[SCzEngrgGroup]

First, this new forum software SUUUUUCKS! I made this post earlier, and it just disappeared into the ether - not an uncommon occurrence since the "upgrade". And the forums hang frequently, and go completely off-line pretty much daily since the change. NOT an upgrade from this users POV. I sure hope they get it sorted out soon.

Anyway, my new mill (Novakon Torus Pro) uses inductive proximity sensors for homing/limits. These are great - they're cheap, and can be easily used on pretty much any machine. They're really good for benchtop machines, because some of them are quite small, and very easy to mount. But, I discovered one "gotcha" - they can be triggered by aluminum chips! On my machine, the Y axis sensors are exposed at some table positions, allowing chips to spill off the table, and land on the sensor, causing false limit triggers. I didn't expect this, but it makes perfect sense once you understand how the sensors work.

Today I took about an hour and a half and did something about the problem. The range of a prox sensor is about 40% less for aluminum than it is for steel or iron, so it seemed reasonable to me that if I could prevent the chips from getting too close to the sensor, it would prevent false trips, but still work properly with the steel limit "flags". A little testing determined that about 0.100" space worked well. So, I machined a couple of covers out of some scrap UHMW, and screwed them over the sensors. Works perfectly! I can pile 6061 chips 1" deep on top of the cover, with no false trips, but when the limit flag approaches, it triggers perfectly.

These covers were only needed on the two Y axis sensors, as the X axis sensor is always covered, and the Z axis sensor is out of range of any chips.

The bottom side:
Attachment 181519

The top side:
Attachment 181520

And installed over the sensor:
Attachment 181521

Regards,
Ray L.

[DonKes]

A couple questions.

What is the repeatable?

Do they reach the same limit from both directions?

Thanks
Don

[SCzEngrgGroup]

A couple questions.

What is the repeatable?

Do they reach the same limit from both directions?

Thanks
Don

I haven't measured the repeatability of these specific sensors, but I've tested several different ones, and they were all on the order of a few tenths of a thou. It depends a WHOLE lot on how you trigger them.

I don't understand the second question....

Regards,
Ray L.

[mduckett]

Ray, thanks for posting this info. Do you have any favorite prox switches, and links for suppliers or manufacturers? I'm wondering if these are passive or powered, and any special signal conditioning for input to the BOB?

Thanks.
-md

[SCzEngrgGroup]

No real "favorites" - they all seem about the same to me, except for interface type and operating mode. All are active devices with three wires - power (usually 3 to 35V, more or less, ground, and output). They come in normally open, normally closed, NPN and PNP, so you have to understand the differences, and use the right kind, depending on your application. There are gazillions of them on E-Bay for as little as about $4 each, in a variety of form factors, sensing distances, etc.

Regards,
Ray L.

[JRouche]

First, this new forum software SUUUUUCKS! I made this post earlier, and it just disappeared into the ether - not an uncommon occurrence since the "upgrade". And the forums hang frequently, and go completely off-line pretty much daily since the change. NOT an upgrade from this users POV. I sure hope they get it sorted out soon.

Hi Ray.. Your post does not belong in this sub-forum. This sub is for Benchtop Machines and the Torus Pro is NOT a benchtop.. Please use the correct sub-forum. Thanks, JR

[SCzEngrgGroup]

Hi Ray.. Your post does not belong in this sub-forum. This sub is for Benchtop Machines and the Torus Pro is NOT a benchtop.. Please use the correct sub-forum. Thanks, JR

I posted here, as the post indicates, because these things are applicable to all kinds of machines, and I thought it would be helpful for people to know about this issue, to save someone else the headaches I had. This forum is about the most active one on CNCZone, so gets the most visibility.

Regards,
Ray L.

[JRouche]

I posted here, as the post indicates, because these things are applicable to all kinds of machines, and I thought it would be helpful for people to know about this issue, to save someone else the headaches I had. This forum is about the most active one on CNCZone, so gets the most visibility.

Regards,
Ray L.

