4th axis wire....

[deeherk]

i'm fooling around with 4th axis and am wondering why when i do a backplot, i can see the u & v moves running .300 and more, but when doing my post, i only see .010 at most...

i'm putting a small clip of the post to show what i mean...

 

N164 G01 X0. Y.19008 U0. V-.00297

N166 G01 X0. Y.20009 U0. V-.00296

N168 G01 X0. Y.21009 U0. V-.00296

N170 G01 X0. Y.2201 U0. V-.00296

N172 G01 X.0003 Y.22981 U-.00031 V-.00266

N174 G01 X.01031 Y.22981 U-.01001 V.00705

N176 G01 X.02032 Y.22981 U-.01001 V.00704

N178 G01 X.03033 Y.22981 U-.01 V.00705

N180 G01 X.04034 Y.22981 U-.01001 V.00705

N182 G01 X.05035 Y.22981 U-.01001 V.00705

N184 G01 X.06036 Y.22981 U-.00593 V.00296

N186 G01 X.07036 Y.22981 U-.00296 V0.

N188 G01 X.08037 Y.22981 U-.00296 V0.

N190 G01 X.08039 Y.2398 U.00703 V-.00999

N192 G01 X.08039 Y.24981 U.00704 V-.01001

N194 G01 X.08039 Y.25981 U.00705 V-.01001

N196 G01 X.08039 Y.26982 U.00705 V-.01

N198 G01 X.08039 Y.27983 U.00705 V-.01001

N200 G01 X.08039 Y.28984 U.00704 V-.01001

N202 G01 X.08039 Y.29985 U.00705 V-.01001

 

any ideas why it is doing that...i changed post to incremental, but no help...i did a screenshot, but can't seem to get it on here..

any wire gurus out there?

dan

forgot to mention that i am using version 9.1

[mayday]

been a while since I did 4x wire but if i remember backplot outputs absolute distances and the post does inc. for u,v.

did you analyze the distances to verify the backplot u,v moves ?

[deeherk]

hey mayday,

we changed the post to incremental and the moves seemed smaller than what the backplot was showing just like in absolute...as far as analyzing, all i seem to get is the x and y dim.

[mayday]

quote:

---

all i seem to get is the x and y dim.

---

well, they sort of are if that makes sence. if the sync points are lined up viewing straight down on your model the x,y distance from top chain to bottom chain is the u,v in the machine. I wound up ignoring it and did some test cuts and everything was ok. I was proramming an FX10

[deeherk]

i guess i should start from the beginning....

we are switching to pro e and want to show that we would like 4 axis options in pro e as we did im mc..i fooled around with doing a TFS on top and EDM on bottom...i had to put letters together so i could chain them..i realize that a u or v movement of 1.00 is probably way over,which backplot shows...we are wondering if something is wrong with our post, because backplot and post don't look the same...it'ds hard to explain. and i don't want to sound like a dumb a--...

if i could show the screenshot, maybe it would help...

dan

[mayday]

well, can you post and dry run in the machine? you should see right away if it looks close to what your after.

sorry, not much help I learned not to try and analyze to much 4x code for wire, drives you

[deeherk]

i know what you mean..i'm kinda beating a dead horse...i tried to load on machine, but won't take it.. actually, this is probably something that is impossible to edm...just thought that the u,v moves in the post or wire path should match what the backplot shows...wondered why not??

thanks for your help...will have to investigate some more...

dan

[Lars Christensen]

deeherk,

Try two draw different size squares and chain those, that should bring your code to something a little easier to follow. My code will come post with "U" and "V" as incremental. And Backplot is absolute, just as Mayday said.

 

Lars

[deeherk]

thanks lars,

i drew a 1" circle and .8 x .8 square and in backplot, i have u,v moves of .4 and such...sure coincides with absolute...and in the post it shows u,v moves of only .001 and .00005, not necessarily in same position, but seems like awful small u,v moves..i guess i should really burn one and see how it comes out..this is a little more realistic than what i was trying earlier..will keep you all posted ..thanks alot..

dan

[3rdstring]

Just curious but how high is the part your programming? A .001 incr. move on the upper head just doesnt sound like enough.

