Shop OBEX P1 Docs P2 Docs Learn Events
Community supported PCB milling machine - Page 14 — Parallax Forums

Community supported PCB milling machine

1111214161719

Comments

  • John KleinbauerJohn Kleinbauer Posts: 3
    edited 2005-12-08 14:30
    When measuring runout on the Dremel Multipro measure it on the inside lip of the shaft with the collet removed. You will find no runout. Then put your collet back it. Put in a piece of drill rod and lightly tighten the collet and check the run out. Tighten the collet tighter. You will now see runout. Over tightening the collet will distort it causing runout. Hundreds of people are using Dremels. I would also like to say I sell plans for two mills that will mill and drill PCBoards. The Brute (uses window channel) and the Hawk which uses drill rod. Both machines uses Gibs and Jibs made out of Delrin and not bushings. Gibs and Jibs can be adjusted so there is no play just like on a lathe or real mill. Questions?

    John

    www.kleinbauer.com
  • John KleinbauerJohn Kleinbauer Posts: 3
    edited 2005-12-25 13:24
    Three boards milled with dull bit. From Eagle BRD to GCode using a ULP Outline

    Three boards Milled with Brute and Dremel using the
    same dull bit


    http://www.crankorgan.com/threeboards.JPG


    Boards Drilled-Sanded and cut

    http://www.crankorgan.com/perfchop7.JPG


    Parts added
    http://www.crankorgan.com/perfchop9.JPG

    Post Edited (John Kleinbauer) : 12/25/2005 2:35:25 PM GMT
  • NewzedNewzed Posts: 2,503
    edited 2005-12-25 15:12
    John, what size cutter did you use?

    Sid
  • John KleinbauerJohn Kleinbauer Posts: 3
    edited 2005-12-25 18:42
    I used a 60 degree bit from Think & Tinker. The deeper it mills the wider the isolation.
  • Kaos KiddKaos Kidd Posts: 614
    edited 2006-01-26 15:58
    Has this project died?
    I'd be interested in it if it weren't...

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

    Just tossing my two bits worth into the bit bucket
    KK
    ·
  • steve_bsteve_b Posts: 1,563
    edited 2006-01-26 17:16
    You're the first post in a month.

    I wouldn't way its dead....but certainly on hiatus!

    I think the lead guys are too busy with their other jobs!

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    ·

    Steve

    "Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
  • stamptrolstamptrol Posts: 1,731
    edited 2006-01-26 17:31
    ··· I ran across this page while looking at the MAKE webzine ( which is worth a look in its own right ). The attached website has numerous milling machine plans which may be of interest to the thousands who are interested in this project.

    ····· www.crankorgan.com

    ·· Good luck. If any of you have access to an old New Hermes V6000 lamecoid engraver, it has the power to do boards and can read CAD drawings directly.

    · Cheers
  • BeanBean Posts: 8,129
    edited 2006-01-27 03:27
    Y'all might find this interesting... http://www.ares-server.com/Ares/Ares.asp?MerchantID=RET01229&Action=Catalog&Type=Product&ID=82573
    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "SX-Video·Module" Now available from Parallax for only $28.95

    http://www.parallax.com/detail.asp?product_id=30012

    "SX-Video OSD module" Now available from Parallax for only·$49.95
    http://www.parallax.com/detail.asp?product_id=30015

    Product web site: www.sxvm.com

    "Ability may get you to the top, but it takes character to keep you there."
    ·
  • PLJackPLJack Posts: 398
    edited 2006-01-28 02:17
    Kaos Kidd said...
    Has this project died?
    I'd be interested in it if it weren't...

    Well, lets put it this way, we still eagerly click every time there is a new post in this thread.

    In my case, Steve_b is right, I'm very busy at the moment.
    I still want a SX milling machine on my desk, but alas, not yet.

