rolling nest 4.62 errors

Discuss Thermwood 3-axis Machinery, Controller, and Software.

Moderators: Jason Susnjara, Larry Epplin, Clint Buechlein, Rich Kasten

Jeff Norris
Guru Member
Posts: 321
Joined: Fri, May 20 2005, 1:44AM

rolling nest 4.62 errors

Postby Jeff Norris » Mon, Aug 08 2005, 9:46AM

i ran this job in 4.62 last week and got some bad parts. similar to some that i got a few versions ago. the problim namley is tennons being cut improperly. the cutter over shooting parts into other parts. i seem to recall that this revolved around blind dados. the nests gave no indication that this was going to happen. i have now upgraded to 4.63 and im a little gun shy. all of our cabinets use blind dados with adjustments to the tennons. if the nest gives no indication there will be bad parts how else can i verify this. i have attached the twd. thanks

User avatar
Jason Susnjara
Thermwood Team
Posts: 1722
Joined: Tue, May 10 2005, 1:26PM
Location: Thermwood
Contact:

Postby Jason Susnjara » Mon, Aug 08 2005, 11:00AM

Hi Jeff,

Can you email me the eCabinet job, the twd file, and the cnc code for this job so I can take a look at it? If you can also sketch this out on paper or software and send that to me as well, it would help.

My email is cabinets@thermwood.com

thanks,

Bill OConnell
Senior Member
Posts: 121
Joined: Mon, Jun 13 2005, 12:26PM
Location: Calverton, New York
Contact:

Postby Bill OConnell » Wed, Aug 17 2005, 3:17PM

I may be experiencing a similar problem. We are using blind dado construction and up until recently all of our joints were dead on. We ran, what was for us, a big job on Saturday and it seems that our tops are down 1/16\". Before running the job I mic'd a couple of sheets and changed the dimensions to reflect these measurements. However where we have sides that go full length the top is down 1/16\" from flush, however if you measure the bottom edge of the top piece it is .75\" from the top edge of the side (this is for 3/4\" material). The outside dimensions of the boxes are all exactly where they should be, which indicates that part of the dimension change was handled properly, but somehow the dadoes are off. We are on v4.63 of rolling nest. I can't seem to attach the files but I can e-mail them to whomever wishes.

The actual dimensions of the material are as follows:
3/4 Prefinished Ply - 0.72
1/2 Prefinished Ply - 0.485
3/4 Unfinished Ply - 0.73
1/2 Unfinished Ply - .45
Woodworking the Way You Want It

Jeff Norris
Guru Member
Posts: 321
Joined: Fri, May 20 2005, 1:44AM

Postby Jeff Norris » Wed, Aug 17 2005, 5:54PM

yea bill i almost posted the same thing yesterday but i ran out of time, when we went to assemble the cabinets from the same job that i posted about previously i found that all our joints were off by .04, terribly loose for my shop. of course i usally blame myself first, but upon inspection i found that all my tooling was set perfectly, my daylight values and everything else were correct. i opened the twd file and my adjusted material thickness were correct. but my joints are the worst yet, ive used these files before 4.62 and they worked fine. now im definatly getting gun shy

Bill OConnell
Senior Member
Posts: 121
Joined: Mon, Jun 13 2005, 12:26PM
Location: Calverton, New York
Contact:

Postby Bill OConnell » Wed, Aug 17 2005, 8:53PM

Jeff,

I was previously on 4.57 and everything worked fine. I reported a problem I had when I added a sheet that measured 1/2\" wider than the material in the database. I think it shifts the X and Y axis off of the database values rather than the added sheet dimensions. The net of it was the router ran 1/2\" further off the table than usual and clipped the aluminum 902 fence. :shock:

It was suggested that I upgrade to 4.63 to fix that problem, which it didn't. As I previously mentioned we ran a large job (18 sheets) and now I have someone hand routing 1/16\" off all of the cabinets.

I will be implementing some change control procedures that if we are on a stable version we don't upgrade without a compelling reason. And if that reason doesn't pan out, like in this example, roll back to the version that we were on before. I will likely run a test tomorrow checking the fit on a cabinet, rolling back to 4.57 and re-running the exact same cabinet to check the fit.
Woodworking the Way You Want It

User avatar
Jason Susnjara
Thermwood Team
Posts: 1722
Joined: Tue, May 10 2005, 1:26PM
Location: Thermwood
Contact:

Postby Jason Susnjara » Thu, Aug 18 2005, 9:49AM

Hi Guys,

I talked to the programmer that handles control nesting and said nothing has been changed in that area. The more the info the better on trying to fix this problem.

There are a few important settings that you are probably aware of are:

1. Make sure that you have the correct diameter set in the tool manager of the tooling being used in your application.

2. Make sure that your sheet material is the correct thickness. If not, you will see a slight offset between two parts instead of being flush.

3. Dados will overextend by the radius of the tool.


Jeff,

Can you email me a few more files from the controller?


Located in C:/Rolling Nest folder
machine.ini
tools.thm
Located in C:/system folder
toolman.thm

Bill,

I think what you might be seeing is a common problem of sheet material thickness being off. Plywood is notorius for not being .75 or .5 but maybe a little thinner. I have seen sheets where the sheet is off by at least .04 within the sheet. This will cause for a small offset where two parts come together using blind dado. When the machine clipped off part of the fence, was it cutting an offall piece? If your material thickness is off, the depth of the dado will change as well causing tennons to touch the bottom of the dado.

thanks,

Bill OConnell
Senior Member
Posts: 121
Joined: Mon, Jun 13 2005, 12:26PM
Location: Calverton, New York
Contact:

Postby Bill OConnell » Thu, Aug 18 2005, 10:17AM

1. Make sure that you have the correct diameter set in the tool manager of the tooling being used in your application.


Our tooling has basically remained the same before and after the problem. One tool was changed, but I just remeasured both and the difference is 0.004 inches. Not enough to make a 1/16\" difference in the fit.

