Setting X and Y zero with Prox Sensors

[sagreen83]

Ok... Every day I find out something new that I didnt know the day before with the bot and its control software..

Today that enlightenment was the "Home X and Y axis's using the prox sensors". I found this after I spent the time to modify the ZZero program into something that I could use with the plate to zero the X and Y axis's. I dont know why I didnt think of this to start with (Somewhat Miffed that I didnt) but this seems like a GREAT way to do the X and Y zero.

So... If you dont know how to do it, or havent beore, hit the "Wrench" in the control software / ShopBot position screen. Then select the "Home X and Y axis's using the prox sensors" option. You will have to modify the file that is used to do this by setting the distance from the prox sensors to your X and Y zero locations. The file that you need to modify is c:\SBParts\home.sbp lines 59 and 62 respectively.

Now, all jubilation asside, there must be a catch. Did I just find something that everybody else alredy knew, or is there something wrong with using this function? Is there an accuracy problem / limitation with using the prox sensors to position your X and Y zeros?

Thanx,
Scott...

[srwtlc]

The key word here is "proximity". If it's not important how close you are to the exact corner or your part, then it's fine. If you need to be exact though, like when you want to place a cut of some sort at an exact place on something, then you'll need to do better. Once you home it, move back untill the prox switches stop it and check the values a bunch of times. I've noticed quite a variance from time to time on a tool that I set up for a nearby school recently. I have never had any prox switches for my machine. I use a zero X & Y routine with a corner plate/block that I place on my blank/part whenever I want. I have it ask what bit size I'm zeroing with at the time also.

[fleinbach]

I use the proximity switches for zeroing the X & Y axsis and find them to be apsolutly dead on acurate. I use 1/4 inch dowel pins to alighn parts that need to be cut on both sides. I have alighnment holes drilled into my table to insert the dowel pins at precise positions. For my porposes they are spaced at 1' intervals on the X & Y axsis. After alignment using the prox switches I can run the bit back down into these holes and it is so precise the 1/4" dowel pins still fit tight. This is acurate enough for my porposes.

[dingwall]

I added speed and ramp settings into our homing routine so that the tool is always travelling at the same slow speed when it hits the detectors.

[paco]

Sheldon
Is this getting you a very accurate position within MANY homing? Or you too wish for better? Do you use SB prox. or other?

[elcruisr]

We slowed down the speed settings for the homing routine and we hit dead on every time as well. Now that I've said that it probably will give me fits next time I run it...

Eric

[stevem]

I'm using micro switches on my Bot and home built router. The switches reliably repeat within .001". The small switches for my home built router cost $2.00 each and should be available at most electronic supplies stores.

[bill.young]

Hey guys,

The newest homing routine records your current speed settings, slows the tool down to 1"/sec. while the routine is running so that it doesn't approach the prox switches too quickly, and then resets the speeds to their original settings. I believe it's in the current Alpha software and will be in the next PRT version.

Bill

[paco]

What are thoses micro switches? Should I ask supplier for micro switches? I have here at home some "telephone switches" (switches to get the line; main telephone switches...)... is it something like that Steve? Where did you get thoses specs. about your switches?

[fleinbach]

Paco,

Just as Steve said you can pick up micro switches from almost any electronics store. Micro swithes have been around for a very long time. You can get them as either N/O N/C or both. They have been used for years to stop travel or signal position of many meachenical componants. They are used to trigger alarms. I have some old ones from the 1950's. I don't exactly know how much farther back there history goes. I'm not sure how they got the name micro because there is nothing micro about some I have seen some several inches accross. The one thing most all have is a spring steel trigger mechanisum. Some have roller bearings to prevent wear. The spring steel is nessacary to prevent the movung part that triggers it from damaging the fragil switch.

[paco]

That's what I thought... I've been told by electronic supplyer that would'nt suit my accurate need... But as I was in front of, I had this voice again in my head that gently said: "don't buy thoses 65.00$ca prox. switches... you don't need it..."

Do you get a satisfying and accurate positionning from thoses? At least as good as prox. switches?

[fleinbach]

Just remember the positioning of a micro switch is different then the prox switch shopbot uses.
There are 2 things to consider.

