Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
- CAD-CAM-EDM-DRO
- Messages
Search
|
Re: AUTOMATIC TOOL CHANGER
Brian Fairey
Dan, did you find a 2deg stepper? if not I have some.
toggle quoted message
Show quoted text
Now you have solved the tool changer for your lathe design one for the verticle mill???????? Brian, Ont, Can. ps what plastic parts do you want to make? I have a Plasticor from Simplomatic in Chicago. Dan Mauch wrote: From: "Dan Mauch" <dmauch@...> |
|
Re: AUTOMATIC TOOL CHANGER
Dan Mauch
I am currently working on two new projects.
1. An automatic tool changer for my CNC lathe. 2. An injection molding machine to make plastic parts. On the tool changer I bought a heavy duty tail stock turret attachment for my JET 13X25 lathe. It has 6 posistions. I plan to drive it directly with a nema 34 stepper motor rated at 150 oz inches. The tool changer will be for those pesty jobs where you have to center drill, step drill, step drill, ream and chamfer. The key was to find a 2 degree per step stepper motor . This makes dividing the 360/6 easy and it comes out to be an even number. I am figuring that with everything on center that there should be little forces trying to make the turret rotate axiallly for the lined up posistion. The radial forces are dealt with within the turret attachment. Projects finished. Adding a servo system to my mill drill Solving the mill/drill quill backlash problem. DRO BTW the DRO boards are on order and the scheduled date is 6-24. Dan |
|
Re: Printed Circuit Boards.
Dan Mauch
That would be great about putting the converter on your web site. I dreaded
toggle quoted message
Show quoted text
having to convert excellon drill files to G code and add the Z up and down. Even cut and pastig is inefficient.. Especially when your have 400 holes to drill. BTW where is your source for that dry film resist. I have been thinking abouut picking up a laminator for several reasons. Dan -----Original Message-----
From: Jon Elson <jmelson@...> To: CAD_CAM_EDM_DRO@... <CAD_CAM_EDM_DRO@...> Date: Sunday, June 06, 1999 5:31 PM Subject: Re: [CAD_CAM_EDM_DRO] Printed Circuit Boards. From: Jon Elson <jmelson@...> |
|
Re: plastic strips for encoders
Dan Mauch
I caled some time ago and US digital Tech support said it had to do with HP
toggle quoted message
Show quoted text
not making a encoder head that would work. This sounds dumb but thats what they told me. Dan -----Original Message-----
From: Mark Ehle <mehle@...> To: CAD_CAM_EDM_DRO@... <CAD_CAM_EDM_DRO@...> Date: Sunday, June 06, 1999 9:26 AM Subject: Re: [CAD_CAM_EDM_DRO] plastic strips for encoders From: "Mark Ehle" <mehle@...>be a better unit than 200 or 360!discussion of shop built systems in the above catagories. |
|
Machine geometry
Per the recent request for lurkers to de-lurk:
I recently received some linear components from Bill Scalione, and now have to decide what kind of machine to build with them. I have two sets of 1" dia. rails, 54" and 66" long. Also two tables, nominally 22" square with two open bearings at each corner along one edge, and one open bearing centered along the opposite edge. These tables weigh about 30 lbs apiece. My question is what geometry will be best for a (mostly) wood router? I like the greater working envelope that a moving-Y gantry system would provide, but this requires a longer leadscrew and probably heavier motor. The moving-X table system, while easier to build, and probably more accurate, provides only half the travel. Any suggestions appreciated. Geoff Roehm |
|
Re: EMC & Linux
This may not be a practical solution for your radius problems, I'm just
learning about this stepper motor cnc stuff myself, but how about adding a rotary table to your program "Z"? Axis? Tracey By the way I'll probably be asking a lot of dumb questions in the future...please bear with me thanx |
|
Re: Digest Number 37
Andrew Werby
From: WAnliker@...[What a good idea! I second the motion. I've been trying to follow this discussion, but this would help put it in perspective. I think there's a market for a CD that would install realtime linux and a compatible version of EMC which would just work, without all the headaches so exquisitely detailed in some of the foregoing posts. I'm sure there would be enough challenges for most of us in just making parts...] Andrew Werby Andrew Werby - United Artworks Sculpture, Jewelry, and Other Art Stuff |
|
Re: Linux vs. DOS
Jon Elson
