Z axis set too low on large shop shapeoko

[Kurt Krummel]

I was using the large shop shapeoko tonight and after manually setting my z height, not using the probe, I proceeded with my cuts. First I was using a 60 degree v bit and thought it was cutting a bit deep, but wasn’t sure. However, when I switched bits and again reset Z manually, this time with a bowl bit, I was absolutely convinced that my z was about 1/4” too deep. Any ideas why this happened? I made sure in v-carve that z was set to measure from the top surface and I double checked my material thickness matched the value in v-carve. Any other ideas? Could the system be out of alignment somehow?

[Rob Sandstrom]

Did you figure out what was happening and get any replies to your question?  Sorry for the late response.  I don’t use the shop CNC’s because I have a Lguna IQ CNC at home.  Based on your description facts:

1. You validated you were zeroing off the proper location (ie. Material surface)
2. You zero’d after each bit change before machining

I can initially only think of one other thing that you didn’t say was checked:  Did you ensure that in vCarve the proper post processor was picked.  See picture below.  Based on your other discussion facts I would assume the answer is yes.  If this was the case, it would seem it almost has to be something going on with the machine control and it would be good to know this and have someone that knows more about the Shaepoko look at the machine such that someone doesn’t ruin a project that the bit of the depth is critical.  Good luck on resolving why this happened.

[Rob Sandstrom]

As you know vCarve is built to work with many machines, each one has different ways they read/interpret the g-code  (machine language) to perform the tasks they are going to perform.  I think of it as many people say they speak "English" but if down in the deep south, up in Boston, or over in London, one group my have a hard time understanding the other.  So although g-code is machine language, how it is arranged and understood by a specific CNC controller (brain) can be slightly different.  If that makes sense.

So when saving the toolpath to go to a given machine in vCarve - see previous picture, there is a drop down menu in which the operator is expected to pick the right Post Processor, so the g-code (tool language) is structured to allow the given machines controller (brain) to interpret the code correctly.  In my picture of my previous post you can see that I selected the Laguna IQ - the arrow is pointing to it.  If I were going to take my same design and cut on a Shaepoko I would pick the Shaepoko (inch) or Shaepoko (mm) for the post processor.  As can be imagined, if someone picked Shaepoko inch vs. Shaepoko mm machine this could still run the Shaepoko but the results may be different on depth, etc.

See example of Shaepoko (inch) below.  Hope this helps.

 

[Mark Stook]

Hi All,

I haven't been to the shop in a very long time so I'm not sure what kind of touch probe they are using.  Is it the one where you need to subtract the thickness of the probe or the new one that you have to place on the material differently based on if you are zeroing just Z or X, Y & Z?  If I knew that I may be able to make a more educated guess.

Mark

--
You received this message because you are subscribed to the Google Groups "SDFWA Digital Tools" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sdfwa-digital-t...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/af49e5e5-692e-4539-8a50-4aa0aad47ccdn%40googlegroups.com.

[Travis Good]

The latter ... on the surface for Z-only and over the corner for XYZ zeroing. 

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAKoM_Z-FWe8pQErux9cZMnsd5PcPLnYegwPcGbR_8Q5feO_DnA%40mail.gmail.com.

[Rex A Schildhouse]

Everybody knows this but I didn’t pick it up so that opening statement is false. I have learned not enough about CNC and been frustrated by my Shapeoko more than I want to be. One of the lessons learned was to use the probe to zeroize X, Y, and Z with something like a ¼” end mill. Then zeroize only the Z azimuth with the V-bit. Once X and Y are set, I am committed to not moving the stock or turning the Shapeoko off until the project is complete.

I bought one V-bit on line, a 90 degree bit by order and container. It was frustrating me to no end. Measured it and it was actually a 60 degree bit in a 90 degree box. Amazon sent me a 90 degree V-bit, correcting the order, and told me to keep the 60 degree bit.

I have tried to set zeros by sight, while occasionally good, found the Carbide Probe best.

Agree with the “Make sure you are using the correct G Code option.”

Thanks,

Rex

[krum...@gmail.com]

This is super-helpful. Thanks everyone.

I'll definitely check the post-processor next time and can try to set X/Y/Z first using an endmill, then reset Z only with the V-bit, but it's strange that the X/Y set manually were perfect... Only Z was wrong... 

I'm 100% positive I have the right V-bit.

[Susan Lively Klug]

I have had the bit slip down after setting Z because I didn't tighten the collet enough.  

I have no idea why the machine even worked at all at that point.

At least I believe that is why my Z was too deep as I hadn't done anything different before or after and was always ok with my Z.

Are you doing a 2 sided project?  I was unable to complete a 2 sided spoon before the covid times, the 2nd side always went deeper into the material than I expected it to based on my preview and measurements.  I didn't have time to debug.

Susan

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/2bc25140-0527-4b34-8cb5-c62d4e572e06n%40googlegroups.com.

[love.evan]

Before the shop closed I had a huge project completely ruined by exactly what you are referring to...

... the z randomly going deeper...

...I don't think you made any mistakes...

...I think there is something wrong with the machine or software...

... possibly that the z, sensing resistance due to particularly hard wood or an end mill who's cutting edges don't go all the way to center and thus can't plunge at a steep downward angle, then tells the massive to dive more...

... or that the head unit just doesn't count z units correctly, particularly when encountering resistance higher than a sharp bit into soft wood. 

I mentioned the issue at the time.   I think it's a major flaw in the larger shapeoko that until fixed makes the unit unreliable, particularly for expensive or important projects. 

