Since i have only worked with big machines, and mostly milling aluminum, i guess some of my knowledge does not apply to a pcb. I suppose cutting fluid would be a problem here, but maybe a small air nozzle by the mill tip to blow away shaving might help to prolong the life of the bit. I understand the expense of the bits, some i use at work are 50 bucks a piece, and if there is a coolant problem, it will be toast in about 5 seconds. Anyway, i read for pcb drilling that a layer of material should be used on top of the board to get a cleaner cut. I am thinking it may secure the board better, thus reducing the vibration / pulling of the copper when being drilled. Now, what material that would be, i don't know yet, but should not be hard to find out. Steves' idea of moving the bit around in the hole is good, the side of the mill bit will actually deburr the hole some amount. A circular pattern would be perfect if it could be done.
Ha! I knew I would be of some use at some point in this project!
For the problem with the burr's on Sid's machine, try layering the whole PCB with Duct tape on the bottom of the board before running the mill's drill operation.
I know it sounds weird, but try it. It helps with woodworking and "chip out", and I've heard it works well with metal too.
Knight.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-
This message transmitted with 100% recycled electrons.
-=-=-=-=-=-
Gravity doesn't exist. The Earth sucks.
-=-=-=-=-=-
Make a man a fire, and he will be warm for the night.
Light the man on fire, and he will be warm for the rest of his life.
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-
Kelvin, the material you lay on top of the PCB is called "entry board".· It is very thin laminated phenolic, about .030 thick.· I used to work for a company that sold materials to the PCB industry and we sold tons of the stuff, as well as backup board, which is a 1/8" thick masonite type of material.
I have no problems with the holes - my drill goes in rather slow compared to the speed of an Excellon machine.· The test hole pattern I drilled yesterday produced no burrs at all.· They all come from the etching process.· I usually deburr the boards with 400 grit wet-dry but it would be nice to be able to do it with the mill.
Knight, thanks for the duct tape suggestion but putting it on the bottom of the board that would accomplish nothing that I can see since I usually make only single-sided boards.· I use a piece of 3/8" oak for backup and normally the bottom is burr-free unless I happen to hit a hole that was drilled before.· I flip the backup over when it gets full of holes.
If anyone has a little single-sided board they need, I would be happy to make it for them.· Max size would be 3 to 4 inches left-to-right and 1.6 inches top-to-bottom, and it would have to be laid out in Expresspcb.· Minimum hole size would be .035 and I can go up to .125 if I have the drill bit you want.· If the board has an SMD chip it will have to be an SOIC .050 pitch.· SSOPs I can't handle.
I'd like to see if I can make a board that would be acceptable to someone besides me, even though I consider myself very picky.
I don't know if you have a solution for this, but thought i would bring the subject up.
I was looking into the idea of getting some output from the gcode to do a circle. Unless someone knows, or there is a resource available for the math to accomplish this, it looks like it would be futile to bother with. And even if there was, i don't know if the stamp could handle everything for it to run properly. But, i did find something interesting. Arcs and eclipses are handled differently by gcode, because there is no diameter input for the gcode to translate it to the I and J commands, as normally output with a closed circle. So, to make a long story short, it decodes the line to a series of small straight lines. This way the mill would still be able to accept and run the same data format to mill a circle. The pcb software i am playing with can place an eclipse, and then stretch it into a circle. Another way would be to have 2 or 4 arcs made into a circle. Either way, when changed to gcode, it outputs the xy moves. Obviously, it would make a big circle a little rough looking, but a small one would probably be okay. The xy axis would have to be run at the same time, like doing a diagonal cut.
By the way, i am using a trial version of Diptrace , very nice, lots to play with.
Kelvin, I have spent hours and hours trying to figure out the math to do a circle, all to no avail.· I can cut a diagonal by making successive small X and Y cuts, but the basic problem is that I can only drive one stepper at a time.· I thought about using two Stamps, one for X and one for Y, but they would have to be perfectly synced and have exactly the same operating speed, so I did not pursue that approach.· If you come up with a program that cuts a circle using the ellipse approach I would very much like to try it out.
What version of DipTrace are you using?· Will it generate .dxf files of the traces?· The Home Edition is only $75, but it is limited to 250 pins and I don't know if it will generate .dxf files.· I might try out the full-featured version with the 30-day limit to see what I could do with it.· I would be interested in any comments you have about the .dxf files.
I would probably not use the Autorouter since I'm sure it assumes you are making double-sided boards with PTH holes.· Being limited to one side I would have to do the routing manually, which is really no problem.
Kelvin, I sent you a PM but haven't heard back yet, so I'll respond here.
I downloaded a trial version of the program you referred to and I do not like it.· Here are a few of the reasons:
It is not near as user-friendly as Expresspcb.
Most import, when you are drawing a trace, it will not "snap to grid" as ·· Express does.· If you draw a trace to·X 200, the trace might end at ·· 196 or at 211.· The coordinate locator at bottom·right indicates in ·· increments of .010 so you neveer know exactly where you are.
With a few exceptions, the pattern names bear no resemblance to ·· what you are actually looking for.· For example, a 24-pin DIP .600 ·· wide is a DIP-24/H310.· I found a TO-92 transistor in the Cylinder ·· library labelled BCY3 something.
The .dxf export function does not generate a .dxf file I can use.· As far ·· as I can tell, the closest you can come is to select Top Edge and ·· Bottom Edge, which I think traces a path that outlines a trace.· I ·· drew a small trace pattern and the resultant file I got was 937 XY ·· lines, which translates into 1874 DATA lines.
