they are at..and when to slow down.

But it really isnt all that fast, and the slides arent all that heavy.

Gunner

One could not be a successful Leftwinger without realizing that,
in contrast to the popular conception supported by newspapers
and mothers of Leftwingers, a goodly number of Leftwingers are
not only narrow-minded and dull, but also just stupid.
Gunner Asch

I got an authoritative answer from EMC guys:

1) EMC can use limit switches when homing.

2) When not homing (ie when it knows its position and limits), EMC will never
permit any command to overrun limit switches.

3) When not homing, EMC will treat any limit switch as an alarm signal
and will stop.

4) EMC can handle both NO or NC limit switches. Mine are NC and no
problem for EMC.

I will wire my stuff to use limit switch signals through software and
will double test the operation of same.

i

On 2010-06-27, Joseph Gwinn wrote:
In article ,
Ignoramus31989 wrote:

Just wanted to do a little sanity check.

The question is about wiring power for all motive power (XYZ and
spindle) on this Bridgeport CNC mill.

The mill will be wired for single phase, with spindle motor on a VFD.

My current plan is as follows.

1. The CNC PC would be fed with incoming power without ANY regard for
any switches. When I need to shut it down I would type "sudo shutdown". I
will use this PC for many other shop purposes.

2. Have a switch (original switch on the door) be the first control of
incoming power to the mill (besides the PC, see 1)

3. Have all incoming power go through two fuses.

4. Jon's PPMC control box will be powered by 220v (more like 243 VAC
in reality). So any one fuse that blows, would cut the power to PPMC
control box.

5. Servo power supply, as well as power to the spindle VFD, will be
provided through a contactor.

6. Regular start and stop buttons would simply send signals to
software via PPMC.

7. The motive power contactor is actuated by a solid state relay built
into PPMC, but its control line will also go through a relay
controlled by estops, so that pressing ESTOP automatically cuts all
power to any moving equipment. I may need an extra relay in the estop
circuit, so that ESTOP not only cuts off motive power, but also sends
a signal to the control box and thus to the software that estop
occurred.

What I like about this approach is that it is relatively idiot proof
(estop means stop regardless of any software mistakes), and yet by
preserving power going into PPMC, I can keep track of my position
etc. It also keeps the PC going.

The minus is that in case of estop, the motors will coast to stop
instead of braking. But it is more simple.

Does this make any sense?

In addition to all the good points made by others, I would add one thing:
Ensure that if prime power is lost and later regained, all motion will stop and
stay stopped until you give permission.

I am pretty sure that if power is lost, EMC will lose contact with the
PPMC controller and will go into some sort of emergency mode.

I have l1 and l2 go through fuses and will make sure that ppmc runs
from both legs of incoming power, not just form one leg.

I would also go through your design and ask the question for each and every
component: what happens if this fails (stuck on, stuck off, or stuck open)? I
particularly liked the part about busted encoders breaking machines as the
servos manfully tried to achieve the unachievable.

I agree totally.

i

"Ignoramus28517" wrote in message
...
On 2010-06-27, Joseph Gwinn wrote:
In article ,
Ignoramus31989 wrote:

Just wanted to do a little sanity check.

The question is about wiring power for all motive power (XYZ and
spindle) on this Bridgeport CNC mill.

The mill will be wired for single phase, with spindle motor on a VFD.

My current plan is as follows.

1. The CNC PC would be fed with incoming power without ANY regard for
any switches. When I need to shut it down I would type "sudo shutdown".
I
will use this PC for many other shop purposes.

2. Have a switch (original switch on the door) be the first control of
incoming power to the mill (besides the PC, see 1)

3. Have all incoming power go through two fuses.

4. Jon's PPMC control box will be powered by 220v (more like 243 VAC
in reality). So any one fuse that blows, would cut the power to PPMC
control box.