All that to say, other than the possibility of using an end mill who's cutting edges don't go all the way to center which then resists a step plunge, i don't think it was you, I think the machine has a serious flaw that shows itself seemingly randomly, but far too often to ignore. 

:)

Evan

evan moulton

love-real-estate.com 

619 823 9650

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAF8oYNHu8GNQ0Snrywuc%3DZtfpdUQwznzjcRxxzC-CXjNWc-bTw%40mail.gmail.com.

[krum...@gmail.com]

That's such a bummer. Glad I'm not the only one though! I think my project might be salvageable, but yes, it definitely turned some beautiful wood into a potential scrap.

Does anyone know if both shapeoko's have this problem? I could have used the smaller one in this case, but had just assumed the larger one was in good working order. Sigh...

[Doug LaRue]

On Friday, August 14, 2020 at 12:22:38 PM UTC-7, Evan Moulton wrote:

Before the shop closed I had a huge project completely ruined by exactly what you are referring to...

<clip>

... possibly that the z, sensing resistance due to particularly hard wood or an end mill who's cutting edges don't go all the way to center and thus can't plunge at a steep downward angle, then tells the massive to dive more...

... or that the head unit just doesn't count z units correctly, particularly when encountering resistance higher than a sharp bit into soft wood. 

The Shapeoko machines I've seen have no motor positional sensing and just use stepper motors to move X, Y and Z. There is no way for it to 'sense resistance' and instruct the Z axis to dive more and as a matter of fact, usually what happens is the stepper motor can not make the step it was instructed to do, skips a step, and from then on it moves relative to a different "zero" position since it no longer is set to what it was originally zero'd at.  High end machines will have position sensors on the motors and at that point there can be more power to the motor applied if a step is skipped to get it back in the correct position.  A VERY poor sensing system would allow errors to creep into the system and they would be thrown off the market in a heartbeat since position sensing does not come cheap and it must work.

I would tend to think there could be a stepper motor and/or stepper motor driver over heating issue which could allow the weight of the spindle and forces on it let it drop a step. Having higher feed rates and dull end mills will put lots of forces on the motors and drivers and can cause over heating.  Software normally( non Windows ) does not randomly do things and especially g-code since it's just a text file being read and there's no writing back to the file when running the CNC operation.

My small CNC did experience a 'deep dive" once but it was due to the coupler between the stepper motor and Z axis lead screw coming loose. With the stepper motor no longer holding it in place, the weight of the spindle went into free fall as the lead screw freely turned. 

But it sounds like possible over loading causing overheating and loss of steps more than anything else.  

[love.evan]

Yes, but wouldn't a loss of steps on z axis result in a shallower cut rather than a significantly (¹/⁴ inch) deeper cut? 

evan moulton

love-real-estate.com 

619 823 9650

-------- Original message --------

From: Doug LaRue <doug....@gmail.com>

Date: 8/14/20 12:42 PM (GMT-08:00)

To: SDFWA Digital Tools <sdfwa-dig...@googlegroups.com>

Subject: Re: Z axis set too low on large shop shapeoko

--

You received this message because you are subscribed to the Google Groups "SDFWA Digital Tools" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sdfwa-digital-t...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/174291a6-e63e-44f6-a333-1675f723c734o%40googlegroups.com.

[Doug LaRue]

On Friday, August 14, 2020 at 12:48:58 PM UTC-7, Evan Moulton wrote:

Yes, but wouldn't a loss of steps on z axis result in a shallower cut rather than a significantly (¹/⁴ inch) deeper cut? 

Not if you lost a step rising up out of a cut. Remember, this is where the system is moving against the cutting forces, against gravity and against frictional loads of the lead screw and bearings.

 

[Doug LaRue]

I might add that things like the speed if the moves can effect this. I don't know about Shapeoko settings but on my CNC I can specify move speeds and accelerations and I can set these to the point where steps are missed.

Throw in motor and driver overheating and what was thought to be a safe setting can become inadequate. 

[Rex A Schildhouse]

Attached are a couple of MANY efforts to get VCarve to do pockets on my Shapeoko XXL properly. Every pocket border was dug deeper than the pocket. Tried to work with Vectric and got asked the same questions every single time, providing the same answers, only to be asked the same questions again. Never got any help out of Vectric support. Finally gave up, asked for a refund, got it, and uninstalled in on my system. Went back to Carbide Create, not as good a program, but it does what it says it will do.

The preproduction image of what VCarve was supposed to do was correct, the results were not.

So, I feel your frustration. All I want to do is make some neat signs and toys for my neighbors kids and my granddaughter. And that is NOT happening.

Thanks,

Rex

[love.evan]

Interesting, yes, missing a z axis up (or several) would do it. 

If that were the case on this particular machine, is there a part that could be replaced with a better part that would be able to handle a longer job before failing (during the job) in this way? 

evan moulton

love-real-estate.com 

619 823 9650

-------- Original message --------

From: Doug LaRue <doug....@gmail.com>

Date: 8/14/20 12:55 PM (GMT-08:00)

To: SDFWA Digital Tools <sdfwa-dig...@googlegroups.com>

Subject: Re: Z axis set too low on large shop shapeoko

--
You received this message because you are subscribed to the Google Groups "SDFWA Digital Tools" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sdfwa-digital-t...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/dc41511d-8717-47f3-b0a1-6b05fbe8f4d4o%40googlegroups.com.