2. Make sure that your sheet material is the correct thickness. If not, you will see a slight offset between two parts instead of being flush.


We learned this the hard way a while back. We now take micrometer readings typically of three sheets in three places and use those values as we prepare the twd files to be sent to the machine. I provide our measured values earlier in the post. The outside dimensions of our cabinets are CORRECT, so I know that rolling nest is making the proper adjustments there for the material thickness. However the bottom surface of one part is 0.75\" from the edge of the joining part while the top surface is not flush but set in. I would think that rolling nest would move the part up by the difference in thickness of the actual from the nominal.

3. Dados will overextend by the radius of the tool.
We don't have an issue with the length of the dado or the fit along that line.

:?: Is there a procedure to roll back to a previous version of Rolling Nest? I would like to make up a sample cabinet, cut it of the same material on v4.63 and then again on v4.57 and see if there is a difference in the fit.
Woodworking the Way You Want It

User avatar
Jason Susnjara
Thermwood Team
Posts: 1722
Joined: Tue, May 10 2005, 1:26PM
Location: Thermwood
Contact:

Postby Jason Susnjara » Thu, Aug 18 2005, 10:25AM

Hi Bill,

Here is the link for V4.57

http://order.ecabinetsystems.com/update ... te4-57.exe

Install this as you would with a newer update.

Bill OConnell
Senior Member
Posts: 121
Joined: Mon, Jun 13 2005, 12:26PM
Location: Calverton, New York
Contact:

Postby Bill OConnell » Thu, Aug 18 2005, 2:06PM

Back to head scratching time. :?

I just created a test cabinet 3/4\" sides, deck and top, 1/2\" back. It is basically an 18\" cube without a front. My hypothesis was that this would have the top slightly down from the sides and the deck slightly up from the sides. I would then roll back to 4.57 and re-run the job with the same sheets of material and the joints would all be flush.

I ran the job using 4.63 and the joints were all dead on. No point, I figured in wasting more material in rolling back to 4.57 and running the same job. My new hypothesis is that perhaps after changing the material thicknesses, somehow they didn't take. Either human error in not saving it properly, or something in the translation. Is there an easy way to check the CNC code to see what thickness it thinks the material is? Would that appear in the ZSHIFT command? If so, I can check this job and the one I ran on Saturday.
Woodworking the Way You Want It

User avatar
Jason Susnjara
Thermwood Team
Posts: 1722
Joined: Tue, May 10 2005, 1:26PM
Location: Thermwood
Contact:

Postby Jason Susnjara » Thu, Aug 18 2005, 2:31PM

Hi Bill,

You can look at the Z shift for the thickness of the sheet. I still think that even if you use a micrometer to measure the thickness of the sheet that the sheet might change thickness from one side to the other. There is still a possibility that there is a bug with changing material thickness in ecab and RN not recongnizing it. I know that when I cut my kitchen out that I had some cabinets where the top and sides meet were not flush and others that were. If I measured the thickness of the two cabinets, the thickness between the two were different. When deciding which material thickness to use, I took a few sheets and measured all four corners of each sheet and then took the average. The only way that I can think of that will fix the problem is to machine both sides of the tenon, but that will lead to a lot of flip parts.

Bill OConnell
Senior Member
Posts: 121
Joined: Mon, Jun 13 2005, 12:26PM
Location: Calverton, New York
Contact:

Postby Bill OConnell » Thu, Aug 18 2005, 2:58PM

I just checked the test box that I ran. For the ZSHIFT values it had the same values I keyed into change the thickness. I went back and checked the job that I ran on Saturday and all the 3/4\" pieces had a ZSHIFT of 0.75 and all the 1/2\" pieces had a ZSHIFT value of 0.5, this is despite having two measured values for the 3/4\" stock (0.73 and 0.72) and two measured values for the 1/2\" stock (0.45 and 0.485). Could I have somehow put the values in and somehow wiped out my changes before they took? I suppose that could happen, however, before we learned to check and update our thicknesses, our cabinets were coming out the wrong size on the outside dimensions. In this case the outside dimensions were correct, which suggests that rolling nest was making adjustments for some operations but not others.
Woodworking the Way You Want It

Jeff Norris
Guru Member
Posts: 321
Joined: Fri, May 20 2005, 1:44AM

Postby Jeff Norris » Thu, Aug 18 2005, 3:49PM

sorry guys im out of the office today but i would like to add that the first thing i checked was my adjusted material thickness. i reopened the twd file in rolling nest and my thickenss' were correct. i have gotten quited good at finding an average thickenss. it is very easy to not save the changes when working in ecabinets so im extra cautious of this. i have noticed some funny behavior when adding sheets. there has been several times when i added sheets of diffrent thickness' and RN nested parts 3/4\" parts on 1/2\" . this does not always happen but i seen it several times. now i did not run the program to find out if it was merely the graphic on the screen reporting this incorrectly. i have also noticed when seleting material to add that if you dont make sure reselect the same material after adding a sheet of the same material that this produces some funny behavior like nesting parts on the wrong sheet or just not using the sheet at all. ill send you the files tomorrow jason im out today, i would really like to be sure there is a fix for this, a whole set of cabinets that were going to have to use construction adhesive to assemble.


Return to “Thermwood 3-Axis Machinery”

Who is online

Users browsing this forum: Jeremy Schiffer, Stuart Douglas, Yahoo [Bot] and 9 guests