[Lars Christensen]

It defently sound like something is wrong with your post, check your mail.

 

Lars

[Roger Martin from CNC Software]

deeherk ,

 

"very small U & V moves"

 

Are you by chance posting for a Mitsubishi EDM ?

[Kannon]

Roger,

I work with Deeherk, and yes the post is for a Mit. edm..

FX10K, SX10, FA10

[Guest]

my uv numbers in the post never match the machine

uv.the actual angle though is the same.

[Roger Martin from CNC Software]

Mitsubishi vs. "other" EDMs on calculating the the U&V.

 

The U&V coordinates on Mits machines are “incremental-incremental” (or "incremental * 2" if you prefer) when compared with most other machines. I hear you! "What @#$*@$#? is he babbling about ?!?" Read on...

 

That is why the U&V coords you see in your example NC code are so "small". (The U&V's are not always smaller, but usually are.)

 

The U&V positions in the NCI data, which is what you see in backplot, are the absolute coords of the U&V axis. For “most” other (non-Mits) machines you can look at those values reported by backplot and take the difference between the U and X = the U axis output and the difference between the V and Y = the V axis output to the NC file.

 

As previously mentioned, the Mitsubishi type machines the U & V are "incremental-incremental".

On the MITS, each the U&V is the incremental distance from the previous U&V position and the U&V are the shifted distance (aka. incremental) from the X,Y axis positions.

 

To demonstrate the difference in the U&V calculations for a Mits vs. “other” machines, I created a sample 4-axis program.

A 1” square (centered at X0,Y0) and at Z0

A 1.1” square (centered at X0,Y0) and at Z1

Break them both at the mid-point of the lower horizontal

Create a Thread Point at X0, Y-.4

Create a 4-axis wirepath program

(Chaining the Thread Pt, then the lower 1” square in CCW direction and then the upper 1.1” square in the same direction with Sync Mode set to By Entity)

When you backplot to where the wire is at the lower right corner, you’ll see coord. values of X0.5, Y-0.5 and U0.55, V-0.55

 

Post this wirepath using the std MPWFANUC.PST and you get ---

O0001

N100 G0 G20 G90

N102 G92 X0. Y-.4 I1. J0.

N104 G0 X0. Y-.4

N106 M60

N108 M83

N110 M81

N112 S101 D1

N114 G41 G1 Y-.5 V-.05

N116 X.5 U.05

N118 Y.5 V.05

N120 X-.5 U-.05

N122 Y-.5 V-.05

N124 X0. U0.

N126 G40 Y-.4 V0.

N128 M50

N130 M30

%

 

Post this wirepath using the MPW-FA10.PST and you get ---

%

L001

(03/01/06)

H1 =.006

N100 G90

N102 Z1 = 0.

N104 Z2 = .5

N106 Z5 = 1.

N108 G92 X0. Y-.4

N110 M20

N112 M80 M82 M84

N114 E101 H1

N116 M90

N118 G41 G01 X0. Y-.5 U0. V-.05

N120 G01 X.5 Y-.5 U.05 V0.

N122 G01 X.5 Y.5 U0. V.1

N124 G01 X-.5 Y.5 U-.1 V0.

N126 G01 X-.5 Y-.5 U0. V-.1

N128 G01 X0. Y-.5 U.05 V0.

N130 G40 G01 X0. Y-.4 U0. V.05

N132 M91

N134 M21

N136 M58

N138 M02

%

 

Now this really doesn’t demonstrate why the U&V’s you see in a Mits NC are usually “smaller”, (in fact it shows when they will be "larger"!), but it does show how the U&V calculation method is different.

On N118 of the MITS sample code you see the upper guide shift off (V-.05) from the XY plane. This being the incremental distance between the X axis and the V axis at the end of the lead-in to the part profile.

