New Machine Build- Generic Emergency Stop (E-Stop) Design Concepts?

[zipzit]

I'm kind of a CNC newby, working on the design of a CNC plasma cutter. Most of the hardware is figured out, prototype stuff works okay. I'm struggling with understanding what to do about Emergency Stop (E-Stop) design. In my case, I'm using closed loop feedback encoders with brushed motors (Rutex driver boards, LinuxCNC control software, PC with parallel port input/output).. but my questions are more of a general nature.

I've got a couple of ideas in my head, not quite sure what the best practices on good design are. In general, the E-stop shuts down things that move immediately, but keeps retains power to the control system (to allow troubleshooting, records of errors, etc..)

There are a number of ways to design this thing. I see five different ways to stop a rotating 'motor'.

1) Send a signal to the control system, tell it to stop all motors now. Hope the motors stop when the control system tells them to do so.
2) Cut power to the DC voltage supply that feeds (thru the driver boards) to the motors.
3) Isolate individual motors (with a single pole relay per motor) between the driver board and the motor. Let the motor freewheel to a stop.
4) Swap individual motor voltage input with a circular loop.. In essence short out the motor, to kill inductive / inertial loads. This would require one Double Pole / Double Throw Relay per motor.
5) Swap individual motor voltage input/output with a high amperage resistor bank. This requires one Double Pole / Double Throw relay per motor. This would kill inductive / 'residual inertial' loads.

Obviously, these are ranked from easiest to implement to more difficult.

#1 only offers a limited safety net, but its easy to do. (Obviously this implies that a CNC control system includes an E-Stop routine... I'm guessing most of 'em do.) I'd do number 1 in every case as a redundant stop. I don't think I'd do this as a stand along option.
#2 requires only one additional relay (SP/ST).
#3 requires three SP relays on traditional X,Y,Z axis system
#4 requires three DP/DT relays
#5 requires three DP/DT relays, three resistor banks.

I haven't had a run-away condition yet, not sure that I really want to create one just to test the E-Stop system.

Any ideas here? What are best practices? Other comments?

Many thanks in advance,
Zip
Metro Detroit, Michigan

[Al_The_Man]

The industry standard to conform to the code etc, is to physically disconnect by contactor etc the AC supply side to drives, VFD, AC motors etc.
The controller itself, and if desired, any I/O of an input nature can be left energized to allow ease of trouble shooting.
The control or E-stop relay that achieves this can also have a contact advising the CNC controller a E-stop has occurred.
It is also customary for a watch dog timer or charge pump to detect if the controller (PC) malfunctions, the output of which is usually placed in the E-stop relay series string.
Al.

[mfgbydesign]

Sometimes overlooked is that an E-stop circuit needs to be wired in series and as a Normally Closed circuit. Essentially any switch thrown or open circuit in the wiring throws the machine into E-stop. You don't want faulty wiring or switch to show up only when you hit the button and nothing happens.

[zipzit]

Originally Posted by Al_The_Man The industry standard to conform to the code etc, is to physically disconnect by contactor etc the AC supply side to drives, VFD, AC motors etc. The controller itself, and if desired, any I/O of an input nature can be left energized to allow ease of trouble shooting. The control or E-stop relay that achieves this can also have a contact advising the CNC controller a E-stop has occurred. Al.

Al, wow, quick response. Okay, You are saying my selection #2 (with the substitution of the words 'contactor' for my word 'relay') are all that is required by industry standard code.

What about residual / inertial effects on a run away motor after the power is cut? Is there a need to 'shut down that motion NOW' to protect people, the machinery or the work piece? This would be beyond the industry standard of cutting power to the actuator's power supply.. Are my concerns much ado about nothing?

thank you, thank you for sharing your experience. Oh, and apologies for posting in the wrong forum. thanks for the help there.
zip.

[Al_The_Man]

If going by the code, it generally states that the function of an E-stop is not only to protect personnel but avoid machinery destruction also.
I think what you are talking about is when the cutting of power to a piece of machinery prevents a rapid controlled stop in the interests of safety.
An example of this is a large spindle that if power is cut can take an extremely long time to wind down, the code in this case allows for 'controlled stop' IOW a normal fast controlled stop could take place BEFORE power is removed.
There are specialized E-stop relays that do this called Safety Relays.
In the case of E-stops where say an axis with large inertia may crash into a physical stop even though a limit has been reached, one answer is to fit a cushioning device that prevents damage to the machine.
For a one off machine it may not be seen as necessary, but for anyone contemplating building more than one machine, the ANSI/NFPA79 Machine Control Wiring Regs are useful.
Al.