[krum...@gmail.com]

I think the overheating or load causing it to skip z is not too likely. The cuts started out from the very beginning too low and did not get worse over the course of the cut. I think I may try to redo this with carbide create and see if that fixes the issue.

[Doug LaRue]

On Friday, August 14, 2020 at 1:21:04 PM UTC-7, krum...@gmail.com wrote:

I think the overheating or load causing it to skip z is not too likely. The cuts started out from the very beginning too low and did not get worse over the course of the cut. I think I may try to redo this with carbide create and see if that fixes the issue.

If there are no random variations in depths of cuts or even changes in depths where they are not supposed to be then it's unlikely to be the machine.

If you are just not getting the expected depths or sizes of things then I would expect that to be software settings and setup issues.

I've seen a few examples of people setting up their zero manually and after the paper test they moved to a different location and verify the Z0 command got the endmill right to zero at that location.

I have a zero probe and I will do that same check just to be sure Z0 is Z0 at the location I expect it to be and that depends on where I set that in my slicing software( Kiri:Moto ) as it lets me specify zero on the build surface or zero on top of the work piece.

Sorry, for bringing up all the missed steps descriptions as I thought that was something which was happening, not a setup issue.

Probably need to go do some careful checking and rechecking to see where your workflow and setup went wrong.

[Lance Grucela]

I still think it may be worth exploring the missed step issue as I experienced something similar recently when cutting some thicker stock.  I learned a valuable lesson to pay attention to something that I wasn't really aware of when the depth of cut was waaaaaay off despite my double and triple checking software and setup.  Also, the issues didn't repeat itself every time.  

What was happening was that after correctly zeroing XYZ using the probe, the motor skipped steps in the z axis when it retracted upwards (trying to go higher than the machine limits).  I found that the safe z height in vcarve was unnecessarily large (.08") and causing this behavior just before the cut began.  Changing the value to (.02") and also paying attention to be sure the bit does not stick out too far has prevented this from occurring again for me this far. I think there was a post on the shapeoko forum about this and Ill be sure to pass along if I come across it again.

It was very frustrating for a while because it had occured just after I installed the bitsetter, so I was pretty convinced that was causing the issue.  I've since added a step to double check the z height using rapid position immediately before starting the cut.  

Hopefully this is helpful... 

--

You received this message because you are subscribed to the Google Groups "SDFWA Digital Tools" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sdfwa-digital-t...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/24fdc3bf-c1cd-4ba6-917f-e31990eb29abo%40googlegroups.com.

[Evan Moulton]

Ya, the missed z up steps would actually explain the experience I had.  And it was randomly happening and at random amounts of z not going up between levels of cuts.

It did actually cut into material that was between it and the next spot to start a depth cut because it didn't rise up to clearance depth between levels of z cuts.

Perhaps we could replace the z step motor.

:)

In the absence of concepts

Is the absence of separation

And the presence of Love

 

evan

619 823 9650

love-real-estate.com

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAKK5Qt6QTvZ_UnU74VdfW34T%3Dbe0USv5WYmL1QZFFcBa%3Dv3OeA%40mail.gmail.com.

[Susan Lively Klug]

 It would be nice to have a reproducible short test case so we can prove any changes to have fixed the problem.  

My test was a small bowl shape, I carved the inside then flip, then carve the outside. Not really a quick test... But may be reproducible.

Is this only happening if you zero z at the spoil board?  I believe that was the instruction for two sided projects.

Susan

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/577726968.2253773.1597440099627%40mail.yahoo.com.

[love.evan]

I zeroed at the top surface of the project. 

But the z definitely did not go back up to clearance depth (or even surface depth) before it went across my project to do its second (and third) passes at the next depth. 

So, I would say it's a case of the z not going up in increments equal to its down (thus losing its original z setting) and ending up too deep.

If it is mechanical, rather than software, perhaps the z stepper motor is missing teeth or otherwise can't lift the router under certain circumstances. 

Somehow, between steps, it doesn't get back up as much as it came down. 

evan moulton

love-real-estate.com 

619 823 9650

-------- Original message --------

From: Susan Lively Klug <slive...@gmail.com>

Date: 8/14/20 4:11 PM (GMT-08:00)

To: Evan Moulton <love...@yahoo.com>

Cc: Doug LaRue <doug....@gmail.com>, Lance Grucela <lance....@gmail.com>, SDFWA Digital Tools <sdfwa-dig...@googlegroups.com>

Subject: Re: Z axis set too low on large shop shapeoko

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAF8oYNFh40%3Dd1Qtc%2B01s2ivH7aBOYCE_%2B9E3YP1ZvP%3DcPpMHqA%40mail.gmail.com.

[krum...@gmail.com]

This is exactly what I saw when I had switched over to my bowl bit to cut 2 handles on my board. After the first handle was cut, as the bit came across the piece to cut the 2nd handle, it dug a channel right across my piece. The first handle it cut was definitely too deep though, so I had assumed that it just went too deep overall, but then I guess it should have at least come back up to clear the surface between cuts, so it definitely seems like a hardware issue now that you bring this up.

"I zeroed at the top surface of the project. 

But the z definitely did not go back up to clearance depth (or even surface depth) before it went across my project to do its second (and third) passes at the next depth. 

So, I would say it's a case of the z not going up in increments equal to its down (thus losing its original z setting) and ending up too deep."