Looks just like N144 in the MPWFANUC output sample file. Now is where the difference kicks in. Where “most” non-Mits EDMs continue to use the “shifted distance “ between the XY and UV for the UV coords to be fed to the machine, the Mits does that plus each UV is incremental from the prior UV. Note in N122 of the MW-FA10 sample, the V.1 compared to the same move in the MPWFANUC sample, where in N118 it is V.05. The V.1 for the Mits is incremental from the previous V axis position. So from V-.05 (at the lower-right corner) to V0.5 (at the upper-right corner) = V.1 of motion.

 

That concludes today’s lesson.

[savagkd]

Great info Roger. Thanks for the skewlin.

 

Keith

[deeherk]

thanks roger,

i really needed that!!!

 

i'll have to fool around with it when i get to work....i'll post a segment of the post that i'm doing now with the circle and square..

maybe kannon has an answer!!!

[Kannon]

I dont have a answer,

but what I did was open your wire 9.1, drew a 1 inch circle at zero, then a 2x2 square at z-2.

Then broke the circle so the chain would start in both places at both entitys.

Here is what backplot shows when it just meets the entitys..

and here is a snip of gcode from your mits post...

L0

N100 G90

N102 G92 X0. Y0.

N104 G00 X0. Y0.

N106 M20

N108 M78M78

N110 M80

N112 M82

N114 M84

N116 G00 X0. Y0. U0. V0.

N118 G00 G42 H1 X.0101 Y0. U-.00505 V0.

N120 M90

N122 G01 X.0202 Y0. U-.00505 V0.

N124 G01 X.0303 Y0. U-.00505 V0.

N126 G01 X.0404 Y0. U-.00505 V0.

N128 G01 X.05051 Y0. U-.00505 V0.

N130 G01 X.06061 Y0. U-.00505 V0.

N132 G01 X.07071 Y0. U-.00505 V0.

N134 G01 X.08081 Y0. U-.00505 V0.

N136 G01 X.09091 Y0. U-.00505 V0.

N138 G01 X.10101 Y0. U-.00506 V0.

N140 G01 X.11111 Y0. U-.00505 V0.

N142 G01 X.12121 Y0. U-.00505 V0.

N144 G01 X.13131 Y0. U-.00505 V0.

N146 G01 X.14141 Y0. U-.00505 V0.

N148 G01 X.15152 Y0. U-.00505 V0.

N150 G01 X.16162 Y0. U-.00505 V0.

N152 G01 X.17172 Y0. U-.00505 V0.

N154 G01 X.18182 Y0. U-.00505 V0.

N156 G01 X.19192 Y0. U-.00505 V0.

N158 G01 X.20202 Y0. U-.00505 V0.

N160 G01 X.21212 Y0. U-.00505 V0.

N162 G01 X.22222 Y0. U-.00505 V0.

N164 G01 X.23232 Y0. U-.00505 V0.

N166 G01 X.24242 Y0. U-.00505 V0.

N168 G01 X.25253 Y0. U-.00505 V0.

N170 G01 X.26263 Y0. U-.00505 V0.

N172 G01 X.27273 Y0. U-.00505 V0.

N174 G01 X.28283 Y0. U-.00505 V0.

N176 G01 X.29293 Y0. U-.00505 V0.

N178 G01 X.30303 Y0. U-.00506 V0.

N180 G01 X.31313 Y0. U-.00505 V0.

N182 G01 X.32323 Y0. U-.00505 V0.

N184 G01 X.33333 Y0. U-.00505 V0.

N186 G01 X.34343 Y0. U-.00505 V0.

N188 G01 X.35354 Y0. U-.00505 V0.

N190 G01 X.36364 Y0. U-.00505 V0.

N192 G01 X.37374 Y0. U-.00505 V0.

N194 G01 X.38384 Y0. U-.00505 V0.

N196 G01 X.39394 Y0. U-.00505 V0.

N198 G01 X.40404 Y0. U-.00505 V0.

N200 G01 X.41414 Y0. U-.00505 V0.