One the prox switch has no moving parts hence less maintenance from wear. For this reason the prox switch will far outlive the micro. But of course you could replace the micro dozens of times before reaching the cost of the prox.

Two the prox switch scenes a metallic mass when in very close proximity to it. This gives it the capability to pass over the target should the tool be moving to fast. You can achieve the same results with the micro switch if you position it carefully. But the activation of the micro switch is dependant upon the depression of the spring steel trigger when it is depressed by the target stop post. You also have to be careful in positioning the micro switch so that there is a ramping action of the spring trigger. Depending upon the switch this spring can have quite a bit of travel to activate. This is not a problem in obtaining an accurate stop but should the tool pass completely over the stop post in might get caught on the return trip destroying the switch. There are several ways to prevent this but these are the things you need to consider if using them.

As for the acuracy of the micro switch. They can be extremly acurate but this acuracy will be slightly affected by temperature changes.

I had considered using them myself but didn't have time to work out all the details so I just bought the prox switches and moved on.

[gerald_d]

If we are quite happy to use a router bit to hit a chunk of aluminium to precisely calibrate our z-axis, why do we look for something more complex for the x and y axes?

We don't have any sensors on the x and y, but if we find that we need them, our first attempt will probably be with "ground-wire-touching-plate" technology. There has been mention on this Forum of guys who have magnet-base plates that they can move around as they want to. This sounds very practical and convenient.

[fleinbach]

My opinion for the reason for two differant methods of zeroing

Simply because the Z axis varies according to the material thickness. Plus the reading you are looking for is an exact location, the bottom of the bit. This seems to me to be best achieved using the aluminum Z-Zeroing plate.

While the X/Y zero is typically an exact location on most tables. The use of the prox switches is better in this application. By just entering one simple command C3 they are zeroed without further thought. The prox switch being the easier method, since it is done by entering a simple command. Where as the zeroing plate needs to be placed into position then removed.

I was thinking about putting a prox switch on the Z axis to be used for zeroing. This would work but I will need to make several files to zero to the top of different material thicknesses. Or I could just use the table surface for all files and then will only need one Z-Zeroing file. I also read another post a week or so ago where they have a permanent Grounding block to run the bit down to. Of course this too would either necessitate either zeroing to the table surface or creating different files for different thickness materials.

[Ryan Patterson]

Frank
I use a permanent plate for my z. The zzero file will zero the bit to the top of the spoil board. If I need to have z zero to the top of my part I use the VA commad to adjust z zero.

[fleinbach]

Ryan

Thanks for the tip. That sounds like a great way to go. I just may give it a try

[paco]

VVEERRRYYYYY interesting Ryan! And VERY simple...

[sagreen83]

Ok, Frank and Ryan...

I agree on the prox sensors for the X/Y alignment, but The Z axis?

If you ZZero to the spoilboard, and your spoilboard expands because of moisture, or you have to re-surface it then the permanent Z plate idea would be inaccurate because of expansion or way off after you resurface. Ryan, do you compensate for the expansion?

Scott...

[cnc_works]

My question of those of you using the Z-Zero plate is to what accuracy you commonly hit using the plate. Within .002? More? Less?

I've seen it implied that it can vary according to the diameter of the bit or the amount of the bit that hits the plate. Would I get the same accuracy using a 1/2" mill as I would with a V bit?

My current method of setting zero is getting a slip fit with a piece of .001" brass shim stock which is quite accurate, but slow.

Any further input on easing the re-zeroing process after a bit change would be greatly appreciated.

Donn

[fleinbach]

I'm not very worried about spoil board expansion as most of the time I cut through my material and the rabbit depths have no discernable affect on parts fitting properly. I usually cut thru .01" this works fine for me. Actually after cutting several sheets of parts the raised area from the bit holds the next sheets by just about that amount so I can barley see the new cut thru marks.

On my old Prt I only cut a bit over 100 sheets and never had the need to resurface.

When I'm ready to do something that requires closer tolerances I will double check my zeroing to make sure it is accurate for that cut

[gerald_d]

Frank, you said yesterday: "... the X/Y zero is typically an exact location on most tables." This is true if guys are trying to clamp stock down onto the table in an exact location, and if one of those edges of that stock is not cut by the SB.