[Doug LaRue]

On Friday, August 14, 2020 at 6:22:28 PM UTC-7, krum...@gmail.com wrote:

This is exactly what I saw when I had switched over to my bowl bit to cut 2 handles on my board. After the first handle was cut, as the bit came across the piece to cut the 2nd handle, it dug a channel right across my piece. The first handle it cut was definitely too deep though, so I had assumed that it just went too deep overall, but then I guess it should have at least come back up to clear the surface between cuts, so it definitely seems like a hardware issue now that you bring this up.

"I zeroed at the top surface of the project. 

But the z definitely did not go back up to clearance depth (or even surface depth) before it went across my project to do its second (and third) passes at the next depth. 

So, I would say it's a case of the z not going up in increments equal to its down (thus losing its original z setting) and ending up too deep."

And that sounds like what would happen if SOMETHING about the lifting of the Z axis results in missed steps. Things I would check on my machine first would be how hot the Z axis motor gets and how hot the Z axis driver gets.

Then I would check the Z axis driver power setting. That could be a software setting or it could be a voltage read on the metal calibration screw to ground. Shapeoko should have information on this or something in the support forums.

Then I would check the Z axis lead screw and bearing to see that nothing is adding excessive loading on the Z axis making it too difficult to raise the motor and sled without skipping steps.

  

[love.evan]

Sounds like good things to check.

Do we have a person at the shop that is reading this email chat who is ready, willing, and able (and allowed) to check these things? 

evan moulton

love-real-estate.com 

619 823 9650

-------- Original message --------

From: Doug LaRue <doug....@gmail.com>

Date: 8/14/20 6:53 PM (GMT-08:00)

To: SDFWA Digital Tools <sdfwa-dig...@googlegroups.com>

Subject: Re: Z axis set too low on large shop shapeoko

--
You received this message because you are subscribed to the Google Groups "SDFWA Digital Tools" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sdfwa-digital-t...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/558951d6-fd35-4802-91a7-9a9e8384236bo%40googlegroups.com.

[krum...@gmail.com]

Pat Duffy offered to help me troubleshoot. It sounds like a number of us have had similar issues, so we're either all making similar mistakes, or there's a HW/SW issue on our shop CNC.

I'll try a few tests of various conditions to see what I can figure out over the next couple of weeks and report back.

[Jeff Romek]

I have CNC time on Monday.  I'll do a test cut on MDF first and report back.

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/f50b7dc4-8368-4275-a298-888635f84f1bn%40googlegroups.com.

[Gerald Mackelburg]

The limit switches will only disable stepper motor stepping when actuated during jogging.  If the software commands the steppers past the limit switches for whatever reason at any other time, they will hit hard limits.  The steppers will (quite audibly) slip (skip steps) until the software thinks they’re where they should be (much higher than reality in this case?). 

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAEHz7mSgjAehZsjD2idh%2Bk3yHheXkG6FdFor-qOSG%2BbE0MjKYw%40mail.gmail.com.

[Gerald Mackelburg]

OOPs – maybe not “quite audibly” if the router is spinning at the speed it should be.  Since there’s only 3 switches (instead of 6), I guess I should have called them “homing switches” rather than “limit switches”?

 

From: Gerald Mackelburg <gmack...@san.rr.com>
Sent: Saturday, August 15, 2020 9:20 AM
To: 'Jeff Romek' <jeff...@gmail.com>; 'krum...@gmail.com' <krum...@gmail.com>
Cc: 'SDFWA Digital Tools' <sdfwa-dig...@googlegroups.com>; 'gmack...@san.rr.com' <gmack...@san.rr.com>
Subject: RE: Z axis set too low on large shop shapeoko

 

The limit switches will only disable stepper motor stepping when actuated during jogging.  If the software commands the steppers past the limit switches for whatever reason at any other time, they will hit hard limits.  The steppers will (quite audibly) slip (skip steps) until the software thinks they’re where they should be (much higher than reality in this case?). 

 

From: sdfwa-dig...@googlegroups.com <sdfwa-dig...@googlegroups.com> On Behalf Of Jeff Romek
Sent: Saturday, August 15, 2020 7:52 AM
To: krum...@gmail.com <krum...@gmail.com>

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAEHz7mSgjAehZsjD2idh%2Bk3yHheXkG6FdFor-qOSG%2BbE0MjKYw%40mail.gmail.com.

[Travis Good]

For everyone’s benefit ...

I was at the Shop yesterday with Pat as he tried to recreate the problem. He couldn’t and he tried hard. A problem can’t be solved if it can’t be recreated. Hopefully Kurt (or another) will provide the missing insight. 

We all want to figure this out and fix what ever the problem is: hardware, software or process. 

Thanks, Travis

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/f50b7dc4-8368-4275-a298-888635f84f1bn%40googlegroups.com.

[Susan Lively Klug]

I could send you my spoon file but 

I think a comparison of work flow may help.

I use vcarve desktop at home, bring file in and use shop computer to create toolpaths. (I think.. it has been too long)

Susan

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAJ6RBKdnOGQrHpSrJtT%2BNdVuejOfv%3DmoigvyByAsP9STjgM-GQ%40mail.gmail.com.

[love.evan]

Just to be completely fair to Pat...  ... the problem has been recreated (re-experienced) ...  ... just not during intentional observation. 

;)

evan moulton

love-real-estate.com 

619 823 9650

-------- Original message --------

From: Susan Lively Klug <slive...@gmail.com>

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAF8oYNE-r%3Ddw-qU7Bdgd9OdZFjLVxwyX6EZHBMMLYdy%3D3fQwbQ%40mail.gmail.com.