    I must say that I am off the idea of milling PC boards.
    But using the same machine to pass a bright light over a photo-treated PC board does turn me on.
    Same machine, same software. Sounds like a winner to me.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    - - - PLJack - - -



    Perfection in design is not achieved when there is nothing left to add.
    It is achieved when there is nothing left to take away.
  • allieallie Posts: 109
    edited 2006-02-19 08:21
    Hi, this sounds like a really good idea. I hope it grows as I already have a sherline milling machine and would try to use your control hardware and software. I was hoping some one would come up with an idea for milling circuits boards. I'm new to micro computer programming and don't have much to offer. I"ll be checking in to see how this project goes. I,m currently designing an interface board to hook my sherline controller's to my commodore 64 and the basic stamp 2. I'll keep in mind that some how I might be able to use your interface too.

    regards
    allie
  • allieallie Posts: 109
    edited 2006-02-19 08:23
    Hi, this sounds like a really good idea. I hope it grows as I already have a sherline milling machine and would try to use your control hardware and software. I was hoping some one would come up with an idea for milling circuits boards. I'm new to micro computer programming and don't have much to offer. I"ll be checking in to see how this project goes. I,m currently designing an interface board to hook my sherline controller's to my commodore 64 and the basic stamp 2. I'll keep in mind that some how I might be able to use your interface too.

    regards
    allie
  • NewzedNewzed Posts: 2,503
    edited 2006-02-19 17:12
    Good to see some life in this thread.

    My little Super Mill is still doing great.· I've taken to writing complete etching programs for all the boards I am making.· Takes a bit of time to write the program, but when it's done the mill runs unattended until the board is finished.· I suppse I could also write a program to drill all the holes but so far it hasn't been worth the effort.· I have a subroutine that will drill holes for a DIP, vertically or horizontal, any width from .100 up to the limit of my board size.· If I want to drill just one row of holes, the routine gives me an option to continue or return to Start after the first row is drilled, so I just return to Start.· Works out pretty good.· Odd placed holes require the mill to be manually positioned for each hole and then I hit a key to drill the hole, ·but that isn't really a problem.

    Hope to see some more activity here. yeah.gif

    Sid
  • Kaos KiddKaos Kidd Posts: 614
    edited 2006-02-20 21:02
    I checked out the web site bean, as well as the one posted by stamptrol.. interesting...
    I did further lookup and research on it...
    for now, until I get my laptop 100% up and running, and I get some of my other little things done, it's on hold for me...

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

    Just tossing my two bits worth into the bit bucket
    KK
    ·
  • NewzedNewzed Posts: 2,503
    edited 2006-03-22 15:24
    There has really been a dearth of activity on this thread - I hope no one is giving up.

    Finally managed to make a double sided board on Super Mill.· I have attached pictures of the front and back, which are certinly not the highest quality.· Difficult to take good pictures of a shiny copper board.

    Anyway, it didn't turn out too bad.· You will will note that some of the isolation paths run into the hole pads - my mistake was in trying to go between the pins of a DIP.· Also, there are a couple of pads that are spaced .050 on a horizontal or vertical axis.· Even though I'm using an .015 router, it still takes away part of the hole pad.· I'll have to use .100 min spacing between pads and traces in the future.· All the other holes seem to be right where they should be.

    Etching was done via program rather then manually punching in each step.· That was a job - 192 data entries for the bottom and 203 data entries for the top.

    We are making progress. [noparse]:)[/noparse])

    Sid
    450 x 266 - 7K
    450 x 270 - 10K
  • PLJackPLJack Posts: 398
    edited 2006-03-22 22:52
    They just keep looking better Sid.

    "Etching was done via program....."