We normally work with sheet stock (MDF and plywood) and we never use the supplied edges as finished edges. We always run the cutter right around all our parts and therefore don't have to worry about positioning the sheets precisely, and don't have to zero the x and y axes precisely.

On the rare occasion when reference clamping positions are needed, we screw/clamp some scrap strips/blocks to the table which are faced with a cutter of known diameter and that gives our reference marks. If the motors are switched off overnight, we just make another facing cut on the reference strips/blocks the next day.

I cannot see where sensors on the rails are going to help us. Maybe they would have helped in the early days when we were still pushing the wrong buttons, but even then the problem was more with the z-axis rather than the x and y.

[fleinbach]

Gerald, I too cut mostly full sheets up to 4 X 8' to 5' X 12' and I usualy trim my edges. But a lot of my parts go from edge to edge. So I use a 1/4 bit to get maximum parts from a sheet. They are spaced very close together. The parts fit so close sometimes I am just taking 1/16 off the edge of the sheet.

I surfaced my table along the 0 axis both X and Y. I just align my sheets with these edges. I have used other methods to get this close to the edges and have considered going back to one of them. In fact it was simular to what you mention about surfacing scrape material as a guide. I may still go back to that method but will still use the prox switches for Zeroing the X/Y.

[Ryan Patterson]

Scott,
I resurface once or twice a week and I have to make an adjustment in the zzero file (plate thickness). I make the addjustment by 1/32" before surfacing then surface with z depth of zero(0). I have also added the zzero file to the homing file so when i type C3 the file will ask me if i want to zero the Z only or X,Y or X,Y,Z.

Gerald
The proxy switches are helpful when in the middle of a cut the power fails, the motors skipped steps or lost placment. With homing X,Y you will easily be able to start where you left off.

[gerald_d]

For power failures and lost steps, reference position sensors would certainly be useful. Thankfully, those incidents are extremely rare over here.

[paco]

Where have you installed your permanent Z-plate Ryan?

[Ryan Patterson]

Paco,
I Have it mounted to the right of table zero it is about 3/8" below spoil board. I jog x to 0 and y to -.875 this will put router over plate.

[bill.young]

What about inlaying some kind of soft metal like copper directly into your table top...or maybe just maybe melt a big glob of plumbers solder and pour it into an undercut recess...and use that as your z-zero contact. That way it would be flattened when you trued your table top, so it would always be the same level as the table. Whenever you wanted to zero your Z-axis you would just move to that spot on the table and run the routine.

There might be some expansion issues with the metal and some safety issues to think through for sure so that the metal plug couldn't come loose if hit by the bit, but if you could get it worked out it would sure be convenient.

Just a thought,
Bill

[bill.young]

How about pouring a recess full of some of the "liquid metal" stuff that looks like fine metal flakes in a resin matrix...is any of that electrically conductive? That would be a lot safer than my original idea of a soft metal insert in the top.

Bill

[beacon14]

Bill,

The problem with having the z-zero recessed into the table is what do you do when your part covers the recessed plate (full sheet of plywood, for instance), and you need to change bits or re-zero in the middle of a cutting sequence?

David B.

[bill.young]

Hey David,

That's a good point. I still like the idea of machining the contact surface when you flatten your table top, though, so that they're always in sync. I guess it needs some more thought....

Bill

[Ryan Patterson]

Bill,
The zero plate is in sync if you surface the table after making an adjustment in the zzero file. Then when you surface the table surface with a Z depth of zero. I do like your idea of useing plumbers solder. When I zero on my indexer the top of the lath is not flate. I can form a flat square and pour the plumbers solder in to get flat surface to zero to.

[richards]

It seems that measuring from the spoil board would be less accurate, in production, than measuring from the top of the material, at least when cutting grooves and dados in sheet goods because of the possibility of debris being caught between the spoil board and the material to be cut. I know that there shouldn't be any debris left on the spoil board, but tired eyes or lazy arms sometimes miss a few specks.

In addition, plywood is not always flat or even the same thickness. Although it is possible to measure the material's thickness at several points around the edge, how would you measure thickness plus a bow in the center of a 4 x 8 sheet?

