[FordGalaxy]

It's time to put some limit/home sensors on my X2 conversion.

I would rather use proximity sensors than micro switches.

Anyone have some suggestions as to which ones work well, which ones to avoid, any help would appreciated thanks.

Bill

[Zaggy]

Hi,
I've tested Honeywell Hall SS443A associated with DealExtreme rare eath magnets (Super-Strong Rare-Earth Square RE Magnets (30-Pack) - Free Shipping - DealExtreme) with good success.
The working distance between the sensor and the magnet is around 5mm. Precision and repetability are satisfying : 1/100 mm.
Zaggy

[wizard]

I ask because for limits anyways a mechanical switch might be more reliable. The last thing you want is some swarf triggering an axis limit.

It's time to put some limit/home sensors on my X2 conversion.

I would rather use proximity sensors than micro switches.

Are you sure about that's?

Anyone have some suggestions as to which ones work well, which ones to avoid, any help would appreciated thanks.

Bill

I'm sure there are different opinions here but for limits I'd suggest mechanical switches.

[Flexo!]

Another vote for proximity switches.

ALL limit switches have to be covered... chips can get into mechanical switches, or optical switches so its not a valid argument.

Inductive switches are easier to mount - espcially the 2 hole ones like these: SNO-4 PNP Proximity Switch, they run off breakout boards 5v line, and give very good accuracy (better than a 10 microstep stepper can resolve ~0.003mm)

[RTTIM]

I ask because for limits anyways a mechanical switch might be more reliable. The last thing you want is some swarf triggering an axis limit.

Depends on what type of proximity switch you use - there are several options, including inductive, capacitance and magnetic/hall effect. The ones Zaggy mentioned, for example, are hall effect units that are triggered only by a specific pole of a magnet (usually S). They are very reliable and will not be triggered by swarf.

OTOH, mechanical limit switches can be less expensive (depending on type), and the wiring is certainly simpler.

Tim

[FordGalaxy]

The two hole inductance sensors seem the easiest to use. I'm wondering about how repeatable they are. Do they return to zero, or is there a plus or minus a few thou?

[FordGalaxy]

109jb
I thought someone might answer how repeatable the inductive proximity switches are, but RPC jacked the thread and took it to the basics.

[RTTIM]

109jb
I thought someone might answer how repeatable the inductive proximity switches are, but RPC jacked the thread and took it to the basics.

Too general a question - again it depends on type and packaging. Some are very precise, other less so.

Repeatability in application is more a function of the control software. Most CNC programs stop the axis drive motor when the home/limit switch actuates, then slowly reverse it until the switch just clears. If your system is calibrated correctly, repeatability should be just fine. For really precise applications you'll still want to zero to the part prior to machining.

[SCzEngrgGroup]

109jb
I thought someone might answer how repeatable the inductive proximity switches are, but RPC jacked the thread and took it to the basics.

You're kind of asking "What gas mileage does a car get?", or "How bright is a light bulb?". Is the car a Pinto or a Ferrari? Is the light bulb 15W or 100W? Different prox switches have different specs. You have to look at the specs for the specific switch you're interested in using.

Regards,
Ray L.

[Flexo!]

I thought someone might answer how repeatable the inductive proximity switches are

The cheap $5-$10 chinese proximity switches will give you better than 0.01mm. I've tested them pretty thoroughly on my 2 CNC mills and you get a bit of variation in multiple home attempts (likely from the CNC electronics/software etc).

[FordGalaxy]

Thanks flexo
For homing .01mm should be fine.
When I was a machinist we would home the machine and zero the registers when we powered up, and then set X,Y,Z zero on the part.

[wizard]

Ah! I didn't even think of doing it that way. You have to forgive me, I come from an electronics background where I am always thinking of tying every to an individual input (I guess it comes from my microprocessor background).

It really depends upon how fancy you want to get with your controls. On smaller machines though there is little value in wiring up axis limits to discreet inputs. So you can either electrically AND them or OR them together.

That being said if you want to directly disable your motor drives with the axis limits then you should pay attention to how the ENABLE works on your axis amps.

So, I guess I can just tie them all into my E-Stop circuit?

Thanks!

No! Repeat No. An E-Stop circuit is an entirely different animal from an Axis over travel limit. This is an area that is often not given the right level of importance on home built machines, but I might suggest that it is a critical area for any automatic machine. Thus I'd reccomend becoming familiar with proper E-Stop implementations. Oh and buy good quality E-Stop buttons.

On a side note people often use the axis over travel limits as an homing limits. I generally don't like this approach, better to implement entirely separate circuit(s) for that.

As to the discussion about prox switches I will just mention that I don't see many of them used on metal working equipment. For machines destined for other uses, yes you will see prox switches being used for limits. Maybe machine tool manufactures are old fashion, but I also see it as a sign of experience.