Tim Goldstein wrote:
From: "Tim Goldstein" <timg@...>It really doesn't matter. this is the ABSOLUTE position. You generally never work in absolute positions, except maybe when calibrating permanent fixtures onto the machine. Normally, you work in relative coordinates, ie. relative to some edge of the part. So, you put an edge finder in the spindle, and jog the part over until it deflects the edge finder. let's say we're doing the X axis, and the part is to the right of the spindle, and we have a .2" diameter edge finder. So, when it deflects, we enter by MDI, G92 X0.1 This sets the relative coordinate offset such that the relative coordinate system is aligned with the left edge of the part. (The X0.1 is to compensate for the radius of the .2" end of the edge finder.) You can now do this with the mouse (I think it's a right click over the axis in question, and you get a little menu.) There is also the G10 L2 Pn X1.23 Y4.56 command, which sets up to 9 work coordinate offsets from the main relative coordinates. Very handy if you have multiple parts mounted in a fixture. So, the only time you should see absolute coordinates is after poweringFrom my limited home shop type experience this is a MAJOR pain for settingup a switch for the Y axis (X in lathe terms) of a lathe. up, and before initializing anything. (You can use it later to check for lost steps, by re-homing.) The relative offsets from the absolute home position are stored on graceful shutdowns, so you could come back the next day and power up, home and switch to the relative coords, and you'd be right at the same place. (The G10 L2 work coord offsets are saved, too.) The # key (I think) switches the display between relative and absolute. The @ key switches between commanded position and actual position. I keep it in commanded because the flickering digits are distracting. Jon |
|
Re: EMC & Linux
Jon Elson
Tim Goldstein wrote:
From: "Tim Goldstein" <timg@...>OK, I'll take a look at it and see how badly it looks as a .gif or whatever. You won't believe how scrambled the original printed document was. If the henscratched corrections are not readable after scanning, it will be useless. Jon |
|
Re: Linux vs. DOS
Jon Elson
Tim Goldstein wrote:
From: "Tim Goldstein" <timg@...>Yeah, but it has to match to within .00000001" or so! Which is crazy, since my machine only has resolution of .000025" on the finest axis, and probably most people don't even have that much. Jon |
|
Re: Linux vs. DOS
Matt Shaver
The one thing I am disappointed in the home function is that the positionof the switch seems to always correspond to 0.00 for the axis. In looking realat exactly 0.00. All I had to do was home up one and then take a cut on aGee, I've never ran a cnc lathe before. I would think that you would set the switches up to home wherever it's convenient and then use G92Xnn.nnnn, or manually right click on the X axis display (or press ALT-X for the X axis) and enter the value you need to shift the zero point to where you want. How are you doing this now in the mill? Like when you use an edge finder to find the corner of your workpiece. I'm guessing that you could line up your tool tip to the center of rotation by chucking up an edge finder and feeding the tool into it. When it goes off center, you are half the edge finder diameter away. Zero the axis, remove the edge finder, jog in that amount (.1" if you use a .2" edge finder) and zero again. Then face off the stock (or use a gage block to establish a precise distance) and zero the other axis. The problem comes in with tool offsets in a multi-tool job. The tool table in the EMC is linked to the Z axis for length (H) and X/Y for the radius (D) and this will need changing to really get lathes going good. We need to be able to freely associate [AXIS_n] with the desired axis letter (or make a quick change wiring setup for you as the changing axis letter assignment problem is specific to the 3-in-1's unusual geometry), and take into account that there's no Y to worry about with respect to tool radius compensation. Until then I guess you write single tool programs substituting the X mill axis for what should be Z on a lathe, and Y on the mill for the lathes' X. Matt P.S. Soon I'll bet you'll want a spindle encoder for threading! Isn't there an LED thing or something on a Shoptask that takes the place of the threading dial? |
|
Re: plastic strips for encoders