    What program did you use?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    - - - PLJack - - -



    Perfection in design is not achieved when there is nothing left to add.
    It is achieved when there is nothing left to take away.
  • NewzedNewzed Posts: 2,503
    edited 2006-03-22 23:26
    I wrote my own program, which reads a DATA list.· Here is the heart of the
    program:

    eePntr = Pattern1
    · GOSUB Make_Moves
    · GOSUB advise

    · DEBUG· cr, "Board finished.· Press any key to return to Start", cr
    · DEBUGIN com
    · x = 2
    · RUN 0

    ·Make_Moves:
    · READ eePntr, numMoves······ ' get number of moves from table header
    · eePntr = eePntr + 1············ ' point to first record
    · DO WHILE (numMoves > 0)··· ' loop through number of moves

    ··· READ eePntr, Word f, theMove· ' retrieve distance and direction from table
    ··· LOOKDOWN theMove, [noparse][[/noparse]"IOLR"], theMove··· ' convert letter command to value, 0..3
    ··· ON theMove GOSUB goin, goout, goleft, goright·· ' call the move code

    ···· eePntr = eePntr + 3···························· ' point to next record
    ··· DEBUG "eePntr = ", DEC eePntr, ":·· X = ", DEC posx, ":··· Y = ", DEC posy, cr, cr
    ······· numMoves = numMoves - 1···················· ' update moves count
    · LOOP
    · RETURN

    Then I have to write a DATA statement for each etch operation.· Here
    is a bit of my DATA table:

    Pattern1··· DATA··· 203··············· 'number of etch operations
    ··········· DATA··· WORD 0, "L"······· '1
    ··········· DATA··· WORD 1000, "L"····'4··········
    ··········· DATA··· WORD 900, "L"···· '7
    ··········· DATA··· WORD 1200, "I"··· '10
    ··········· DATA··· WORD 1000, "R"··· '13
    ··········· DATA··· WORD 900, "R"···· '16
    ··········· DATA··· WORD 1200, "O"··· '19

    The numbers on the right are the eePntr values.· Takes me quite awhile to write the DATA statements and, invariably, I make an error and have
    to debug the program.· I run the entire program on my office PC - the mill is out in the shop - and if the ending X-Y positions aren't what they are supposed to be I know I have an error.· With a new program I may have 10 or 11 "checkpoints" to see if the program is OK so far.

    Did I answer all your questions?

    Sid
  • NewzedNewzed Posts: 2,503
    edited 2006-03-26 23:09
    And I thought I was so smart with my little double-sided board.· I was sitting at my workbench admiring it when all of a sudden I realized that some of the traces going to the pins were on the component side could not be soldered since the holes were not PTH.· So........I redid the layout for the whole board, rewrote the programs, and now I'm ready to try again.· I'll post new pictures when I get it finished.· Oh, the agony of it
    all. cry.gif

    Sid
  • PLJackPLJack Posts: 398
    edited 2006-03-28 00:30
    Oh, by program you mean stamp code.

    Very nice.

    So you enter your board layout with the stamp IDE?
    That's resourceful!

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    - - - PLJack - - -



    Perfection in design is not achieved when there is nothing left to add.
    It is achieved when there is nothing left to take away.
  • NewzedNewzed Posts: 2,503
    edited 2006-03-28 00:42
    All the etching operations are done with a BS2E driving the X and Y steppers.· Etch operations are entered as DATA statements.· The Stamp reads the DATA statements and then executes the instructions contained in each DATA statement.· Takes a while to write the DATA statements and debug them, but once it's done, the BS2E runs the mill runs on its own from start to finish.· And I got a proven program if I want to make the board again.

    Sid
  • Kaos KiddKaos Kidd Posts: 614
    edited 2006-03-28 15:42
    NewzEd:
    Just a thought on creating the data statements... use PAINT, and draw the circuit in monochrome.
    A program could read the .bmp and generate the data statements from there.
    I once (many years ago) used an old etch-a-schetch, fankin-morfed to my PC's joystick and a QBasic program to read and output the data needed.
    That idea, along with some features in the software could make programming the system better.

    Just an Idea...

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Just tossing my two bits worth into the bit bucket


    KK
    ·
  • NewzedNewzed Posts: 2,503
    edited 2006-03-28 17:56
    How is this possible?· First, Paint does not have a .100 grid, which is essential.· Second, all I want to program is the etch lines, not the circuit.