N202 G01 X.42424 Y0. U-.00505 V0.

N204 G01 X.43434 Y0. U-.00505 V0.

N206 G01 X.44444 Y0. U-.00505 V0.

N208 G01 X.45455 Y0. U-.00505 V0.

N210 G01 X.46465 Y0. U-.00505 V0.

N212 G01 X.47475 Y0. U-.00505 V0.

N214 G01 X.48485 Y0. U-.00505 V0.

N216 G01 X.49495 Y0. U-.00505 V0.

N218 G01 X.50505 Y0. U-.00506 V0.

N220 G01 X.51515 Y0. U-.00505 V0.

N222 G01 X.52525 Y0. U-.00505 V0.

N224 G01 X.53535 Y0. U-.00505 V0.

N226 G01 X.54546 Y0. U-.00505 V0.

N228 G01 X.55556 Y0. U-.00505 V0.

N230 G01 X.56566 Y0. U-.00505 V0.

N232 G01 X.57576 Y0. U-.00505 V0.

N234 G01 X.58586 Y0. U-.00505 V0.

N236 G01 X.59596 Y0. U-.00505 V0.

N238 G01 X.60606 Y0. U-.00505 V0.

N240 G01 X.61616 Y0. U-.00505 V0.

N242 G01 X.62626 Y0. U-.00505 V0.

N244 G01 X.63636 Y0. U-.00505 V0.

N246 G01 X.64647 Y0. U-.00505 V0.

N248 G01 X.65657 Y0. U-.00505 V0.

N250 G01 X.66667 Y0. U-.00505 V0.

N252 G01 X.67677 Y0. U-.00505 V0.

N254 G01 X.68687 Y0. U-.00506 V0.

N256 G01 X.69697 Y0. U-.00505 V0.

N258 G01 X.70707 Y0. U-.00505 V0.

N260 G01 X.71717 Y0. U-.00505 V0.

N262 G01 X.72727 Y0. U-.00505 V0.

N264 G01 X.73737 Y0. U-.00505 V0.

N266 G01 X.74748 Y0. U-.00505 V0.

N268 G01 X.75758 Y0. U-.00505 V0.

N270 G01 X.76768 Y0. U-.00505 V0.

N272 G01 X.77778 Y0. U-.00505 V0.

N274 G01 X.78788 Y0. U-.00505 V0.

N276 G01 X.79798 Y0. U-.00505 V0.

N278 G01 X.80808 Y0. U-.00505 V0.

N280 G01 X.81818 Y0. U-.00505 V0.

N282 G01 X.82828 Y0. U-.00505 V0.

N284 G01 X.83838 Y0. U-.00505 V0.

N286 G01 X.84849 Y0. U-.00505 V0.

N288 G01 X.85859 Y0. U-.00505 V0.

N290 G01 X.86869 Y0. U-.00505 V0.

N292 G01 X.87879 Y0. U-.00505 V0.

N294 G01 X.88889 Y0. U-.00506 V0.

N296 G01 X.89899 Y0. U-.00505 V0.

N298 G01 X.90909 Y0. U-.00505 V0.

N300 G01 X.91919 Y0. U-.00505 V0.

N302 G01 X.92929 Y0. U-.00505 V0.

N304 G01 X.93939 Y0. U-.00505 V0.

N306 G01 X.9495 Y0. U-.00505 V0.

N308 G01 X.9596 Y0. U-.00505 V0.

N310 G01 X.9697 Y0. U-.00505 V0.

N312 G01 X.9798 Y0. U-.00505 V0.

N314 G01 X.9899 Y0. U-.00505 V0.

N316 G01 X1. Y0. U-.00505 V0.