[Susan Lively Klug]

Yes thank you to Pay & Travis for trying! 

I am sure combined we can figure it out!

And can I say that I am glad it wasn't just me?

Susan

On Sat, Aug 15, 2020, 1:33 PM love.evan <love...@yahoo.com> wrote:

Just to be completely fair to Pat...  ... the problem has been recreated (re-experienced) ...  ... just not during intentional observation. 

;)

evan moulton

love-real-estate.com 

619 823 9650

-------- Original message --------

From: Susan Lively Klug <slive...@gmail.com>

Date: 8/15/20 1:30 PM (GMT-08:00)

To: Travis Good <travi...@gmail.com>

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAF8oYNE-r%3Ddw-qU7Bdgd9OdZFjLVxwyX6EZHBMMLYdy%3D3fQwbQ%40mail.gmail.com.

[krum...@gmail.com]

Hey everyone, 

 I just wanted to give an update on my troubleshooting efforts. The good news is that I I was able to successfully make my cuts and confirm that there is indeed nothing wrong with the hardware (or likely software) on the Shapeoko. The bad news, is that my "control" condition, repeating the same cut that caused problems last time, did NOT cause the same problem! I couldn't do the EXACT same cut, as my workpiece was about 1/4" less deep in Z (after some thickness planing), but it was as close as I could get in every other way. I used the same V-carve file, same export procedure, and same bit-setup. This time, I confirmed that I had the "shapeoko" gcode file selected when saving from V-carve (and it was the default). I was hoping that would have been the cause, and I guess I can't fully rule that out, if for some reason, last time I did my cuts, a different default was popping up, and I didn't notice, but I think it's unlikely, as it would mean someone would have changed the default last week, and then now changed it back. The ONLY other thing that I can think that it could have been is that I DID put the bit more fully into the collet this time and perhaps that extra 1/4 inch, plus another 1/4 inch thinner piece, made some difference in clearance over the workpiece and a more accurate Z-height, but other than that, I'm out of ideas.

Thanks everyone for the great ideas and support. I'm sorry I couldn't replicate it, or figure it out definitively, but I am glad it is working! If  anyone else has this problem in the future, maybe we can all figure it out next time.

Kurt

[Travis Good]

Thanks for updating us. 

We can’t solve a problem that we can’t recreate. That said, there is something going on. It might be as simple as people not tightening end mills sufficiently in the collet. Feedback such as yours will help us figure it out. 

I hope everyone will continue thinking about this as they use the CNC. In the end we can make changes to solve the problem. We want members to have a better and more consistent experience for all. 

Your efforts are appreciated, Kurt.

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/f8df7558-f71a-4adc-b9c3-c78458d1880dn%40googlegroups.com.

[Tim Peachey]

I can't offer direct experience, but I discussed this with my brothers who both use routers much more than I do. Both of them have experienced what Susan suggested - the strong pull created by the twist of the bit can pull the bit down in the collet. This was observed both in a CNC router and a hand router. They specifically looked at the position of the bit after it happened to confirm that the bit had in fact moved. Since then, they make sure that the bit and collet are clean, and are sure to tighten the bit securely. For them, this has eliminated the problem.

On Sat, Aug 22, 2020 at 5:52 AM krum...@gmail.com <krum...@gmail.com> wrote:

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/f8df7558-f71a-4adc-b9c3-c78458d1880dn%40googlegroups.com.

[Doug LaRue]

In a membership shop, maybe there's something to the use of those endmills which have the ceramic depth/insertion rings on them. They would make sure the users insert the endmill in fully and if there's any issue with bit creep it would be visible by an equal sized gap between the ceramic ring and the collet as there would be too much depth cut into the work piece.

One could also make sure in their workflow they have z=0 after zeroing and if there's an issue with the Z axis, do the same check for z=0.  If it's off it's either bit creep or step loss and without the ceramic ring it's pretty much going to be a guess as to which is the cause.

[Tim Peachey]

I'm still a novice on the CNC, but I recall seeing an option in Carbide Motion to jog the bit to x,y=0 and z=6mm. I'm going to make a 6mm spacer and use this position for the check, so that I don't crash the bit into my workpiece. 

On Sat, Aug 22, 2020 at 12:59 PM Doug LaRue <doug....@gmail.com> wrote:

In a membership shop, maybe there's something to the use of those endmills which have the ceramic depth/insertion rings on them. They would make sure the users insert the endmill in fully and if there's any issue with bit creep it would be visible by an equal sized gap between the ceramic ring and the collet as there would be too much depth cut into the work piece.

One could also make sure in their workflow they have z=0 after zeroing and if there's an issue with the Z axis, do the same check for z=0.  If it's off it's either bit creep or step loss and without the ceramic ring it's pretty much going to be a guess as to which is the cause.

--

You received this message because you are subscribed to the Google Groups "SDFWA Digital Tools" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sdfwa-digital-t...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/d07aa293-c2f1-406b-9e7f-c9b87567140fn%40googlegroups.com.

[Lance Grucela]

Tim-- that's a good idea to make a spacer to double check.  (I have a small veritas setup tool which I find extremely useful to keep in my pocket in the shop but of course, it's imperial.)

After ruining a piece due to missed z steps, I typically use the Z+6mm rapid position to verify by switching to the 1mm increments and then lowering 5-6 steps to verify correct z height (i.e. no missed steps) before running.

