Taking the CNC Plunge

[TomS]

After gathering some information from others I've decided to use proximity sensors for my limit switches. So I go to eBay thinking this should be easy. Was I mistaken! After perusing page after page of proximity sensors my brain is mush. Do I need inductive or capacitive, normally open or normally closed, NPN or PNP, shielded or unshielded, or DC or AC powered? I'm certain their are other considerations but I'll reserve those until I work my way through this first iteration.

What I think I need are inductive, normally closed, unshielded, DC sensors. What I don't know is do I need a NPN or PNP sensor? Your input is welcome and appreciated.

I knew the electronics part of this conversion was going to be difficult.

Thanks

Tom S

 

[JimDawson]

After gathering some information from others I've decided to use proximity sensors for my limit switches. So I go to eBay thinking this should be easy. Was I mistaken! After perusing page after page of proximity sensors my brain is mush. Do I need inductive or capacitive, normally open or normally closed, NPN or PNP, shielded or unshielded, or DC or AC powered? I'm certain their are other considerations but I'll reserve those until I work my way through this first iteration.

What I think I need are inductive, normally closed, unshielded, DC sensors. What I don't know is do I need a NPN or PNP sensor? Your input is welcome and appreciated.

I knew the electronics part of this conversion was going to be difficult.

Thanks

Tom S

The most important parameter is not listed above. You need a sensor that will operate at 5 volts, because that is your available supply voltage on the break out board. The lowest operating voltage that I have seen is 6 volts. Normally proximity sensors operate at 10 to 30 volts. And you want a shielded sensor so it doesn't pick up the tab from the side. Normally closed would be ideal, but they are not common. I think for a breakout board you want a NPN output, I'll check that tomorrow when I go out to the shop. I think you want to pull the output low on activate.

An option would be to operate the prox sensor at higher voltage and use a relay to switch the input on the breakout board. That would give you the equivalent of a NC prox. Google ''Phoenix Contact 2966171'' this relay is ideal for the application, and using a relay, you can use either a PNP or an NPN output.

 

[TomS]

The most important parameter is not listed above. You need a sensor that will operate at 5 volts, because that is your available supply voltage on the break out board. The lowest operating voltage that I have seen is 6 volts. Normally proximity sensors operate at 10 to 30 volts. And you want a shielded sensor so it doesn't pick up the tab from the side. Normally closed would be ideal, but they are not common. I think for a breakout board you want a NPN output, I'll check that tomorrow when I go out to the shop. I think you want to pull the output low on activate.

An option would be to operate the prox sensor at higher voltage and use a relay to switch the input on the breakout board. That would give you the equivalent of a NC prox. Google ''Phoenix Contact 2966171'' this relay is ideal for the application, and using a relay, you can use either a PNP or an NPN output.

Thanks for your comments. I thought proximity sensors was the simple solution. Maybe not. Now I'm thinking limit switches are the way to go. I'll reserve my decision until I hear back from you.

Tom S

 

[JimDawson]

Thanks for your comments. I thought proximity sensors was the simple solution. Maybe not. Now I'm thinking limit switches are the way to go. I'll reserve my decision until I hear back from you.

Tom S

I would use limit switches in this application. Prox sensors have their place, but I don't think your application is the proper use.

I am using limit switches like this one on my mill on the X and Y axis. http://www.automationdirect.com/adc...tal_Plunger_with_Roller_Actuator/AEM2G12X11-3
The major features of this switch is that it is slow acting, much more accurate and repeatable than a snap action switch. This makes it suitable for use as an accurate home switch. Most limit switches that you find on ebay are snap action because they are more common. It is also IP67 rated which means it's water tight to 1 meter submersion, pretty much coolant proof.

My Z axis is using a different switch, also slow acting. I needed a switch that I could bury in the head, so I went to Radio Shack and picked up a pack of these. http://www.radioshack.com/pushbutto...tary-switch4-pack/2751548.html#start=24&sz=12
This one would not work well if it got wet with coolant, and can't be used with a cam type actuator.

You only need 3 switches, and use the same switch for travel limits in each direction, and also they double as the home switches. You don't have enough inputs for separate switches for each end.

You are correct that you want to wire the switch as normally closed, that way the input is pretty much immune to stray electrical signals, and will fail safe in the case of a broken wire.

 

[TomS]

I would use limit switches in this application. Prox sensors have their place, but I don't think your application is the proper use.

I am using limit switches like this one on my mill on the X and Y axis. http://www.automationdirect.com/adc...tal_Plunger_with_Roller_Actuator/AEM2G12X11-3
The major features of this switch is that it is slow acting, much more accurate and repeatable than a snap action switch. This makes it suitable for use as an accurate home switch. Most limit switches that you find on ebay are snap action because they are more common. It is also IP67 rated which means it's water tight to 1 meter submersion, pretty much coolant proof.