On Sun, 6 Jun 1999 12:27:54 -0400, "Mark Ehle" <mehle@...> wrote:
I have been in contact with them, too, about a 250 line/inch scale and theyIt's really quite simple. I chased this waskelwy wabbit down a few months back, meself, so hopefully I know of what I speak. The SENSORs that US Digital relies on, namely those made by HP, are NOT made in a resolution that will accomodate 250 lines/inch, at present. NO bloody reason by HP couldn't make some at that res, but they don't, and although I DID talk to the Product Marketing Mgr. of this sensor line, for some reason I couldn't seem to convince him that us hobbyists should be considered in their overall product scheme. B) Bottom line is that the quadrature sensors with their arrays of leds & sensors, are NOT made by HP at the moment in such a resolution to allow US Digital to make the films/gratings work, and make us all happy. You CAN use gratings of slightly varying resolutions from the nominals that the sensors are rated at, but apparently 250 *IS* too far off from 200 to just whip out a 250lpi grate, and get it to work properly with a 200lpi sensor from HP. Sad, but apparently we're toast on that hope. Gar |
|
Re: Linux vs. DOS
Tim Goldstein
I hooked up my home switches for the lathe and tried the function out. Over
toggle quoted message
Show quoted text
all it works just as Jon described. HOMING_POLARITY controls the direction the axis moves when you hit the home button and HOME_SWITCH_POLARITY takes care of whether you switches are NO or NC. I found that if you set the HOME_SWITCH_POLARITY wrong what happens is the program functions just like it does with no home switches attached. Click home and it zeros out the selected axis without any table movement. The one thing I am disappointed in the home function is that the position of the switch seems to always correspond to 0.00 for the axis. In looking real quickly I didn't see a way to indicate home as being a different position. From my limited home shop type experience this is a MAJOR pain for setting up a switch for the Y axis (X in lathe terms) of a lathe. With a mill all movements are relative and if you move your home point .100 other than having to move your tooling you will still get parts that are the correct dimension. On a lathe the 0.00 point on the Y (X) axis has to be dead on to the centerline of rotation or you get off size parts. In DeskNC I was able to tell the program what coordinate I wanted home to be. This way I didn't have to try and adjust the position of the switch to get it so it clicked at exactly 0.00. All I had to do was home up one and then take a cut on a piece, measure it and enter 1/2 of the amount I was off as the coordinate for the home switch. Matt, any chance you can ask Fred if we can get the ability to set the coordinate we want to be home in the ini file? Tim [Denver, CO] -----Original Message----- |
|
Re: EMC & Linux
Tim Goldstein
Jon,
toggle quoted message
Show quoted text
If you want to scan it I have an FTP server available to me that we could put the images on. Tim [Denver, CO] -----Original Message----- |
|
Re: Linux vs. DOS
Matt Shaver
What might be good is some sort of "arcAt the risk of replying to myself I think that such a program would need to check arcs and for erroneous ones in IJ format it could suggest solutions that move the center point keeping either the start or end radius (user's choice) or furnish the minimum radius value that will connect the start and end points for commands in the R format. I doubt anyone wants to change the start or end points, and if they do, the number of correct answers is infinite! (I suppose one solution would be to propose solutions for moving these points along an imaginary line that connects them, or only in one axis, but the utility of this in the real world is questionable) Matt |
|
Re: Linux vs. DOS
Matt Shaver