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAMWu08VN0suY8h%2BjeqVFjKNY3S5jHdSundAQdkK6P6df5%3DQWEQ%40mail.gmail.com.

[love.evan]

Unfortunately for the current theory of bit drop in the collet, I am absolutely certain that is not case in my experience of the large Shapeoko on two completely separate occasions, failing to rise back up completely between depth passes, and thus cutting too deeply and destroying projects (and fancy wood). 

Which means, unfortunately again, that the issue still lurks in the background ready to affect another project. 

evan moulton

love-real-estate.com 

619 823 9650

-------- Original message --------

From: Tim Peachey <tim.p...@gmail.com>

Date: 8/22/20 1:55 PM (GMT-08:00)

To: Doug LaRue <doug....@gmail.com>

Cc: SDFWA Digital Tools <sdfwa-dig...@googlegroups.com>

Subject: Re: Z axis set too low on large shop shapeoko

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAMWu08VN0suY8h%2BjeqVFjKNY3S5jHdSundAQdkK6P6df5%3DQWEQ%40mail.gmail.com.

[Doug LaRue]

On Saturday, August 22, 2020 at 2:11:39 PM UTC-7 Evan Moulton wrote:

Unfortunately for the current theory of bit drop in the collet, I am absolutely certain that is not case in my experience of the large Shapeoko on two completely separate occasions, failing to rise back up completely between depth passes, and thus cutting too deeply and destroying projects (and fancy wood). 

But if the bit did slip down in the collet that is exactly the type of result you would get.  Unless you are talking about being off by 10mm or so since that would be quite noticeable with the endmill almost falling out of the collet.

Maybe we need a torque wrench type collet wrench so there's no question about tightening correctly.  If it doesn't exist, it should. ;-)

[Evan Moulton]

Yes, it was down by over a quarter inch and the bit was still fully set in the collet.  So, definitely not the issue.

In the absence of concepts

Is the absence of separation

And the presence of Love

 

evan

619 823 9650

love-real-estate.com

--

You received this message because you are subscribed to the Google Groups "SDFWA Digital Tools" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sdfwa-digital-t...@googlegroups.com.
To view this discussion on the web visit

https://groups.google.com/d/msgid/sdfwa-digital-tools/383d80fb-e673-4f1c-9d68-9ff5067a542cn%40googlegroups.com

.

[ptsch...@gmail.com]

I suppose using the inside and outside calipers that are in the toolbox can give a more accurate indication of the router slippage to start to narrow down the issue.

Paul

[Travis Good]

Pat, our CNC guru at the Shop, has created such a space measuring tool.

He basically created a isosceles triangle where along the diagonal

he has steps at various heights, e.g. 2mm, 4mm, 6mm, 8mm, etc.

Slide the base under the bit to the snuggest fit ... voala! Travis

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAMWu08VN0suY8h%2BjeqVFjKNY3S5jHdSundAQdkK6P6df5%3DQWEQ%40mail.gmail.com.

[Mike McElhiney]

All this focus on the Z axis has to stop.  The poor Y and X axies are complaining about being ignored.  They might be planning something.  We need to be careful with our only X, Y, and Z axises - or is it axisici?

Best,

Mike

On Aug 22, 2020, at 4:37 PM, Travis Good <travi...@gmail.com> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAJ6RBKeNHm1nGNZk-Tev%3DM3bwgy%3DqPigz_DWVBwk8p1R4s_qfA%40mail.gmail.com.

[Jeff Romek]

Hey Kurt, sorry I didn't get back to you sooner but I had no issues at all when I used the machine last week.  

Jeff

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/f8df7558-f71a-4adc-b9c3-c78458d1880dn%40googlegroups.com.

[Rex A Schildhouse]

I have read almost every one of the e-mails about inappropriate depth on the Shapeoko. I have a Shapeoko XXL and I had VCarve as well as Carbide Create for creation and Carbide Motion for operations. I use the Carbide 3D probe for setting X, Y, and Z zeroes. I had inappropriate depth issues on anything I created with VCarve with pockets. While I cannot comment on the issues or processes of others or their results and causes, I can relate my total experience with the issue. And I will attach at least one photo for clarification.

Among other things, I found an Internet reference that stated there is a difference between an end mill and a square / rectangular router bit. An end mill is supposed to cut on a straight plunge easily while a square / rectangular router bit will not plunge cut as easily. The reference stated that most end mills the individual had bought were actually router bits, some nice photos were included. I have yet checked mine to see exactly what I have. His statement was, “when able, ramp any and all bits into the cut.”

When I put in an eighth inch end mill in my Shapeoko and carve a name plague I get decent and expected results. When I check X, Y, and Z zero it is still correct.

When I take the Pocket option in VCarve and use a clearing bit, again an eighth inch end mill, everything goes right UNTIL the VCarve is done with the 90 degree V-bit. Using the probe to reset only the Z zero, the values are verified to the pocket depth and the V-bit digs in 0.125 inch to 0.1875 inches, or more, below the depth of the pocket. And it is uniform on the outer frame of the pocket and uniform surrounding the letters of the name.

Depth to the point that it destroys the name.

So I step back and check Z zero and it is still correct. To believe, accept, that the bit creeped out due to insufficient tightening, and then it reset itself to the original depth is a bit past my understanding of physics and process.

The other issue which I base my disagreement on is that the depth of the V-bit carving is consistent throughout the work. It never starts correct and then increases, it never returns to expected. It starts in excess, stays in excess, and still checks to Z zero after the work is done.