    Sid
  • Kaos KiddKaos Kidd Posts: 614
    edited 2006-03-29 16:11
    It's all in the scale you want.
    For example: In it's simplest form, one pixl could = .100 X .100 dot, or one pixle = .050 X .050 dot (so you could put holds in the etch for through holes)
    Draw the dots (lines etc) where you want...
    Next, the program would follow the lines. (it's like the maze solver programs...)
    It's somewhat simple to read the bitmap, and the logic is stright foward...
    Start at the top left of the image...
    ..scan until you found a set bit.
    ..follow that series of set bits (they will always be one bit next to it that is set) until no further bits are found.
    ..continue scanning from where you left off
    there's more logic to include, but the general idea is to make a maze solver that will know when it's traced a path...
    and continues to look for additional paths...
    Using this logic, and knowing the max capabilities of your mill, you could reduce the gid size if the hardware was capable of doing it.
    Remember, in programming, it's all in how you interpet things, not how things really are.
    Keeping things in a relitive scale makes the software easier to develope as well...
    Another idea is after you get the reader working, creating a paint program that was designed for this would make the entire process even easier.
    Making a paint program is another simple program...


    YOu would scan the image until you found a set bit...
    then the logic would follow that series of bo set bits until no others were found...
    Next, go back to the top and scan

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Just tossing my two bits worth into the bit bucket


    KK
    ·
  • kelvin jameskelvin james Posts: 531
    edited 2006-03-30 07:18
    Sid

    Glad to see someone is still pursuing the DIY cnc. I may give 'er a shot when the propeller gets out, looks good for controlling 3 motors at the same time. Anyway, on the programming issue, what values are you using in the data? Is it in the form of measurement being translated to steps, or just step values that you pre-determine? Reason i ask is because, a gcode compiler gives all the info needed for the milling. If the circuit is done in pcb software, and saved as a dxf, it can be imported to a gcode program, and it basically spits out the moves and values needed for the milling. The idea from here would be to use the gcode text to import into the stamp data. The G moves themselves are pretty basic, and could be translated in your program to motor movement. The most important thing is the move values in the code, 2 lines of code for a straight line, a starting xy point in the first line, and an ending xy point in the second. I was thinking you may be able to structure the stamp program around this info, and possibly do a copy / paste from the gcode output into the data area. Or even get the stamp to read the data info from your computer. It is just an idea, and i will do some playing around to see if it can work. Just thought it would be a lot easier then manually entering all the data.

    kelvin
  • NewzedNewzed Posts: 2,503
    edited 2006-03-30 13:41
    Kelvin, I am interested in anything that will eliminate the edious and error-prone process of entering DATA statements to control the mill.· Here is an example of the DATA statements I use:

    Pattern1··· DATA··· 191
    ··········· DATA··· WORD 0, "L"
    ··········· DATA··· WORD 1000, "L"··· '4
    ··········· DATA··· WORD 1000, "L"··· '7
    ··········· DATA··· WORD 300, "L"···· '10
    ··········· DATA··· WORD 1300, "I"··· '13
    ··········· DATA··· WORD 1000, "R"··· '16

    I use Expresspcb, and if I draw the etch lines as a silkscreen layer then I can view them in a .dxf file viewer.· How do I convert this .dxf file to something that will spit out the direction and distance as·DATA statements, or is this possible?

    I will take all the help you can give me.

    Sid
  • NewzedNewzed Posts: 2,503
    edited 2006-03-30 21:49
    Kelvin, I Googledaround and found all the programs needed to convert a .dxf file to gcode.· I haven't figured out exactly what I am getting, but it is academic at this point because my system won't run on gcode and I don't know how to make it do that.· I drew a simple square, 1000 x 1000 in Expresspcb, exported the .dxf file and converted it to gcode.· Here is what I got:

    attachment.php?attachmentid=73575

    I don't know where the first three lines came from and it appears the next 4 lines are duplicated by the last 4 lines.· Don't understand gcode, I guess.

    Anyway, you gave me an idea - I always do the etch instructions in Excel to check them out, so I modified by Excel spreadsheet to read:

    attachment.php?attachmentid=73576

    Now when I get it all done I can just copy and paste it into my Stamp file - that will save a tremendous amount of work.· BTW, the " '4· ", etc.,·after the cap letter is the EEPROM pointer, which I use at various checkpoints·to see if the mill cutter is where it is supposed to be.