From: "Tim Goldstein" <timg@...>for the code my CAM program is generating R is the better way to go. Here is anI drew this in CAD and it's an 1/8" arc, but I think there must be some teeny, tiny error at maybe the .0000000000000x level. I'll ask Fred how we can verify code produced by CAM systems and what we need to do to reconcile the difference between theory and reality. The up side of this is that these errors are pretty rare (at least for me using the system I've got) although they do crop up from time to time. What might be good is some sort of "arc validator" program that would check G2 & G3 moves for correctness and allow you to change the start, end, or center point to fix problems (probably the center is the most useful). Also in a previous post I said: "Since switching to IJ format there haven't been any problems since for any two points (start and end) and a specified center point, an arc of some radius exists!" Please ignore this obviously untrue (and stupid) statement on my behalf! I think what I meant was that using the IJ format there is only one possible arc between the start and end points, but the distances from the center of the arc to the start and end points have to be equal! Matt |
|
Re: EMC & Linux
"Mo" <[email protected]
From: "Tim Goldstein" <timg@...>Hi Tim, From my understanding it should be possible to use more than one parallel port. Mo |
|
Re: Linux vs. DOS
Tim Goldstein
Matt & Jon,
Thanks very much for the explanation of what I, J, and K mean in circular moves. I think I see the light and hope it is not a train!! Had a little bit of a chance to play with using I & J vs. R and at least for the code my CAM program is generating R is the better way to go. Here is an example of the code that is causing errors for me: N14 G01 X0.155190 Y-2.331463 N15 X5.844810 Y-2.331463 N16 G03 X5.875000 Y-2.250000 I5.750000 J-2.250000 N17 G01 X5.875000 Y-2.237713 Line 16 is the one that is blowing up. The error message says "arc_data_ijk:error 116: radius to end of arc differs from radius to start of arc" If I plot this out using my new found understanding of I & J it appears to me to be the same radius. Any ideas? Tim [Denver, CO] |
|
Re: Printed Circuit Boards.
Jon Elson
Dan Mauch wrote:
From: "Dan Mauch" <dmauch@...>I generate plot files and drill files from PCB CAD software. I used to use Accel's Tango, but I'm using Protel's Adv Pcb for windows, and using Protel 98 at work. Each has pluses and minuses. But, I can't imagine designing complex boards on a mechanical CAD system. In fact, both the 'Gerber' files for photoplots, and the 'Excellon' drill files ARE 'G-code'! Yes, they are ALL variants of RS-274D! It really isn't that hard to write converter programs that chop the commands into pieces and make the changes necessary to go from one format to the other. Both the excellon and Gerber formats are fixed point, where the decimal point is not shown. Some of these are leading zero suppressed, some are trailing-zero suppressed. That is the biggest difference. Also, the drill file just has to name a position and a hole is drilled there, as it sets a canned cycle and G00 positioning mode. As for PC boards, I got a steal on a dry film laminator, so I now coat my own boards with DuPont Riston film resist. It works VERY well. I don't know of any other system that works anywhere near as well as this stuff. I used to buy my boards pre-coated from Kepro with this resist. I hadn't thought about making my 'Excellon' drill file to RS-274D public, but I will check to make sure there aren't any annoying bugs that require editing the output file by hand, and put it on my web site. It has some rather tricky features, such as you set the machine so that with the drill point just touching the top of the board, that is Z=0.0. Then, it computes the correct plunge to make the specified size drill's flank come completely through the bottom of the board. So, it automatically computes the length of the drill's point. Jon |
|
Re: Linux vs. DOS
Jon Elson
Dan Falck wrote:
From: Dan Falck <dfalck@...>I don't have my home or limit switches hooked up, either. But, I went through this with Fred (or maybe Matt) some time ago, and think I understand it. The home switch must be set so it closes when the machine is between the limit switches. It makes sense to put the home switch near one of the limits, for the following reason. In the .ini file, you define which way each axis is to move to go home. So, you put the machine in a position on the 'right' side of the home switch manually, then hit the home button while that axis is highlighted (selected) on the screen. The machine moves in the direction specified in the .ini file until the home switch closes. It then advances slowly until the encoder index pulse is seen. The position marked by both the home switch being closed and the next index pulse is 'Home' for that axis. This puts it very close to the same place every time, probably within one or two encoder counts. Jon |
to navigate to use esc to dismiss