My Limit/Estop Strategy - Safety

[1875]

Will this diagram work and provide a safe limit shutdown?

Limit switches are NC. The VFD switch may be an opto-isolated switch that engages free run stop. The idea is that the circuit goes high when any of the limits are overrun as the ground is open. This sends a high to the drivers disable and BoB limit inputs. The momentary limit override switch allows the drivers to engage and drive off the limit, and will illuminate when the circuit goes high. The BoB is a CNC4PC C50; the VFD is an Hitachi WJ200; drivers are Gecko G203V. I was going to use separate home switches in series on a different input pin.

An alternative (addition) to the diagram is to wire the Estop in series and use the dedicated E-stop input pin. I think I like this idea better, but I'm not sure. I would have to hold the limit override in that case to reset/move the machine. Also, my Estop switch is not latching, so I would need to buy a new one or build a latching circuit.

Note that I may do something different with the VFD as I have not yet figured out how I want to control the spindle (Modbus or analog PWM, or even manual otherwise).

Thanks.

Similar Threads:

• Safety relays or just cutting power directly Estop
• limit/homing & estop wiring help
• Need Help!- Limit Switches/Estop
• Limit switch strategy
• Limit/Home Switch Strategy - Comments?

[1875]

OK, I think I need to move the 4k7 resistor to after the VFD, and I'm planning to implement this with the estop circuit...OK?

[1875]

That last thought was just silly - won't work that way... I should be able to use internal pull-up on the BoB and not need the 4k7 resistor to the 5v supply.

[A_Camera]

Will this diagram work and provide a safe limit shutdown?

No, it will not work at all, in fact it is very dangerous to have it connected like in your drawing.

Limit and eStop switches MUST be NC type of switches for safe operation. Limit override must be NO, as in your drawing. The reason for using normally closed switches is safety, and the possibility of serial connection of any number of switches. You can not connect normally open switches in series, as you have done in your drawing.

To check your drawing, just follow the route. What happens if you theoretically push the X limit switch? Nothing, because EN, Y and Z are still open. Your machine will not stop before all the switches are closed, and that will not likely to ever happen. If you would use normally closed switches then if you push ANY one of them, the circuit would be broken and the input would go high, triggering the BoB limit input. By pushing the normally open limit override, the circuit closes again and you could jog away from the tripped limit switch. If a cable is broken in the limit switch circuit, or a connection is lost (lose screw or soldering) then the same thing will happen, the limit input of your BoB will be tripped and your machine will stop. This is an error condition, you will notice it because you will see that the limit switches are not tripped, so something else must be wrong. If you get intermittent stops this is a good indicator of a bad connection which needs to be fixed. If you are using normally open switches you will NEVER be able to detect a faulty wire or connection until you run into the end of the axis and ram into one of the sides and destroy your CNC or cause some damage.

In principle, I have the same connection as you were thinking of, I have all my limit switches and eStop connected in series, one NO push button across the input used as limit override. I also have a real emergency switch, which is a latching type, that one is connected to the mains, breaking the mains power. Remember that the eStop, is NOT a real emergency stop, since it demands not only all the circuitry to work as expected, but also the computer to handle and process the input signal. A real emergency stop on the other hand is just a quick switch on the mains, it does not care about processing at all, you push it and all the power to WFD, steppers and control circuitry is broken, stopping everything. It is though a good idea not to have the computer connected to the same emergency breaker, it is better if you let the computer carry on and actually stop your software in a more controlled way to avoid disc crash or operating system failure.

[1875]

Hi A_Camera. Thanks for responding. My second sentence states the limit switches are NC. I just drew them open so they look more like a switch. The VFD FRS and Limit Override would be NO, and EN would be a single pole toggle.

What I want to happen is that 5V gets to the driver and VFD disable pins to stop the drives in case the computer gets locked in a loop and fails to respond to a limit condition.

One thing I do not understand about the Gecko documentation is that they state you can install a 7A fuse on the DC input, but installing a switch would likely destroy the drive.

[A_Camera]

Hi A_Camera. Thanks for responding. My second sentence states the limit switches are NC. I just drew them open so they look more like a switch. The VFD FRS and Limit Override would be NO, and EN would be a single pole toggle.

What I want to happen is that 5V gets to the driver and VFD disable pins to stop the drives in case the computer gets locked in a loop and fails to respond to a limit condition.

One thing I do not understand about the Gecko documentation is that they state you can install a 7A fuse on the DC input, but installing a switch would likely destroy the drive.

Hi,

I can't help you with the Gecko because I don't use that. It sounds though strange that they can be destroyed by a switch but not by a fuse. If you add a fuse (BTW, don't they already have a fuse on the motor DC input?) and the fuse is tripped due to too high current then that will break the DC current flow, which electrically is the same as if a switch was turned off. So, I have no idea why a switch could destroy it and a fuse could not.

Regarding the VFD controlling the BoB and the motors... I don't understand why you would need that. If for whatever reason your computer would lock up and fail to respond you must break the mains power for your CNC. Even if that may result in a blown driver, that is a serious safety issue, so a driver or two is a cheap price to pay. Though, I still don't understand why it would go up in smoke if you break the power, unless it is a cheap crappy design like the Chinese "blue cards". Perhaps if you would make a better drawing it would be easier to understand what you want to do, because like I said, that drawing has some very serious errors. I include a drawing of limit switch connections which works.



