Proximity Vrs Mechanical Limit Switches

Settle an argument,

Application hoist lifting a heavy load + 300-500kgs

If the hoist over travels substantial damage could be done, not dangerous just more $$$ issue and a loss of production time

System needs to have home sensors top and bottom for programming these are installed as proximity switches.

After these switches there is nothing to stop the hoist from travelling into a mechanical stop.

the argument

one side recommends having additional mechanical limit switches installed inline with the hoists linear slide after the proxy's as he is off the opinion that proximities are not reliable enough to act as limit switches, they have a sensing range of around 2-3 mm and the mechanical switch should be there as back up only.

The other side recommends that it is ok as is and to just let the drive fault out by driving into the mechanical stop and letting the VSD drive fault out on overcurrent.


probably not hard to know which side of the argument I am on, Interested to hear you guys thoughts.

Two ways to go about this.

It is NOT okay to let the drive fault. Period.

You can use mechanical limit switches or a photoeye for overtravel, or the other way would be to use the limit switches as NC so that when they break the machine stops. This does not help if a switch fails closed, though, so you would want one on each side to minimize the chance of impact.

You could also use a pair of proximity sensors on each side of the same type that way you have 4 of the same type of part and avoid the need for yet another part that needs to be stocked for the machine.

IIRC, passenger and goods lifts have both a mechanical limit switch for overtravel, before the final mechanical stop.

Proximity switches are good for positional purposes, but some safety systems would rather rely on a mechanical arm on a limit switch.

Prox to the PLC, mechanical NC to the limit inputs on the servo drive.

taward31 said:

If the hoist over travels substantial damage could be done

Click to expand...

Will the damage be instant, or after some time ?
If not instant, rather than letting the drive fault, you could monitor the current as a "healthy" signal for continue travelling.
If the damage is instant, then definitely have an additional safety sensor, and best in a fail-safe configuration.
IMO a mechanical switch isnt "safer" than a proximity sensor.

We have a conveyor in a project that is finicky and can easily be damaged if not stopped instantly in case something gets jammed somewhere. There is speed sensor to react when the conveyor gets blocked, but it doesnt react fast enough. We found that mounting a prox on a tensioner is a god way of sensing that something is wrong. Has saved the conveyor at least once.

A fail safe prox probably has a a longer MTBF(dangerous) than a regular prox + regular limit switch.

Are you going to have a maintenance interval to test the mechanical limit switches are working?

Redundant sensors are only good if you regularly check the sensors are working. If the mechanical switch fails, you won't know until the prox fails.

I would rather have two proxes at the same height, and a discrepancy alarm. Of course though, there are common cause failures to consider..

Good point from AustralIan that the sensors should be self-checking.
I have done something like this:
2 sensors (A and B) that have to be activated when the carrier stops in the correct position. The tab that activates the sensors is only so wide that both sensors are activated when the carrier is in the right position plus some tolerance.
A activates first, B second.
If A is not activated, but B is activated, then there is an overtravel.
The sensors are automatically checked for every travel of the carrier.

I agree with the above, if the damage is substantial. Fit fail safe Prox sensors.

Stop by drive fault - just bad strategy from so many angles.

Limit switch vs prox is a wash. One of the main advantages of a prox over limit is more lifetime operations. These won't see enough action for that to matter.

As for the prox only seeing a few millimeters, that is only an issue of the mechanical mounting/design allows it to be an issue. Mount the sensor in such a way that the sensing surface can't be further than 2-3 mm away and it is not an issue.

Being overtravel switches, I would always have them NC (Prox or limit)

I had a very similar scenario just recently. What we ended up with was mechanical limit switches for the functional end stops (although prox's or anything else would have been fine), and retroreflective PE's for overtravel sensors. A limit switch can be bent out of position and not detect the object. A prox can be bent out of position and not see the object. If you bend the PE or it's reflector out of the way, it'll switch immediately. We used NC so that it would also fault if the cable was damaged or unplugged.

The only "safer" way to do it would be to use a coded magnet that would be actuated at all times, but would clear if the device overtravels. Possible, but much more difficult mechanically, which is why we settled on the PE.

Waiting for the drive to fault to detect an overtravel, on the other hand, is akin to parking your car using the method "just keep reversing until you hear the sound of breaking glass, then go forward an inch".

ASF said:

I had a very similar scenario just recently. What we ended up with was mechanical limit switches for the functional end stops (although prox's or anything else would have been fine), and retroreflective PE's for overtravel sensors. A limit switch can be bent out of position and not detect the object. A prox can be bent out of position and not see the object. If you bend the PE or it's reflector out of the way, it'll switch immediately. We used NC so that it would also fault if the cable was damaged or unplugged.

The only "safer" way to do it would be to use a coded magnet that would be actuated at all times, but would clear if the device overtravels. Possible, but much more difficult mechanically, which is why we settled on the PE.

Waiting for the drive to fault to detect an overtravel, on the other hand, is akin to parking your car using the method "just keep reversing until you hear the sound of breaking glass, then go forward an inch".

Click to expand...

Yes, also transmitter/receiver photo eyes are great for this. I used to do a lot with a VRT 3-story tall automated freezer system that used this for the elevator and shelve blockage/misalignment/etc.

We had 2 cargo elevators on the last ship I was on.

They had Mechanical Limit Switches top and bottom for over travel and prox switches for stopping position at each deck.

You certainly don't want to jam the hoist tight at the upper limit and pray the drive faults before your cable does.
You would not want to put that kind of stress on your cable and connections. That would not be smart with or without a load on the hoist.

The bottom end is no different, you need something to tell the hoist to stop if it over travels in this direction, before it hits the mechanical stop at that end also.
If you don't have this and it does over travel down, then you will enter a slack wire situation and possibly need to realign your cables on the drum before putting the elevator back into service.
(You do have some sort of slack wire feature that will shut the lift down if it enters a slack wire condition also, correct?)

BCS