Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: tinkerer360 on November 06, 2006, 04:29:31 PM
-
Hi All,
I'm new to this so please bear with me.....I have tuned my motor using the steps per setting and am able to get consistant measurement for the value I used for tuning the motors, i.e. 4inches, but any other distance will be slightly off. whats going on, how do I find out if I'm losing steps, what is the screw mapping function and where is it explained in the manual, is screw mapping standard procedure for "tuning motors?
-Here is what i'm using to set up my mill;
circular "pocket" designed in software
ID 4.015
OD 4.287
after milling I get
ID 4.018+/-0.007
OD 4.330+/-0.007
these numbers are pretty consistant if I run multiple parts
I think the ID is set perfect but the OD is off, I guess my question is what other setting need to be made to get the correct distance value for any distance.
any help understanding what I need to do will be appreciated.
thx, eric
-
What kind of machine do you have? is it ball screws?
Have you used the step per unit button on the settings page to make sure your steps per unit are good?
Hood
-
If your machine is sound, I wouldn't think screw mapping would be necessary. How much backlash do you have on each axis?
-
Try Sherline mode under the ports and pins page. You will need to exit the software to have it take effect.
Thanks
Brian
-
thx for the fast reply guys.....
My unit is a ballscrew driven cnc router, I am using a xylotex board and Mach3 software.
As for setting the values I did that manually...basically I kept on adjusting the value untill the circle's ID matched (using calipers) what I drew in the software, i.e. ID 4.015". At that point I assumed that I was done calibrating/tuning the motor... the number came out to be ~6400 for both the x and y axis
I haven't used the auto steps per unit botton, I will try that tonight and see if my number and the number mach3 spits out are the same...i'll also try to put the unit in sherline mode and see what happens.....
As for backlash I'm not sure how I check the value of this.......I do notice some backlash thought I assume its about 4-7mils, 0.004-0.007", but I'm not sure....any good websites on how a non-pro can learn how to accurately figure out the backlash of an axis?????
thxs a bunch guys, I'll work on it tonight and report my findings...
eric
-
Once you have the steps per unit pretty good you can then jog against a dial gauge, zero the axis you are moving then change to step mode and back off, when the needle just starts to move look at the DRO for the axis and the amount you see is more or less your backlash.
Hood
-
Hello:
I am using Mach2 with a bridgeport shopmaster unit to engrave, well trying to!!!
I tunned the stepper motors fine I think, and I am getting readable text, but somehow it looks like I am loosing steps, the letter "O" for instance, the cutting tool would have to finish right where it started right, well, it does't quite do so, I have been playing with different units/min different minimum pulse width, and still getting the same result, any ideas???
-
Have you checked your backlash?
Hey Hood, thats no fair. I just figured out why I can't get up early enough to get one on you. You are 5 hrs ahead of me. Guess I'm a Geographically challenged fool. ;D
-
Sounds like backlash, do a check for backlash with a dial gauge.
Hood
Edit
too late in my reply :) Now what is puzzling me is the saying great minds think alike, cant be right especially in my case.
Oh OK I got it, should be fools seldom differ ;)
-
well I read your promt reviews, (i am surprised by the length of time that took you to answer). I went to my mill and found a little backlash. I tightened a screw and that almost eliminated the problem, my text almost looks perfect now, I probably have to look into the whole chain of elements to eliminate this completely....
question, shoouldn't there be a setting for backlash in the software itself, I mean, when you are doing things manually you always adjust for this yourself, just a thought.....maybe there is a setting and I don't know!!!!!
thanks guys, appreciate the promt response on the previous question!!!!!
-
There is a setting for backlash but for it to work properly on some machines you need to be in exact stop mode. Best thing is to eliminate the backlash, comp really isnt the best solution. Just my opinion ;)
Hood
-
Oh, he got me, he got me. I've been bamboozled. It's a set up. The old look, over there trick.
Just so everyone will know, that is exactly what I was going to say,........ only better. ;D
-
and slower ;)
-
guys I have a Dayton Router
(READ>>>> home made by some Chinaman on 20Cents per day making his wages on the parts he didn't fit into my machine - Sorry for the tears [for further background check out the CNCZONE forum and search for Dayton routers ] )
Anyway onto my question... I am trying to cut this file out (see attached) but instead of a smooth constant motion i get some smooth tool paths and some jerky cresent type movement and a lot of start stop or as I like to say SLAM/START- BANG/STOP. I and using 300" steppers on a rack and pinion drive the motors are set to (see attached file) the feed is mm/min(G94) and not per REV (G95) I have tried reducing the feed rate fromm 4600mm/Min (which is ok for Xmas cut outs) to 2400mm/Min but I still get the jerky overshoot cutting in some areas. ... Oh the material is an extreamly hard Polycarbonate. If this has any bearing on the whole situation ???