Slowing the feed rate, increasing the travel of the ramping insertion did nothing. I even tried inserting an eighth inch end mill and telling the Shapeoko that it was a 90 degree V-bit and surprise surprise, inappropriate depth of consistent value. And another piece of wood for my neighbor’s fire pit.

One of the minor issues I had was I bought a 90 degree V-bit on Amazon. Did not check it upon arrival. Turned out it is a 60 degree V-bit in a 90 degree V-bit box. Amazon corrected it. However, it yields the same results.

In working with Vectric I got absolutely nothing other than questions.

From Vectric “Can you send us the file you are using?” and I did. When I asked what they found in the file, the answer was “Have you checked your feed speed?”

“Yes, the feed speed is good, the router is not bogging down, burning the wood, chattering, smoking, skipping, or rough cutting the path. What did you find in the file I sent?”

The reply would be “Are you setting the X, Y, and Z zeroes to the top of the project or the bottom?”

In the VCarve application the image is correct. The results are not what is depicted.

Weeks of no support, just question after question after question from Vectric before I finally gave up and asked for a refund.

So, based on my Shapeoko XXL, my results, my investigation, my results on MANY efforts, the V-bit is not creeping in my Shapeoko. Something between design and finished project is wrong.

0-2020-05-08-Rex-A-Schildhouse-005.jpg – Don’t you just love Wyatt’s “t”s?

0W-2020-05-13-Lilli-Name-Plaque-Carbide-Create.jpg – Carbide Create and a 0.125 end mill.

2020-05-13-Lilli-Name-Plaque-154604w.jpg – Telling Shapeoko that I was using a 90 degree V-bit and actually using a 0.125 inch end mill.

Rex A Schildhouse-Vectric-2020-05-05.pdf

One very frustrated Shapeoko - Former Vectric user.

Thanks,

Rex

[davism]

Kurt,

 

My apologies also.  I used the CNC machine a couple of weeks ago to cut a guitar fretboard, something where the z axis position is critical and I had no problems with it at all.

 

 

Mike Davis

 

 

 

From: sdfwa-dig...@googlegroups.com <sdfwa-dig...@googlegroups.com> On Behalf Of Jeff Romek

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAEHz7mTeYMdY0XBa1%2B6oyTMtunbLDX_zguc%2BzuUMzgJ%3DOUwREw%40mail.gmail.com.

[krum...@gmail.com]

Rex,

 Based on your experience, you seem to be saying that it is likely a V-carve issue (with V-carve customer support unable to help). Have you tried using carbide create instead, and does that prevent this type of error? Or is this a problem for which you have yet to find a solution?

[Rex A Schildhouse]

Based on my experience, not professional knowledge, it is a VCarve talking to Carbide Motion talking to the Shapeoko issue. And that answers almost nothing but it does focus the issue on VCarve. Sorry. I have made numerous things with Carbide Create, which does not do pockets well, and I get decent results. My statement is VCarve and Carbide Motion and the Shapeoko are in disagreement whenever I select the Pocket option in VCarve with a clearing bit followed by a V-bit. Played with one simple little block, which took over an hour. Made a letter block about 1 1/2" x 1 1/2" with one single letter in it and pocketed, like the children's' learning blocks. Told VCarve it was a 1/16" end mill while using a 90 degree V-bit. Knew it was going to be rough, it is called troubleshooting. The pocket was cut to the expected depth, just rough, ridged. When I told Carbide Motion to run the pocket and letter border with the 90 degree V-bit, it did exactly as expected, it gouged deeper than set on the borders of the pocket and around the letter. And I had not changed the bit AND DID NOT reset the Z zero. Tried it again, it takes over an hour to run, this time with a 1/8" end mill as a pocket clearing bit, which ate the letter but did not really care. Troubleshooting. Changed to the 90 degree V-bit, reset Z zero, and got the gouge.

My statement, on my setup of computer, VCarve, Carbide Motion, Shapeoko XXL, SOMETHING is wrong. And ZERO support from Vectric. Called Carbide 3D, they have been told by Vectric not to address VCarve issues.

VCarve is off my computers, the money is back in my accounts, most of the pieces are in my neighbor's fire pit.

All I wanted was a tool to make stuff for kids to smile at. I now avoid it as much as I can.

Thanks,

Rex

[Tim Peachey]

Hi Rex,

Thanks for the detailed explanation of your experiences. Seems like there may be several issues experienced by various of us on these machines. Your details give me the idea to do some experimentation and see what happens. In the simple class at the member shop, my results were as expected, but we were following a very canned procedure. When I made my first attempt at an inlay, I used a vcarve/engrave shape rather than pockets, per Howard's 3-2-1 process video, though I did use a clearing bit. There was some behaviour I didn't expect, but being a novice I assumed that either I did something wrong or misunderstood something. It first looks like the vbit might have gone too deep, but when I measured, it seems as though the vbit went to the right depth, but the endmill I used for clearing seems to be too shallow. The engineer in me wants to do some experimentation to troubleshoot the issue further. So I will set up a simple experiment with various features on some cheap wood and see what happens. I'll report back on progress. I want to figure out what I can use reliably before doing long jobs on more expensive wood. I will test with outlines, pockets, and v-carve/engrave.

Variables: feature type, bit type, depth of cut, use/non-use of clearing bit, and pass depth

Checks: zero height before and after, depth of each pass, x/y extents, bit security

While I am planning this, are there any other variables anyone else would suggest?