N318 G01 X1. Y-.02548 U-.0001 V.01547

N320 G01 X1. Y-.05096 U-.0003 V.01548

N322 G01 X1. Y-.07643 U-.0005 V.01549

N324 G01 X1. Y-.10191 U-.0007 V.01549

N326 G01 X1. Y-.12739 U-.0009 V.01552

N328 G01 X1. Y-.15287 U-.0011 V.01553

N330 G01 X1. Y-.17834 U-.0013 V.01556

N332 G01 X1. Y-.20382 U-.00149 V.01558

N334 G01 X1. Y-.2293 U-.0017 V.01562

N336 G01 X1. Y-.25478 U-.00189 V.01565

N338 G01 X1. Y-.28026 U-.00208 V.0157

N340 G01 X1. Y-.30573 U-.00229 V.01573

N342 G01 X1. Y-.33121 U-.00247 V.01579

N344 G01 X1. Y-.35669 U-.00267 V.01583

[3rdstring]

If I understand Roger right that should be close.

If you add up all the U moves (which are incremental e.g. each time the U mooves -.00505 from where its at) it comes up to .5.

[Roger Martin from CNC Software]

Kannon's sample is without any 'Sync'.

Sometimes you can get away with this and end up with a good part - but rarely.

Here I broke the arc into 5 sections (since you have 5 linear segments on the bottom square) and used Sync: by Entity

 

 

%

L001

(03/01/06)

H1 =.006

N100 G90

N102 Z1 = 0.

N104 Z2 = 1.

N106 Z5 = 2.

N108 G92 X0. Y0.

N110 M20

N112 M80 M82 M84

N114 E101 H1

N116 M90

N118 G41 G01 X1. Y0. U-.5 V0.

N120 G01 X1. Y.14286 U-.00314 V-.08688

N122 G01 X1. Y.28571 U-.0094 V-.08757

N124 G01 X1. Y.42857 U-.01552 V-.08898

N126 G01 X1. Y.57143 U-.02146 V-.09106

N128 G01 X1. Y.71429 U-.02712 V-.09378

N130 G01 X1. Y.85714 U-.03244 V-.09713

N132 G01 X1. Y1. U-.03737 V-.10106

N134 G01 X.84615 Y1. U.10866 V.04004

N136 G01 X.69231 Y1. U.10415 V.0343

N138 G01 X.53846 Y1. U.10038 V.02806

N140 G01 X.38462 Y1. U.09739 V.02141

N142 G01 X.23077 Y1. U.09523 V.01445

N144 G01 X.07692 Y1. U.09391 V.00728

N146 G01 X-.07692 Y1. U.09346 V0.

N148 G01 X-.23077 Y1. U.09391 V-.00728

N150 G01 X-.38462 Y1. U.09522 V-.01445

N152 G01 X-.53846 Y1. U.09739 V-.02141

N154 G01 X-.69231 Y1. U.10039 V-.02806

N156 G01 X-.84615 Y1. U.10416 V-.0343

N158 G01 X-1. Y1. U.10866 V-.04004

N160 G01 X-1. Y.84615 U-.04004 V.10866

N162 G01 X-1. Y.69231 U-.0343 V.10415

N164 G01 X-1. Y.53846 U-.02806 V.10038

N166 G01 X-1. Y.38462 U-.02141 V.09739

N168 G01 X-1. Y.23077 U-.01445 V.09523

N170 G01 X-1. Y.07692 U-.00728 V.09391

N172 G01 X-1. Y-.07692 U0. V.09346

N174 G01 X-1. Y-.23077 U.00728 V.09391

N176 G01 X-1. Y-.38462 U.01445 V.09522

N178 G01 X-1. Y-.53846 U.02141 V.09739

N180 G01 X-1. Y-.69231 U.02806 V.10039

N182 G01 X-1. Y-.84615 U.0343 V.10416

N184 G01 X-1. Y-1. U.04004 V.10866

N186 G01 X-.84615 Y-1. U-.10866 V-.04004

N188 G01 X-.69231 Y-1. U-.10415 V-.0343

N190 G01 X-.53846 Y-1. U-.10038 V-.02806

N192 G01 X-.38462 Y-1. U-.09739 V-.02141

N194 G01 X-.23077 Y-1. U-.09523 V-.01445

N196 G01 X-.07692 Y-1. U-.09391 V-.00728

N198 G01 X.07692 Y-1. U-.09346 V0.