-
Hi Salty,
First things first, the code you have here will give you the shutters I think. What cam package did you use to generate the code? It looks to me like a bunch of short point to point moves, no arcs. Do you have an arc fit vectors option in your cam package?
Brett
EDIT
I didn't see a G64 to put you in CV mode either.
-
HEY guys it's me again,
I fixed my backlash problem, and things are working out now, but something extremely weird just happened,
The x axis suddenly stopped moving, see how if you disconnect the parallel port the computer still thinks it's connected and keeps on trying to move the axis, well it doesn't move the x axis, but it moves the other two noramlally, I don't get it, someone HELPPPPPPP
-
never mind, restarted the comp and it fixed itself, I still don't get it though!!!!!
-
Hi Salty,
This may sound daft but drop the feed rate down to 200mm/min and see how that cuts, it should cut sweet.
Chances are it will cut it just as quick.
Graham.
-
Brett,
Thanks for that I had to have a look to see what G64 (CV was) and I will try this to night, the program I'm using is MC9, you might have noticed that the file is a 3D shape and Mach3 originally didn't like changing from G17 (x-y cutting plane ) to a G18 (X-Z cutting plane) it kept claiming the start of arc is different to end of arc in previous line, even when I-J incremental mode was changed. (if you want the DXF or MC9 file I will post)
So I re-drew the shape and made it a spline instead of a chain. then re-analysed the tool path and this is the result. (note: the mold that the sunnies sit on thru which I suck the vacuum was cut using the same point to point style of programming [cutter was 3/8" ball mill with 8% overlap at 4600mm/min ] the only difference between then and now is that I have changes spindles from a Ryobi 2Hp router to a VFD 1.4Kw high speed spindle - I wonder if the spindle frequency is interferring with the steppers??)
Graham
I did notice that cutting at 4600mm/min took 1M43S to cut and at 2600mm/min only took 3Sec -- so to get this clear in my own head (this could take a while ;D ) ....is you theory that the router won't try to accelerate to full speed at the same time as it's trying to slow down??
ONE OTHER POINT OF NOTE... Last night I tried to get a video of the cutter in action, but for me to set up and start filming I reduced the speed using the FEED RATE OVERRIDE and upon dropping the speed to 20% of 4600mm/min, it did as advertised --cut slower.. but it also kept the chunky action at the end on each line or direction change... (that sudden start stop).
I wonder if its worth trying to reduce the acceleration and/or max speed to ensure a softer start and more constant run speed...... but then wont Brett's G64 do the same thing, Find out the answer to these and other intriguing questions in next weeks exciting episode of the day in the life of a Dayton Owner VTIC
-
Hi Salty,
I would like to see the DXF. I may be leading you astray with the G64. That will smooth a point to point program however, Graham is one of the G code Gurus. ;D I'm certainly not. If this program ran to your satisfaction before I would think it should now.
the only difference between then and now is that I have changes spindles from a Ryobi 2Hp router to a VFD 1.4Kw high speed spindle - I wonder if the spindle frequency is interferring with the steppers??)
Hmmmmm, this is interesting. How are you driving your spindle or sending the varying voltage to the VFD? I had another member call me yesterday with some interesting problems with his VFD.
Brett
-
well now I'm annoude ( read Very upset!!!) i have just spen the last ten mins writting a responce setting it up and adding files then I hit thew submit button to find you can't post MC9 files and Oh you need to re-type all the crap you have previousely said.
well here is the DXF file
If you nee more info e-mail me salty72(AT)tsn(DOT)cc
Also attached is the VFD spindle I'm using and it's controller (it's the 1.5Kw 220Vinput spindle)
-
Maybe pressing the back button on your browser would have got you to all your previous typings. You can zip the file before attaching it as the forum accepts zip files.
Hood
-
I don't know what hte hell is going on here every so oftern my broser updates the page and I loos all the stuff I have just typed, I really thought I was over my Dummy spit, but I am starting to see red again
-This reminds me of the video clip of the bloke who looses it and bashes his monitor with the keyboard and then starts laying into the computer, I thought it was funny at the time but am beginning to wonder where I can get another keyboard to bash the living daylights out of this Lappy
So I'll post know B4 loosing it again and continue in a Moment
-
Ok back again!!
the MC9 file is done using layers [auto save]
so if using MC you will be able to view all the different layers etc... [auto save again]
!!!*** guys as this is a customers file [read friend of the family] ...can you do the right thing and ask B4 using it yourself***!!!!!