Laying the Zero plate on top of the material only takes a few seconds, and has proven to be reliable, except with small bits. In that case, I raise the bit a little and check the actual gap with a feeler guage.

Mike

[bill.young]

Hey Mike,

That's a very good point but I use a shopvac vacuum holddown setup without a spoil board on top so I try REALLY hard to not cut below the surface of my table top and ALMOST always zero at the table surface. That way I know that if I keep all my z-axis moves as positive numbers that I don't have to worry about cutting into my table...if I want to leave .02" of material uncut I don't have to measure each sheet before I cut it.

The exceptions are exactly the sort of things you're talking about...things that have to be referenced from the face of the material like grooves and dados and v-carved lettering. However, if I want to cut a rabbet (sp?) in the edge of a piece of material so that the tab that's left is an exact thickness...say a .25" tongue...then zeroing at the table surface lets me do that without having to measure the material and calculate the depth of the rabbet. My z-axis value would be .25 (or whatever the thickness of the tongue I wanted) no matter how thick the material is.

Just one of my many quirks!
Bill

[paco]

Would you (all) say that zeroing from spoilboard surface better than zeroing from surface of material? I have some trouble getting to NOT cut too deep in my spoilboard... or to control depth of cut from material top...

[stevem]

Paco, NO.

Zeroing from the surface of the material is much safer and makes more sense when generating a cut file. You are always removing material, thus, any vertical cutting should be a NEGATIVE.

[bill.young]

Paco,

I'm in the minority for sure...by far most ShopBotters zero at the material surface. They both have their advantages and disadvantages, but for the way that I work and the setup that I have, zeroing at the table surface works best for me.

Bill

[Brady Watson]

Paco,
If you are cutting too deep into your spoilboard, then you should measure your material thickness with a micrometer.

The only times I zero off of the spoilboard is if I am surfacing the spoilboard itself, or need to dimension a rough piece of lumber to an exact thickness.

You might want to check your z-zero plate thickness with a digital micrometer and make sure it is the same as the value in the C2/cuscut2 file. Zero on the spoilboard surface, let the bit jog up to it's safe z-height and turn on the router. Tell the router to go down to 0 and see if it touches the spoilboard. If it does, you need to fine tune the z-zero values in the file. If it doesn't, then you need to more accurately enter your material thickness into PartWizard or your CAM program.

99% of the time, it is human error in entering the exact thickness of the material.

-Brady

[fleinbach]

Personaly even though both methods can work equily well as long as your mind set is well aware of where zero is. I believe it is safer to use the material surface if you are in the habit of using differant material thickness's. I have cut material up to 3 inches thick and I feel safer referancing Zero from where the bit is going to first contact the material.

[fleinbach]

I will add that I do have one very bad habit I wish I could get rid of. I always pull my tape measure from right to left so I am always reading it upside down.

[Brady Watson]

Frank,
I think they make a tape just for people like us:

I'm not left-handed...but I mark with my right.

http://thelefthand.com/lefhantapmea.html

-Brady

[paco]

Yes, I should get a more accurate reading of the material (using a caliper) than as for now (tape... and eye ball)... but I was wondering if using the zero to spoilboard way would prevent from cutting in spoilboard IF no M value (or J !!!)? Like the other days as I was cutting 1/8" Acrylic; I wish I could had a better control (without MANY try) to cut thru material BUT NOT thru "bottom" protective paper...

Steve, Brady, Frank, tell me more about the SAFE aspect please... What is less safe using zero to spoilboard?

And about the advantages vs disadvantages; using zero to material could be better to control depth of cut FROM the top of material... but you need the use of a zero-plate (EACH TIME) which is the way I use for now... but I'm getting bored of it. One the other hand, using a permanent zero-plate with routine thus refering to spoilboard surface could (in my mind...) prevent cutting in spoilboard unless plan to... and the zero plate installation could make Z calibration automatic?... Anyboby have other views? Or aspect that they think it should be emphasize?

Now, I hope I don't get anybody offend or hurt by debating/discussing their views. Please... don't get mad at me...
see this as a "together sharing our views"...


And thanks for sharing your ideas... and telling when and where I may be wrong!!!