I get what you are saying. But I think that's why your post disappeared (not a benchtop machine). Not due to the forum software. Oh, FYI. I didn't vaporize the post

Umm? Nice machine by the way. And if it was the Novakon Torus that looks to be a benchtop machine. Anyway. Just something to think about. JR

[Al_The_Man]

I'm wondering if these are passive or powered, and any special signal conditioning for input to the BOB?

Thanks.
-md

There are also two kinds of powering for Prox switches, they come in 3 wire and 2 wire, the 3 wire are mainly intended for powering devices that require a relatively higher current, e.g. a relays direct etc up to ~200ma.
The 2 wire are usually intended for higher impedance inputs such as BOB's and PLC inputs, the advantage is you do not have to worry about PNP/NPN as they will sink or source, also a separate supply is not required as they are powered from the input itself.
One downside is they posses some leakage when off, so if inputs other than say BOB opto inputs or are TTL type they could trigger in the off condition.
Al.

[SCzEngrgGroup]

Umm? Nice machine by the way. And if it was the Novakon Torus that looks to be a benchtop machine. Anyway. Just something to think about. JR

Soooooo.... Exactly what is the definition of a "benchtop machine". Because I have yet to see anyone put an RF-45 on a benchtop -they always seem to be mounted on stands, very much like my machine. Weighs almost as much too. Then there are the big Weiss and Optima machines. For that matter, the G0704 even comes with its own stand, and I don't think I've ever seen one of those on a benchtop either. What is the dividing line? Size? Weight? Cost?

Regards,
Ray L.

[bjones]

if your machine has it's own forum, i.e Novakon then it doesn't belong in the benchtop forum period.

[JRouche]

Soooooo.... Exactly what is the definition of a "benchtop machine". Because I have yet to see anyone put an RF-45 on a benchtop -they always seem to be mounted on stands, very much like my machine. Weighs almost as much too. Then there are the big Weiss and Optima machines. For that matter, the G0704 even comes with its own stand, and I don't think I've ever seen one of those on a benchtop either. What is the dividing line? Size? Weight? Cost?

Regards,
Ray L.

Here is a rf-45 on MY bench



Haha.. No doubt Ray. I actually DO think the Novakon Torus is a bench top mill. It might even have roots that go back to the the Rong Fu 45 bench top mill from the looks of the diags I have seen.

And the torus pro is the same mill but the base is not an option. Its part of the entire package I think?. If it is available without the apparently required base then YES! She is a bench top mill. But I dont think novakon advertises it as a bench top mill. Just like tormach does not advertise their mill as one.

Novakon markets the Pro as a bed mill and I think thats where they want to be. It looks like a great mill and by all standards it should have a place here along with other small bed mills.

And it DOES. The Novakon CNC Zone forum >>Novakon Systems

I along with others like your support on the CNC Zone. Contributors of information like yourself are the bread and butter. The post re: proximity sensors is useful info not just on the BTM (benchtop machines) sub but prolly more important on the novakon sub due to the likeness of machines. And the novakon sub has a great amount of interest going on now, its a very current sub.

Am I being nit picky? A lil nutty? Pushy? LOL Solly.... NOT my means. Just trying to direct traffic is all. I wont delete, move or shut down a post unless its blatantly wrong in one form or another. I was just trying to redirect your efforts.

Um... Oh. And any opinions re: the site as a whole, like "First, this new forum software SUUUUUCKS!" will perk my ears. The benchtop machine sub (or any other subs that Im a part of) is not a place for that. There are avenues for that feedback and thats why it caught my attention. Other than that Im good to go JR

[SCzEngrgGroup]

If it is available without the apparently required base then YES! She is a bench top mill. But I don't think novakon advertises it as a bench top mill. Just like tormach does not advertise their mill as one.

Uh-oh! By that definition, the G0704 is not a benchtop mill - you can't buy it without the stand either. And, technically, most of the machines on the benchtop forum are bed mills, as that term only defines a machine for which the spindle moves up and down, and the table moves perpendicular to the spindle axis, as opposed to, for instance, a knee mill, where the spindle is essentially stationary, and the table moves in all three axes.