Oh lucky this time I had it saved ( I have no idea what the hell is going on here !!)
sorry back on track
the file is faily basic there is a surface that has a curve in both the X-Z plane and the Y-Z plane. the X-Z is on a radius of about 132mm and the Y-Z is around the 32mm Radius
at Z=0 there is a 2D chain of a pair of sunglasses, (well quite a few if you consider the 5 different layers )
the 2D chain is then projected to the surface this creates and a splined curve, this CORD is not projected perpendicullar to Z=0 as this would mean the [auto Save]
machined sunnies would infact be longer than the 2D chained image. but instead it ahs been wrapped (if this make sence)
anyway here is the Zipped Mc9 file. if anyone has any Q's pls just ask
-
Hi Salty72,
My theory is this :-
based on the code you posted the longest single move in the file is approx 0.9mm, if your router is running at even 20% of 2600mm/min this is still 520mm/min how on earth can your axis drives possibly get any where near that speed and back to 0 in .9mm moves.
If you add up the total distance to travel its only 250mm so at 200mm/min its only going to take 75 seconds.
So 1min 43 secs at 4600mm/min ain't right.
I have 2 machines on the same job 1 rapids at 40 meters/min the other rapids at 10 meters/min over short moves the 10 metre machine is faster.
Graham.
-
I got the 1M43S from the clock that runs in Mach3 when you hit the START button, so it mighit include the movement from X0,Y0, Z30 to the first point of the lead-in the moved to Z-4 the 6mm arc for the lead-in the cutting of hte job the 3mm overlap, the 6mm arc for the overlap, the move from Z-4 to Z30, the return and the rewind of the program..
Don't get me wrong I like the way you worked out hte 200mm/Min feed rate this might be a very usefull thing to remember for the futer especially when making small moves as apossed to long flowing arcs etc...
-
Having looked closer at your code the profile is dodgy, see picture
Graham.
-
sorry Graham, I sent you the file that has 4 half height tabs -these tabs were going to be used to ensure that the sunnies stay attached to the rest of the material but as it turned out with a vacuum of 20"Hg and blue tack sealing it there is no longer a need to use tabs
-
I hate to say it but this one is worse
Graham.
-
Mate, help me out here >>>>
this is the correct one ( i even checked it B4 posting this time)
salty
-
I think it rhymes with Clucking Bell. :D
Its 2:30 am here, I will be back tomorrow, email me all your latest files.
Graham.
-
thanks Graham, BTW it 1:30 here (Oh that would be PM ----Sorry, me so silly,..... LOL )
-
Hi Salty72,
sorry could not get back sooner, I had a busy one yesterday.
Try this new and improved version (sounds like washing powder ad).
Graham.
-
graham,
Thanks for the NEW improved File, (it washes clothes well !) Seriously I have just downloaded it so haven't had a chance to try it out...
One thing I did notice was that every second line had Xnn, Ynn, R1.5 and then the next line had the Xnn Ynn Znn. will this not create a slight scollop shape befor the drop in Z-height? (I also added the G94 [feed per mm] as I have no feed back to the spindle for a G95 [feed per RREV] )
What was the reasoning for adding the R1.5..?
Over the week end I was having a think (couldn't do much else as I fell through the Ceiling onto the CNC table and have hurt my back aslo brused my pride rather surverly) if I can set the Zheight then carry out a Xnn Ynn Inn Jnn G-code would this give me a smoother cutting file, I realise that the Zhieght would remain constant but because of the shape I would either be cutting less or more material depending on where about on the shape I was cutting??
-
I converted the whole geometry into a 3d polyline to get rid of the lumps and bumps.
Try it as it is and let me know what happens.
If you need to rough it out first just off set the Z axis up by about 24mm and then run a cut, then keep dropping it by 3mm or so and re run again. keep doing this until you get back to the proper depth.
Graham.
-
the file with the R1.5 give little scollups around the edge of the sunnies, I noticed that the stop start motion was worse in one direction than the other,
I looked closer, to see that if I applied pressure to the gantry the stop/start action was remarkably reduced, further investigation showed the stepper was loose the vibration of cutting with the 55mm Cutter and the larger router must be loosening the bolts (the stepper is mounted upside down and is suspended by 4 hex bolts I have since added LOCTITE and will hope this problem is fully resolved,
feed and speed are set to 400mm/Sec with 6000RPM on a 2.5mm Standard two flute end mill.
I will post a video tonight for all to see...
thanks for your help...
Ps how do you tell if the speed it to fast or that the cutter is the wrong style
-
You can calculate the spindle speed using this formula :-
In imperial measuments.
Alum, Brass, Copper, Plastics = 200ft/min
Mild steel = 100ft/min
Alloy steel = 80ft/min
Stainless Steel = 40ft/min
Titanium = 30ft/min
High tensile alloy steels = 15ft/min
RPM = S(ft/min)
--------------------------
0.2618* Cutter Dia
So in plastic with a 1/8 cutter
RPM = 200
----------------
0.2618*.125
RPM = 6116
Graham.