Hey Scott, have you figure your way (http://www.talkshopbot.com/forum/cgi-bin/discus/show.cgi?tpc=312&post=16328#POST16328)?!

[Ryan Patterson]

Paco,
As Frank said there is no real difference between spoil or top of material. You just have to keep in mend where the Z is. I zero to spoil board 80% of the time. When you zero to spoil board as long as you have a positive Z value in you part you will not cut into the spoil board. The one thing you do have to remember is to make sure your jogs and safe hight for z between parts is higher than the material you are cutting. This is why as frank said he finds it to be safer to the top of his material.

[Brady Watson]

Paco,
If you zero to the spoilboard and you tell the machine that your safe Z is 1", what if you are milling 1.5" material? It is highly likely that you will not realize that the Z is not high enough until you ram into the material if you have to move it home or run another toolpath.

Again, get better at measuring your material. For instance, 1/4" acrylic may be anywhere from .21 to .25+ inches thick. 3/4" plywood is a lot of times 18mm thick which is about .70" thick.

Measuring from the top of the material block is the standard and accepted way of machining in almost all cases. If you are 'getting bored' of using the z-zero plate, then you don't know how good you've got it. It wasn't that long ago that you had to use a piece of paper, a light touch on the arrow keys and/or an educated guess to determine your zero. All the CAM programs I have seen ask you how thick your material is and assumes you'll zero from the top by default. Other CNC machines can only dream of having the luxury of a Z-zero plate.


-Brady

[fleinbach]

Brady,

That's pretty interesting. I might just try one. I was in a hurry earlier and didn't explain the reason I brought this up. Most of my errors in measuring come from reading the tape upside down. Let's say you want to mark the material at 12 1/4" When reading the tape upside down I will sometimes look at the 12 and use the 1/4" mark on the tape to the right. I attribute this to seeing the 12 and then my mind reverts to reading from left to right. Of course this causes me to make a mark at 11 3/4". After cutting I start looking for the board stretcher.

Paco,
By safer I'm referring to an operation that has less chance of breaking a bit, ruining the part or knocking something on the shopbot out of alignment. None of which are good things. The problem is with the safe Z height. Zeroing in the shopbot software parks the bit at .5" above 0. If zeroing to table surface and cutting 3/4" material you may want to set safe Z at 1 1/4" so it will clear the material safely. Now if you change to say 2" material and forget to move safe Z to a new height you may possibly run the bit into the material if using an M or J move.

Now when you Zero to the material surface it doesn't matter how thick it is, the safe height of .5" will always be above the material.

[ron brown]

Beady mentioned, "Again, get better at measuring your material. For instance, 1/4" acrylic may be anywhere from .21 to .25+ inches thick. 3/4" plywood is a lot of times 18mm thick which is about .70" thick"

Yes Brady - and that can be on the same sheet.

A Lot of this "Top or bottom register" depends on what one is doing, the accuracy needed and the consistancy if the material.

The problem of plowing into material on tranverses is just a matter of paying attintion to details... and being both smarter than the material, machine and person running the machine if you have a"hired" operator. I normally zero to top material. I will put an extra line in files where I want the zero on the table or where I want the bit zeroed and material thickness added telling the file runner - or myself if I run the file at a later time, what the intentions of the file writer were at the time of creation.

Ron

[Ryan Patterson]

You could reference Z from spoil board and input thickness of material to zero to the top.
Add this to your z zero file and add + &mt to the end of &ZBOT= 0.131.
&ZBOT=0.131 + &mt ' 0.131 is the plate thickness)

INPUT "Enter material thickness" &mt
&mt="&mt"
&mt=&mt*-1


If you entered .75 Z will be zeroed .75 above spoil board, and then any Z depth not greater than -.75 you will not cut into spoil board. The disadvantage would be if you needed an accurate pocket depth then you would have to measure material thickness and enter the exact material thickness as Ron and Brandy mentioned.

[Brady Watson]

Beady...and Brandy...Sound like imposters!


-Brady

[ron brown]

keyboards were designed for folks who have dainty hands and do no manual labor in their lives...

deal with it

Ron

[Ryan Patterson]

Sorry brady