I'll try taking some pictures and video to share if I find anything of interest to the group.

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/f24682a2-79c0-4ca4-80ed-8c00a10253c0n%40googlegroups.com.

[James Collard]

Tim,

I am a relative newb to CAD but I was wondering if the depth differences generated for a V-bit and an endmill would show up in the G-Code for cutting profiles generated out of V-Carve.  It would seem based on Rex's explanation that the Shapeoko is just doing what it is told to do.  So if you assume the Shapeoko is cutting accurately then the next issue to rule out would seem to be the actual instructions that it is receiving for the two different bits.

James

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/CAMWu08W1EYzz3pZ_iJaYKfz3OmbX_o1NSuanG2cgZdej4tBLJg%40mail.gmail.com.

[krum...@gmail.com]

It seems like there could possibly be a difference  between V-carve and carbide create in the way they create their G-code files for Shapeoko, so that would be another variable to throw in the hopper (V-carve vs CC). I was going to do that very comparison on my last test, but V-carve worked, so I didn't end up testing it.

Kurt

[Rex A Schildhouse]

Kurt, James, Tim, everyone else,

I cannot tell you that you can duplicate my VCarve issues on the shop's Shapeokos. Different software updates, different defaults, different computer, different firmware in the Shapeoko, etc. Too many variables for an easy solution.

First, I rest fairly comfortable that if you make two identical projects, one in VCarve, one in Carbide Create the resulting G-code would not be identical. The software writers use different code to accomplish their goals. I loved teaching code writing because I saw so many different ways to accomplish the objective and those student generated exposures expanded my capabilities. And yet, the vast majority ended up with the same visual presentation.

Some of the unique code lines I put in some of my programs won my legal battles for me.

Merely my thoughts, and I created the same name plaques in VCarve, before I removed it, and in Carbide Create. The issues are / were Carbide Create would not handle a chamfered by V-bit pocket. So, I compromised and used an end mill. And surprise surprise. The VCarve plaque was "gouged" at the borders and around the letters when Pocket and Clearing Bit options were utilized while the Carbide Create was good. Just not as pretty as a true VCarve pocket should have been.

To clarify, I had issues whenever Pocketing and a Clearing Bit was used for a V-bit pocket detail end result.

Comment from Vectric -

Feed speed is too fast causing the V-bit to plunge deeper. Rejected based on logic and trials. If the feed speed is too fast the bit will NOT try to climb OUT of the cut, nor would it dig into it, it is a straight fluted bit, not climb or dig angles. I changed the feed speed to be painfully slow. Increasing the project time 200%-300%. I use "rectangular" up cut and straight flutes on my routers. The different bits are interesting to watch when routing signs and such. Especially if routing gouges or "troughs."

Feed speed is too slow. Rejected based on logic and trials. If the bit is moving slowly through the project its probability of wanting to climb or dig due to pressure of cutting into the wood is totally illogical to me. Sorry, my technical background is aeronautical, a mix of just about every science known to man.

Router RPM speed is too fast. Rejected based on logic and trials. Too fast would most likely burn the wood as the friction of the cut is focused on a small area for a long period of time - revolutions of the bit. A router speed too fast would not cause a V-bit to dig in or to climb out, especially when the V-bit is a straight cut fluted bit. And most of my wood is softwoods, not hardwoods. I have not looked but I have not seen a spiral fluted V-bit.

Router speed is too slow. Rejected based on logic and trials. There is no chatter from the bit and the router is not burdened, the cut is VERY smooth, the Shapeoko is not kerchunking from excessive pressure on the stepper motors.

The depth of cut is too much. Rejected based on logic and trials. Dealing with clean and clear softwoods and taking 0.10 inch when the "Wood Calculator" says I can cut 0.185 inches, is going conservative, changing the default passes from two or three to FIVE and the bit is not chattering, the router is not bogging down, the Shapeoko is not jumping stepper motor steps, etc.

These are the standard "We didn't really read your e-mail, we didn't really look at your files, we really don't care, please stop e-mailing customer support at Vectric" responses.

Tried to open my shop class projects in my purchased VCarve and VCarve would NOT open them. Interesting. Disappointed.

In my case only, I disagree with the possibility that the V-bit shifted in the collet. As I learned, I set X , Y, and Z to zero with the Carbide Probe using an end mill, do that work. Then put the V-bit and set only the Z to the new zero using the Carbide Probe. We all know you cannot use the probe to set X, Y, and Z with a V-bit. Depth of the V-bit cut, verified to be something like 0.030 inches in VCarve and the resulting cut is something like 0.055 inches. Reset the V-bit cut depth to 0.00 inches and it gouges to something like 0.050. Mathematically inconsistent. The Carbide Probe is approximately 0.25 inches above the wood. Based on the first failure the second run should have been a cut of (0.050 - 0.025) 0.025. Not 0.050.

So, very frustrated. If you want to see some of my projects, go to www.schildhouse.com and scroll down to Woodworking. I have several requests for my CNC clocks. They are now created with Carbide Create to the same precision as VCarve did them. However, putting off using the Shapeoko, just annoying to have that much $$$ wrapped up and cannot get VCarve and the Shapeoko to agree and produce.

Thanks,

Rex

To view this discussion on the web visit https://groups.google.com/d/msgid/sdfwa-digital-tools/8ff13f52-6ca2-4ee4-be7b-f1792bbc0458n%40googlegroups.com.

 

Thanks,
Rex