[wildwestpat]

Hi Zip

You have obviously though hard about what happens when in the event of an accident. As there are health and safety rules that vary with country, location and the type of premisses as well as type of use to comply with. Making contact with your local health and safety people should be the next step. Industrial users will have regular contact. DIY users should access the local regulation from the website of the local health and safety authority. These people are there to help prevent accidents and most of them will happily advise – unfortunately you then have to comply – so the web search will test the water so to speak!

There can be confusion between bringing a machine to rest so that it can be restarted and the functionality required when injury is being inflicted by the machine on either its operator or anyone else.

Sorry to sound so pedantic but it is necessary to have a clear view of the difference as the requirement to halt a machine quickly and the need to stop it killing someone can be different. Machines that use compressed gases or fluids or where there is a lot of stored energy capable of causing further injury once the emergency switch has been operated require further protective gear often operated by removal of the prime eletrical supply. (poder clutches - dump valves etc)

The wall mounted switches at regular intervals are a good starting point with these arranged to remove the prime electricity supply to the workshop. Individual machines need to have ON/OFF switches under local control and these must be reachable by the operator under all conditions. Large machines some times have a rope that surrounds the machine that when pulled activates the switch. A large table plasma cutter might require several OFF switches as it might be approachable from all sides as well as arrangements for fire protection etc. Also what happens if the operator is over come by fumes perhapse by loading the wrong sheet of material.

However in the event of an accident anyone should be able to stop that machine from a distance with the safety of people being the only concern.

Above all stay safe - Regards - Pat

[zipzit]

Ouch. funny.. I believe I'm thinking "big picture" when actually I'm acutely aware of the details in my own case. (Is that think globally, act locally?)

I'm well aware of industrial machine design and lock out standards for energy disippation. The particular machine I'm working on is more of a garage / very small shop machine, with a very limited number of users. Of course, machines like this do get sold, transfered to others, etc, so ensuring safety to standards is responsible behavior.

the machine I'm working on is 4' x 4' top, with the working surface approx 3' off the ground. The control panel is a seperate cabinet, off to the side of the work surface. With the design of this machine, generally you load work material from the front, and sit at the console to perform a run. There currently is one E-stop button at the upper right hand corner of the table in a very convenient location to both the load area and the control console. (I guess it would be very easy to add more buttons.. I can wire them in series.)


I think I'm okay on meeting the minimums.. E-Stop switch is normally closed, devices it drives must be activiated before the machine can function. Key features (think hot plasma arc) are engaged by multiple control devices in series (so a shut off on any one, is a shut off everywhere..) We've got end of travel electrical stops in one direction, and the gear drives run harmlessly off the drive tracks in the other direction.etc... We've added a light to the E-Stop that indicates the button condition. (You've got to spin and pull the button to reset it...) I really think I'm okay on the safety minimums...

I still have a question about stopping rotating mass AFTER the driving power is removed. Is it worth attempting to do an electric circuit 'brake' to further slow down a motor shaft in power off, spin down mode?

Let me ask a few questions about 'braking' actions in a slightly different way:
Has anybody ever seen a runaway CNC device? (something that the movement of the machine wasn't doing what was expected?)
Were you able to react in time to hit a E-Stop?
How long after you hit the E-Stop did the machine finally stop moving / spinning/ etc...?

thanks again to all for your reply's and thought starters..
zip.

[Al_The_Man]

Originally Posted by zipzit I still have a question about stopping rotating mass AFTER the driving power is removed. Is it worth attempting to do an electric circuit 'brake' to further slow down a motor shaft in power off, spin down mode?

A brake is a valid option, my Excello mill has a spindle brake, I connected it so that whenever the VFD shows zero speed after an M5 it is on, or during an E-stop.
The question of steps to take for safety reasons is at the discretion of the MTB or retro-fitter and will depend on how much the process has been analyzed in order to operate safely?
IOW, it is up to you to assess you particular situation and come up with a suitable method accordingly.

Originally Posted by zipzit Were you able to react in time to hit a E-Stop?

Murphy's law, the E-Stop is the hardest button to find on a operator panel.
Al.

[wildwestpat]

Good point Al about the 'E-stop' being the hardest one to find on a control panel. That is why the button of last resort should be with in reach at all times from anywhere in the workshop. This probably means multiple wall mounted 'Red Emergency' stop buttons as a minimum. At home I have five of the switches fixed to the wall. The control panel and the hand controller have their own as they are near to hand when setting up a run.

Yes removing the mains supply may also require other machine specific features to bring the beast to rest quickly and safely. There is also an associated design point concerning glitches in the electricity supply particularly for the supply to the computer.

Regards - pat

i