    So........I thank you for bringing up the subject.

    Sid
    342 x 226 - 3K
    294 x 193 - 5K
  • ktekxktekx Posts: 71
    edited 2006-03-31 04:38
    Hi, Newzed, the first few G commands might be reference points, to zero out the point of origin or something and to measure the length/width of the material being worked on, if I recall correctly. This is really cool, I am thinking of making a miniature PCB router this summer. Have HPGL been considered? It seems pretty straightforward and since PCB etching is 2D, it would be simliar to a plotter I would imagine. Sorry if I'm completely off, I'm new to the whole CNC and software related.
  • kelvin jameskelvin james Posts: 531
    edited 2006-03-31 07:34
    Sid

    I am pretty busy right now, so bear with me, it will take me a few days to put together a detailed plan.
    The gcodes are a non-issue, they only apply to the gcode software that is controlling the mill from a computer. The idea is the mill plot sequence that is compiled automatically for you from the dxf. Now, it is just getting the stamp to translate it. Good for the copy / paste.

    kelvin
  • BeanBean Posts: 8,129
    edited 2006-03-31 12:06
    I may be totally off-base here, but what about a simplifed version of LOGO ?
    On my old (and I mean old) Timex Sinclair 1000 they had a version of LOGO that was written in basic.
    I think what Sid has in excel is pretty close to LOGO already.
    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "SX-Video·Module"·available from Parallax for only $28.95 http://www.parallax.com/detail.asp?product_id=30012

    "SX-Video OSD module"·available from Parallax for only·$49.95 http://www.parallax.com/detail.asp?product_id=30015
    Product web site: www.sxvm.com

    Available now! Cheap 4-digit LED display with driver IC·www.hc4led.com

    "I reject your reality, and substitute my own." Mythbusters
    ·
  • NewzedNewzed Posts: 2,503
    edited 2006-04-01 15:09
    Bean, I'm not familiar with LOGO - maybe I can find out what it is with Google.

    Kelvin, here is where I stand to date.

    I created an etch pattern in Expresspcb using the silkscreen layer to draw the etch pattern.· Exported it as a .dxf file, which I converted to a .txt file using ACE Converter 3.10.· Then I opened the .txt file in Excel.· The Express .dxf file generates X and Y positions as 1.2, .8, and so on, whereas the gcode uses mils for the X and Y positioins.· So I had to multiply all the X-Y data by 1000 to get a format that would be acceptable in gcode.· When that was all done, I copied and pasted the Excel data into the gcode window of CNCSimulator.· I then ran the program in CNCSimulator and it drew a perfect duplicate of what I had laid out in Express.· So now I have a good gcode.· The next problem is to convert the gcode into DATA statements that my BS2E can use to drive the steppers.· I'm going to try to manipulate the Excel file so it can just be copied and pasted into my Stamp program, but I really don't anticipate much success.· I have attached the gcode so that maybe you can come up with something when you have the time.

    Sid
  • kelvin jameskelvin james Posts: 531
    edited 2006-04-02 06:54
    Sid

    Still trying to put a program flow together and find some better compatible software , in between some chores here. The data area looks promising, as it can store / read text and numerical. The stamp program may turn out to be fairly intensive to interperet the gcode, but it only has to do that ' before ' a move, and not during, so no big hit in execution. Programming will basically read the data line by line, decode the gcodes to stamp controlled motor operations, and mathematically change the xy coordinate values to step values in the proper direction. Some gcodes like G00 ( rapid move ) and G01 ( move at programmed feed rate ) are necessary, others can be omitted from the gcode program, same with the mcodes. I think it is worth pursuing, it will be a major bonus if it works.
    I don't know what kind of hardware you are using to drive the motors, so i am just taking the basic approach right now.

    kelvin

    P.S.
    Since this is a " community project", if anyone wants to step in here that has some experience or ideas, you are more than welcome.
Sign In or Register to comment.