Start from there and add the modifications you want to add which explains what you wish to achieve.

[1875]

Hi,

...
Regarding the VFD controlling the BoB and the motors... I don't understand why you would need that. If for whatever reason your computer would lock up and fail to respond you must break the mains power for your CNC...

I include a drawing of limit switch connections which works.

Start from there and add the modifications you want to add which explains what you wish to achieve.

I do not want the VFD to control the BoB. I do want the BoB to control the VFD to turn the spindle on and off, and speed for normal operation.

[A_Camera]

I do not want the VFD to control the BoB. I do want the BoB to control the VFD to turn the spindle on and off, and speed for normal operation.

OK, now I get it. But...

I would not connect the way you want it because it demands common GND and +5V between the VFD and the BoB. I don't think that's a good idea at all.

Another thing is your motivation to do it... "in case the computer gets locked in a loop and fails to respond to a limit" If that happens then the limit switches have no effect at all because the computer reads the input and acts accordingly, if the limit switch is hit, the computer would sense it and would stop sending step pulses, as well as it would issue a spindle stop command to the spindle relay. Obviously, that demands the computer to work properly. I don't know how to make this work if the computer is out of control. Not that I understand why that would happen at all... Yes, computers may crash, but in that case it would also stop sending step pulses and as a result, the limit switch will never be hit. If on the other hand the computer is working as expected, you could use the spindle relay to stop the VFD. Isn't it designed that way? I mean, I don't have a VFD, but I assume every VFD has a hard stop input, similar to the limit switches, where you can connect the relay.

[1875]

The connections to the VFD and step drivers are opto-isolated.

I guess the designers of these devices have wasted their time creating disable inputs because we could just rely on the software to tell the BoB to stop sending motion signals. Or if not, we just rely on the software to tell the BoB to send a disable signal, which seems utterly redundant to simply stop sending motion signals.

It's all good because I have the ultimate fail safe as power comes from an overhead receptacle, so I can just yank the cord.

[UA_Iron]

The connections to the VFD and step drivers are opto-isolated.

I guess the designers of these devices have wasted their time creating disable inputs because we could just rely on the software to tell the BoB to stop sending motion signals. Or if not, we just rely on the software to tell the BoB to send a disable signal, which seems utterly redundant to simply stop sending motion signals.

It's all good because I have the ultimate fail safe as power comes from an overhead receptacle, so I can just yank the cord.

Multiple governing bodies for safety standards (CSA/UL/IEC etc) will not accept a software only solution as a means for safety of machine control devices that have potentially fatal consequences. They will accept it as part of a redundant system though.

There's a reason for this; software is notoriously unreliable, and if it is not unreliable, then it is super expensive, or it has a certain level of fault tolerances. Software expects faults, how it manages them is where the cost is also factored in. The path for a software control is usually in a series with other components critical for it's operation, like embedded chips and chip level logic. Say you have 20 components in series each with 98% operational reliability, then you will have a 67% system reliability. My kill switch will only work 7/10 times at best lol. This is an extremely simplified model, and electronic components are on the order of 99.99999% reliability anymore, but it's a valid argument that the more things in the path, the less opportunities I have for success. If you relied on the stop button on your GUI for mach3, how long would it take to take hold of the computer, move the cursor to the stop button and hit stop - or how many things would have broken in that time period (not saying this is your strategy, it just follows the logic).

You want a hard cut-of mushroom kill switch, one that is easily accessible. Your disconnect plug may work to the same degree.

[1875]

Sorry, I was being sarcastic with that comment. I was serious about the concept and proposed implementation of a hardware-only limit/e-stop solution. As long as there is 5V to the circuit, it seems like it should work. I just wanted to run it by the experts. I might be being penny wise and pound foolish by not implementing a more sophisticated contactor module - I'm just not seeing the need.

[UA_Iron]

Sorry, I was being sarcastic with that comment. I was serious about the concept and proposed implementation of a hardware-only limit/e-stop solution. As long as there is 5V to the circuit, it seems like it should work. I just wanted to run it by the experts. I might be being penny wise and pound foolish by not implementing a more sophisticated contactor module - I'm just not seeing the need.

Oh no need for apologies, I just like to explain something if I think I grasp the concept on 'the why' aspect!

It's always good to ask these questions, keeps everyone thinking and aware of concepts.

I'm not familiar with the C50 breakout board (nor can I find it on the cnc4pc site), but most all the boards clearly state how to wire NO or NC in a network (series or parallel) to achieve the limits. CNC4PC does make some unique solutions that might work for what you are doing:
C17 - MASTER CONTROL BOARD - Charge Pumps - Motion Control
C4- Safety Charge Pump - Charge Pumps - Motion Control

[1875]

For all I can tell, the C50 is the first release of the C10. They must have improved something though?

http://cnc4pc.com/Tech_Docs/C50R1_user_manual.pdf

Edit, or maybe it's newer? The file date on the C50 manual is newer (4/2012) than some of the earlier C10 boards. Odd that I cannot find a product page for it...?