BTW - I did post basically the same prox switch information on my thread in the Novakon forum.

Regards,
Ray L.

[DonKes]

I haven't measured the repeatability of these specific sensors, but I've tested several different ones, and they were all on the order of a few tenths of a thou. It depends a WHOLE lot on how you trigger them.

I don't understand the second question....

Regards,
Ray L.

I was thinking of a HOME switch where you could approach it from either side.

Don

[SCzEngrgGroup]

I was thinking of a HOME switch where you could approach it from either side.

Don

Ah! The sensitivity is generally symmetrical about the center point of the sensing area, so they would work fine. That is exactly how the X and Z limits on both of my machines operate. On Y is was simply more convenient to use two separate sensors. It really just depends on the machine configuration, the form factor of the specific sensor you choose, and how you choose to make use of them. As long as the "flag" that triggers the sensor is larger than the sensing area they should work fine, regardless of the direction of approach.

Regards,
Ray L.

[109jb]

I tested my cheap chinese prox switches in the way shown in the attached sketch. Sorry I didn't get a picture. Doing it this way allows using one prox switch to detect both end of an axis' travel. The gap when the target was over the sensor was 1/16" and the sensor worked perfectly this way. I ran the homing routine about 100 times and the repeatability was within 0.0002". On my G0704 I couldn't figure a clean way to hook them up this way and at the same time protect them from chips, so I am currently slowly working on a setup that only detects one end of the travel of each axis. So I will only have home/limit on one end of travel and the other end will use sofware limits but that's OK for me.

[SCzEngrgGroup]

I tested my cheap chinese prox switches in the way shown in the attached sketch. Sorry I didn't get a picture. Doing it this way allows using one prox switch to detect both end of an axis' travel. The gap when the target was over the sensor was 1/16" and the sensor worked perfectly this way. I ran the homing routine about 100 times and the repeatability was within 0.0002". On my G0704 I couldn't figure a clean way to hook them up this way and at the same time protect them from chips, so I am currently slowly working on a setup that only detects one end of the travel of each axis. So I will only have home/limit on one end of travel and the other end will use sofware limits but that's OK for me.

If you use NPN normally-open switches, you can wire-OR the outputs, and put as many switches as you like on the same signal. My machine has 4 prox switches (one of X, one for Z and two for Y), all wired to the same BOB input, with a pullup resistor. The signal is normally high, and goes low when any limit is hit. And they serve as both home switches and limit switches, so I can home to any corner of the "work cube".

Regards,
Ray L.

[bjones]

Uh-oh! By that definition, the G0704 is not a benchtop mill - you can't buy it without the stand either. And, technically, most of the machines on the benchtop forum are bed mills, as that term only defines a machine for which the spindle moves up and down, and the table moves perpendicular to the spindle axis, as opposed to, for instance, a knee mill, where the spindle is essentially stationary, and the table moves in all three axes.

BTW - I did post basically the same prox switch information on my thread in the Novakon forum.

Regards,
Ray L.

do a google search for bedmill, you won't see g0704 or any of the other mills in the benchtop forum show up, you will see talon, trak, novakon, lagun, ihcnc etc, all very large travel 1000 lb plus machines. you can't be that obtuse to think the torus pro qualifies as a benchtop machine. you already posted the same in your novakon thread and that's where it should be period. your fix is specific to the switch used on your torus. you don't seem to comprehend the reason for all the different forums to keep the site organized or else you don't care.

[waltpermenter]

BTW - I did post basically the same prox switch information on my thread in the Novakon forum.

Regards,
Ray L.

READ YOUR FORUM RULES!

DOUBLE POSTING:
Please do not double post the same topic in multiple forum sections.

[TiagoSantos]

Can we drop the silly politics? The Ray vs. Tormach owners thing is getting really old.

This is pretty much general interest, it's not Novakon specific and it should be where it impacts the most people - the benchtop machines forum is it. I'm not sure what the interest of having so many sub-forums is. Does anyone really have time to trawl the dozens of sub-forums here to check if anything remotely interesting has been posted?