5. Servo power supply, as well as power to the spindle VFD, will be
provided through a contactor.

6. Regular start and stop buttons would simply send signals to
software via PPMC.

7. The motive power contactor is actuated by a solid state relay built
into PPMC, but its control line will also go through a relay
controlled by estops, so that pressing ESTOP automatically cuts all
power to any moving equipment. I may need an extra relay in the estop
circuit, so that ESTOP not only cuts off motive power, but also sends
a signal to the control box and thus to the software that estop
occurred.

What I like about this approach is that it is relatively idiot proof
(estop means stop regardless of any software mistakes), and yet by
preserving power going into PPMC, I can keep track of my position
etc. It also keeps the PC going.

The minus is that in case of estop, the motors will coast to stop
instead of braking. But it is more simple.

Does this make any sense?

In addition to all the good points made by others, I would add one thing:
Ensure that if prime power is lost and later regained, all motion will
stop and
stay stopped until you give permission.

I am pretty sure that if power is lost, EMC will lose contact with the
PPMC controller and will go into some sort of emergency mode.

I have l1 and l2 go through fuses and will make sure that ppmc runs
from both legs of incoming power, not just form one leg.

I would also go through your design and ask the question for each and
every
component: what happens if this fails (stuck on, stuck off, or stuck
open)? I
particularly liked the part about busted encoders breaking machines as
the
servos manfully tried to achieve the unachievable.

I agree totally.

i

You can separate the power to keep power on the controls and encoders, but
kill power to the drives and things that cause motion and can cause injury
to people or equipment. As long as you keep the controls and encoders
powered up, you will keep the positions and shouldn't have to re-home,
unless there is an error. If there is an error in the encoder feedback
circuits it would be difficult to tell if you lost position information
because you had to hit an e-stop. I try to give power to controls and
feedback or input circuitry but kill power to outputs, motors, valves,
contactors, etc. If you had a robotic airplane, if things went wrong, would
you be better off to turn the robot off (or turn the flight data recorder
off) or leave it on so you can later see what went wrong. For the morons
that made up the "Kill all power" rules, does an aircraft kill power to the
flight data recorder, ie the black box? If not, why not, and why should
power be killed to a machines recorder (controls) if it is only recording
data, ie inputs, no outputs? Or to put it another way, I agree with "hands
off" but I disagree with brain off.

RogerN

On 2010-06-28, RogerN wrote:

You can separate the power to keep power on the controls and encoders, but
kill power to the drives and things that cause motion and can cause injury
to people or equipment. As long as you keep the controls and encoders
powered up, you will keep the positions and shouldn't have to re-home,
unless there is an error. If there is an error in the encoder feedback
circuits it would be difficult to tell if you lost position information
because you had to hit an e-stop. I try to give power to controls and
feedback or input circuitry but kill power to outputs, motors, valves,
contactors, etc. If you had a robotic airplane, if things went wrong, would
you be better off to turn the robot off (or turn the flight data recorder
off) or leave it on so you can later see what went wrong. For the morons
that made up the "Kill all power" rules, does an aircraft kill power to the
flight data recorder, ie the black box? If not, why not, and why should
power be killed to a machines recorder (controls) if it is only recording
data, ie inputs, no outputs? Or to put it another way, I agree with "hands
off" but I disagree with brain off.


estop will kill power to servo drives and will stop the vfd.

Computer and PPMC will still be powered.

i

Ignoramus31989 wrote:
The minus is that in case of estop, the motors will coast to stop
instead of braking. But it is more simple.

You have to try a simulated E-stop while in motion to see how far it
moves before stopping.
I have a messy all-relay scheme in my Bridgeport to apply a braking
resistor on E-stop.
It charges the caps in the servo amps with the braking resistor when
coming on, then connects with a contactor and sends enable to the amps.
It cuts power and applies the braking resistor as fast as the relays can
switch when going to E-stop. There is a time delay relay between the
two steps when coming on.

I have a newer circuit I call the "power switch and braking module" but
I'm not sure the braking resistor is suitable for a Series II size machine.
This does it electronically.

Jon

Ignoramus31989 wrote:
PPMC has 16 inputs. Of them, one is a E-Stop. So I have 15 inputs
left.