N200 G01 X.23077 Y-1. U-.09391 V.00728

N202 G01 X.38462 Y-1. U-.09522 V.01445

N204 G01 X.53846 Y-1. U-.09739 V.02141

N206 G01 X.69231 Y-1. U-.10039 V.02806

N208 G01 X.84615 Y-1. U-.10416 V.0343

N210 G01 X1. Y-1. U-.10866 V.04004

N212 G01 X1. Y-.85714 U.03737 V-.10105

N214 G01 X1. Y-.71429 U.03244 V-.09712

N216 G01 X1. Y-.57143 U.02712 V-.09378

N218 G01 X1. Y-.42857 U.02146 V-.09106

N220 G01 X1. Y-.28571 U.01552 V-.08898

N222 G01 X1. Y-.14286 U.0094 V-.08758

N224 G01 X1. Y0. U.00314 V-.08688

N226 G40 G01 X0. Y0. U.5 V0.

N228 M91

N230 M21

N232 M58

N234 M02

%

[deeherk]

looks good roger...we sent program to sx10 wire and graphed it up...we watched the u,v numbers and they looked right, just, diff. than post...

i guess if it walks like a duck, and sounds like a duck. i want it to look like a duck too!!!

i guess we did not sync it either...never messed too much with that...would like to burn something out once and see how it turns out...

thanks for your help....

 

update....

i did a quick sync like you did, 5 segments and wow!!

the post looks great, like it should...so, is it necessary to do a sync of some sort every time you do 4 axis...could be a good lesson learned...

thanks teach!!!! and thanks everyone else...

[Roger Martin from CNC Software]

Once you get a couple 4-axis jobs under your belt, you'll see just how easy the are to do.

 

Yes, assume that you always need to Sync.

The methods used 98% of the time are either; 'By Entity' or 'By Branch'.

Any job that can be done 'By Entity' can also be done using the 'By Branch' method.

The reverse not necessarily true.

 

To get your head around what is ‘Sync Pt’ is ….

It is a moment in the program where the wire must be “here” on the XY plane profile and “here” on the UV plane profile.

I’ve uploaded 4-axis EDM Sample.MC9 to the forum FTP in the MC9_files folder.

Both operations use the same set of chains.

But… OP#1 has NO Sync and OP#2 is Sync’d By Entity

The branch lines (connectors between the bottom and top profiles) are not really needed is in case, but are there to assist is seeing what the finish shape needs to be.

Plotting each operation will show “Why you need to Sync”.

If you do not give the system any Sync (position) information the entire profile becomes a single sync. In this case all the system has to work with is the total length of the profiles and since one is longer than the other...

You can see what happens to what are supposed to be “well defined” features on this part.

 

Sync’ing

By Entity = the system makes sure that the wire is at the end of the entity on the XY plane at the same time the wire is at the end of the corresponding entity on the UV plane. The system just keeps doing this for each entity “set” along the chains.

So of course you must have the same number of entities in each chain.

 

It is not always possible to have the same number of entities in each chain, so that is when you use Sync - By Branch.

 

By Branch = the system makes sure that the wire hits each “branch” condition position in the chains simultaneously. The system just keeps doing this for each “set” of branch points along the chains.

So of course you must have the same number of branch conditions in each chain.

[savagkd]

deeherk and kannon,

 

Check your mail.

[Kannon]

Thanks savagkd for the email, and Roger, could you please email the file unzipped, our firewalls dont allow us to get to the ftp.

Thanks guys,

Brian... aka Kannon

[deeherk]

thanks guys,

savagkd, i replied to e-mail, thanks again.

 

roger, i got the mc9 file from kannon, and i see the difference..makes alot of sense too...like i told savagkd, before we did our own drawings, product design did them...so, any 4th axis work we still do, we already have the program, so no need to go to indepth into this, but now that doors are open, it is very interesting..will be pursuing this matter, for sure!!!

thanks..

dan

oh, and for kannon..