[wizard]

Thanks flexo
For homing .01mm should be fine.
When I was a machinist we would home the machine and zero the registers when we powered up, and then set X,Y,Z zero on the part.

There are many many different ways to home an axis on a machine. Generally on a machine with servos and encoders the "Home" switch is only a course reference. Once the control establishes the course position it finishes finding home by looking for the zero pulse from the encoder.

On stepper controlled machines the accuracy of the home switch can be more important. As you note though one often establishes zero someplace else.

If you look around a bit you can find all sorts of sensors, prox switches and the like designed specifically for homing switches. They may be of value to you, but off the self switches may work just find. As has been mentioned read the manufactures specs closely. Sensed materials, distances and other factors come into play here.

[SCzEngrgGroup]

On a side note people often use the axis over travel limits as an homing limits. I generally don't like this approach, better to implement entirely separate circuit(s) for that.

Because.....??

Regards,
Ray L.

[SCzEngrgGroup]

That being said if you want to directly disable your motor drives with the axis limits then you should pay attention to how the ENABLE works on your axis amps.

This would require a hardware over-ride switch to enable moving OFF the limits....

Regards,
Ray L.

[wizard]

Because.....??

Regards,
Ray L.

For one if you decide that you want to change the position of a hard limit all of your zeros then change. In industry we often set up hard limits to attempt to save expensive tooling from crashes or operator goofs. Yes I realize that most installations put the limits at the physical limits of travel and leave them there. It is just that this isn't always the case.

Another issue is this, if you use an axis limit for homing how do you finish the homing cycle after tripping the hard limit? Especially if you are disabling the amps as well as signaling the computer. If you want the hard limits to be useful, in all failure modes, it is best to have them disable the drives when hit.

Another issue is that sometimes you simply don't want zero where the hard limits are.

Depending upon how fancy you get, software can be written to distinguish the inputs and properly log debugging data.

Remember I indicated that the separate inputs for home versus hard limits is a suggest or preference of mine. There are many ways to zero an axis or to handle hard limits. I even spent some time on a machine that actually faulted the amp by bottoming out the axis to find the zero. It worked well enough for that application but it isn't something I'd reccomend.

[109jb]

For one if you decide that you want to change the position of a hard limit all of your zeros then change. In industry we often set up hard limits to attempt to save expensive tooling from crashes or operator goofs. Yes I realize that most installations put the limits at the physical limits of travel and leave them there. It is just that this isn't always the case.

First, I'll only talk about EMC because that is what I know about. In regard to your concern above, there is no need to physically move a switch in EMC. EMC has the ability to set soft limits. Esentially there is a large box representing the limits of the machine as defined by the switches. Within the software you can set up a smaller box that is defined by the soft(ware) limits. The EMC controller software will not drive an axis past these soft limits.

Another issue is this, if you use an axis limit for homing how do you finish the homing cycle after tripping the hard limit? Especially if you are disabling the amps as well as signaling the computer. If you want the hard limits to be useful, in all failure modes, it is best to have them disable the drives when hit.

EMC knows that the limit/home switch is combined and doesn't disable the drive when the switch is tripped ONLY for the homing sequence. At all other times, if the switch is tripped, the drives are disabled and you have to re-enable the motor drives within the software before you can move the machine

Another issue is that sometimes you simply don't want zero where the hard limits are.

If you are referring to the machine home position, again, in EMC you tell it where the home position is in relation to the switch. For example you can tell EMC that the home position is 0.1", 0.5", 0.163" from where the switch trips. Any number you choose. This is set up in the machine configuration.

Depending upon how fancy you get, software can be written to distinguish the inputs and properly log debugging data.

Remember I indicated that the separate inputs for home versus hard limits is a suggest or preference of mine. There are many ways to zero an axis or to handle hard limits. I even spent some time on a machine that actually faulted the amp by bottoming out the axis to find the zero. It worked well enough for that application but it isn't something I'd reccomend.

No doubt that having separate switches for home, +limit and - limit on each axis would be ideal, but the fact is that for those using a single parallel port for I/O only has 5 pins that can be used for inputs and combining switches preserves those inputs for other uses. On a typical 3axis machine with 3 switches per axis you would need 9 inputs if you want to hook them each up individually. I personally see no reason you need to do this. In the case of a limit switch being tripped, the machine is going to shut down the drives for all axes, so it doesn't really matter which limit is tripped. Wiring them together makes sense in order first have enough inputs on a single parallel port, and secondly to preserve input for other tasks such as spindle feedback, tool changer feedback, etc. Combining the limit switches and keeping the home switches separate still uses up 4 inputs. Combining the home switches would reduce that to 2 inputs used and you would just then have to home each axis one at a time, like Z first then X then Y. Once homed, who needs the home switches. They don't do anything during any operation except homing anyway.

I'll be using one input for each axis initially with a switched home/limit at one end and a soft limit at the other end of each axis.