No, 16 totally free inputs PLUS the dedicated E-stop. Total of 17 inputs.
The first candidates for inputs are limit switches (6), Start/Stop
(2). This leaves 7 more digital inputs. Ideally, I would like to make
my X-Y-Z hand controls to work. For that, I would need at least 6
inputs.

I hope that if I need more buttons to work, I could expand this PPMC
by adding one more I/O cards.

Yes, another DIO card adds 16 more inputs (the E-stop is used only on
the master DIO board)
and 8 more outputs.

The PyVCP (Python Virtual Control Panel) lets you customize add-ons to
the main Axis GUI screen.
I've used it only to add a spindle speed display, but I'm pretty sure
you can add start and stop buttons, and then link external switches to
those functions. There's a whole manual section for PyVCP.

For a pendant control see http://pico-systems.com/pendant.html
I have two versions of this, one on the Bridgeport and one on the
minimill. My second one is better, you can hold down the activate
switch and dial the encoder with one hand. I have an axis select
switch, a rate selector and a button that enables the dial to move the
selected axis. If you aren't holding the button, the jog dial has no
effect. I think that is important, I sometimes bump the dial
accidentally while working. Another button trips the E-stop.

Jon

Ignoramus31989 wrote:

Maybe EMC has some kind of logic that if there is no motion detected
when it should be detected, it would do some kind of software shutdown?

If it is commanding motion, and no movement is detected, then it will
get a following error, after some time (usually short).
However, if the encoder were to fail while EMC was NOT expecting
movement, then it would NOT get a following error.
That is kind of a problem. A velocity servo would then move at a
probably modest rate as long as no motion was commanded.
This could be so slow you wouldn't even detect it!

Jon

On 2010-06-29, Jon Elson wrote:
Ignoramus31989 wrote:
The minus is that in case of estop, the motors will coast to stop
instead of braking. But it is more simple.

You have to try a simulated E-stop while in motion to see how far it
moves before stopping.
I have a messy all-relay scheme in my Bridgeport to apply a braking
resistor on E-stop.
It charges the caps in the servo amps with the braking resistor when
coming on, then connects with a contactor and sends enable to the amps.
It cuts power and applies the braking resistor as fast as the relays can
switch when going to E-stop. There is a time delay relay between the
two steps when coming on.

I have a newer circuit I call the "power switch and braking module" but
I'm not sure the braking resistor is suitable for a Series II size machine.
This does it electronically.

At 30v, the table that is moving in X direction, stops visibly
instantaneously.

i

On 2010-06-29, Jon Elson wrote:
Ignoramus31989 wrote:
PPMC has 16 inputs. Of them, one is a E-Stop. So I have 15 inputs
left.


No, 16 totally free inputs PLUS the dedicated E-stop. Total of 17 inputs.
The first candidates for inputs are limit switches (6), Start/Stop
(2). This leaves 7 more digital inputs. Ideally, I would like to make
my X-Y-Z hand controls to work. For that, I would need at least 6
inputs.

I hope that if I need more buttons to work, I could expand this PPMC
by adding one more I/O cards.

Yes, another DIO card adds 16 more inputs (the E-stop is used only on
the master DIO board)
and 8 more outputs.

Great.

The PyVCP (Python Virtual Control Panel) lets you customize add-ons to
the main Axis GUI screen.
I've used it only to add a spindle speed display, but I'm pretty sure
you can add start and stop buttons, and then link external switches to
those functions. There's a whole manual section for PyVCP.

Yes, I was reading about it last night.

For a pendant control see http://pico-systems.com/pendant.html
I have two versions of this, one on the Bridgeport and one on the
minimill. My second one is better, you can hold down the activate
switch and dial the encoder with one hand. I have an axis select
switch, a rate selector and a button that enables the dial to move the
selected axis. If you aren't holding the button, the jog dial has no
effect. I think that is important, I sometimes bump the dial
accidentally while working. Another button trips the E-stop.

Looks nice. I have some buttons on the machine for X, Y and Z
movements, I will try to integrate those with EMC. John did that:

http://machineability.com/Bridgeport_series_II.html

i