My Z axis is using a different switch, also slow acting. I needed a switch that I could bury in the head, so I went to Radio Shack and picked up a pack of these. http://www.radioshack.com/pushbutto...tary-switch4-pack/2751548.html#start=24&sz=12
This one would not work well if it got wet with coolant, and can't be used with a cam type actuator.

You only need 3 switches, and use the same switch for travel limits in each direction, and also they double as the home switches. You don't have enough inputs for separate switches for each end.

You are correct that you want to wire the switch as normally closed, that way the input is pretty much immune to stray electrical signals, and will fail safe in the case of a broken wire.

Jim - Thanks for your feedback. When it comes to electronics I'm treading in new territory so your guidance and opinion are appreciated. I'll read up on the limit switches you suggested. If I have any questions I'll post them here so others can learn along with me.

Tom S

 

[Karl_T]

I also recommend a roller switch like Jim suggests:
http://www.automationdirect.com/adc...tal_Plunger_with_Roller_Actuator/AEM2G12X11-3

I've done many CNC conversions over the years. Being a cheap bastid, I've always just shopped for a large lot of switches on eBay. You can get them for 10 cents on the dollar. OK, my question for Jim, are all name brand made in USA type roller switches slow acting? Or should I be more careful in the future?

FWIW, eBay quality switches have worked well for me. But my applications don't rely on super accuracy for homing. I like them because they can be located nearly anywhere, then a bump put in place to make them actuate. Normally both limits and home switch all in the same spot for one axis. This allows only one seal tight cable back to the control.

Karl

 

[JimDawson]

OK, my question for Jim, are all name brand made in USA type roller switches slow acting? Or should I be more careful in the future?
Karl

Most limit switches are snap acting. You can tell by holding it up to your ear and listen for the click as you actuate it. If it doesn't click, then it is slow acting. Slow acting limit switches are not common, they are mostly used for precise positioning applications. Snap action are normally used in present / not present type of applications.

 

[jbolt]

Tom,

FYI my proximity switches are HTM: FCU1-0801N-ACS3. 5v-36v, NPN, NO, 1.5mm sense range. I got them through Charter Oak Automation. IIRC the kits were about $35 ea which included the sensor, 2m cable and bracket.

I spoke with HTM today and they say the repeat accuracy should be less than 1% of the max sense range or 0.0006". They did say that raw 4140 steel is used to establish the data and that the shape, size, type and coating of the trigger can effect the accuracy. Some of my triggers are 1/4-20 zinc plated flat head screws so I suspect this may be part of the accuracy issue I am having for homing.

Looking at the data sheet they sent it says the switching hysteresis is less that %15 of the sense range but I'm not sure how that may affect the repeat accuracy.

FWIW- I decided not to bury the switches in the machine because I tend to use all of the available table space and I like to see where the limits are while jogging.

J~

 

[TomS]

As promised here are a few work in progress pictures of my PM-932 CNC conversion. Just the mechanical stuff so far. I'm currently sanding and painting the column and Z axis components. It's been raining everyday for the past week and a half so it takes two days for the paint to dry before I can handle anything. Should have pictures to post of these parts in a couple of days. Electrical and electronics to follow once the machine is assembled and mechanically functional.

I had to re-paint the entire machine. Our favorite freight company didn't do a good job when handling my mill and it got beat up pretty bad. Lots of cosmetic damage but thankfully no mechanical damage.


Saddle installed.


X axis ball nut and screw installed. I want to keep the outside of the machine as uncluttered as possible. That is why I installed the X and Y axis lube fittings under the table. I'll follow the same concept with wiring.


I have 28" of X axis travel but will limit it to about 23" when I'm up and running.


The X axis stepper motor and mount. Thanks to jumps4 (Steve) for posting his motor mount drawings in his conversion thread. I borrowed his basic design.


I'm using BK and BF bearing housings on the X axis. The coupling is a double diaphragm style.


The Y axis motor and mount.


The Y axis uses a BK bearing housing and the same double diaphragm style coupling as the X axis.



So far this has been a fun project. The mechanical part is fairly straight forward you just need to take a lot of measurements. I'm trying to get my mind around the electrical and electronics. That's going to be a challenge.

Tom S

 

[TomS]

More pictures of the assembly process.

Column installed.


Z axis motor mount. I'm using a FK type bearing housing.


Another view of the Z axis motor mount. I had some 3/4" x 4" flat bar so that's what I used.


Here's a shot of the one-shot lube system pump and manifold.


The fittings on the bottom are feeding the X and Y dovetails and X axis ball nut. The top fittings will feed the Y and Z axis ball nuts and the Z axis dovetails.


Tomorrow the Z axis slide and head go on. More pictures to follow.

Tom S