There were several other problems which made the program difficult ·· to use, such as resetting the layer colors each time you open the ·· program, when setting a pad there is only one size which has a ·· .035 hole and can not be modified, and so on.
I shall stick with Expresspcb - I have yet to find a program that is as easy to use and that generates a good .dxf file, even though I have to use the silkscreen layer to draw my etch path and hole patterns.
Since we are not here to promote / criticize anyones software, we will just leave it to whatever floats the boat. The circle / arc thing is a toughie, but i think it is possible. The floating point math is a major obstacle here, as the stamp will have to do some calculations to avoid a huge data entry. Maybe one of those math coprocessors with i2c would help? I will give it a closer look this weekend.
Kelvin, I have attached a little program that came as close as I could get to drawing a circle.· It plots 90 points in a quad, which theoretically is a point every degree.· If you pl;ot the results it comes pretty close to 1/4 circle.· The coordinate resolution would be much better if I had better math for calculating the value of sine.· Maybe you can improve it.
Don't let all the formula-ize put you off.
It is really very simple.
There is an X loop and a Y loop.
How fine the circle is depends on the loop depth.
In other words a loop depth of 8 would render an octagon.
Hope that helps.
Jack
EDIT: Opps! Just noticed that I was not on the last page of this thread when I posted.
BTW: I believe there is an integer only version of this formula.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - 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.
Thanks, PL.· I ran the little program that I posted above.· I didn't want to plot all the points in Express and the convert the resulting .dxf file, so I just took the X and Y parameters my program produced, and by modifying my gcode conversion procedures a bit, I was able to get them into a DATA list format.· I sent the program out to the shop and tomorrow morning I will run it and see what kind of a quarter circle it produces.· If it is even close to decent, I'll post a picture of what I get.
Wel..........here is a picture of my quarter circle - two pictures, actually.· Quad is as close as I could get to the actual size, and Quad2 is a blown up picture so you can see the actual cut.
Not too bad, really.· It looks like that if I had plotted 180 points instead of 90 the path would have been smoother.· Also, if I could move the X and Y tables simultaneously it would have been smoother.· The program defaults to a 1" radius or 2" diameter circle.· I can't draw the left side of a complete circle because my mill travel on the Y axis is 1.7" max; however, I can draw the bottom half.· All I have to do now is to figure out how to rotate the path I have 90 degrees clockwise and move it 1" to the left.· I think if I wrote another program and swapped the X and Y lookup tables, then start/finish the tables at 180/0 that might do it.· Anyway. that's what I'm going to try first.
Here is a picture of a semicircle I just cut.· I'm using a 1" radius and it is exactly 2" from A to B.· I made a couple of minor adjustments to the two programs.· Remember that the origin -·0, 0 - is at the center of my circle.
I'll probably never use this - I just wanted to see if I could do it.
Looks like your plotter for the arc works great, but with only one axis moving at time, the staircase effect is hard to get around. I was trying to find a way to alternate the pulses on the x and y axis to move to a point. But since the distance each axis has to move will not be equal, an increment ratio would have to be determined, through a comparison of the 2 values. eg. the x axis would have to step a certain amount in relation to the y axis. Then it all went to hell when i looked at the math. It looks like having to drive both motors at the same time is the only way for a smooth arc line. The more i look at this stuff, the better the Propeller looks. Still, you have over-achieved for only having one stamp to work with. Gotta like that " nothing is impossible" attitude.
What's the maximum number of "steps" you can use for an arc? The more steps the finer the resolution and the smaller the "staircase".
Can you "computer" a quarter of an arc at a time?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·
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."
Steve, I have written a program that will plot 90 points in a quadrant.· That is one point every degree.· I can't go finer than that, and that dictates how much each step is.· If I could do 180 points in a quadrant then it would be much smoother, but unfortunately, the Stamp can only handle integers so I'm limited to 90 steps.· Also, the math that computes each step distance must work with integers only.·That is why there are "flat" areas at the beginnin and the end of each quadrant. ·I know about Tracy Allen's floating point thing, but I have no idea how I would incorporate it into my program.· I'm just not that smart.
My program uses lookup tables for the next degree to plot, and it reads:
0,1,2,3,4...and so on for X.· I can't add degrees like 1.5 because the Stamp Editor will reject it.· However, I can write:
0, 1, 1*/$017F, 2,...and so on, so this gives me 1.5 degrees.·
The only problem is that the Stamp can't read decimals.· I sent Tracy Allen an e-mail telling him of my math problems, so if he has the time maybe he can come up with something that will let me plot 180 points instead of 90.· That would really improve the circle.· We'll see what happens.
To get your 180points, you could change your gearing....but that's probably more than you were hoping to muck with.
how do you send your commands to the steppers? Serial commands? Could you send two variables separated by a "."?? Would the stepper serial dohickey understand anything other than an integer value?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·
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."
I couldn't generate 180 points with the Stamp, so..............I went to Excel and created a special template.· I added all the intermediate points so now my table reads 1, 1.5,2.2.5 and so on.· This is in Excel, not my lookup tables.
Then I devised a formula in Excel:
degree =(SIN(C1*PI()/180))*1000
SIN in Excel is in radians to 5 decimal points so I had to convert it to degrees - *PI()/180.
This gave me 180 X points and 180 Y points - just what I wanted.· I have the program loaded out in the shop and I'm going to run it as soon as I finish lunch.· I'll post a picture of the results.
I bought the·kit from http://www.drillbitcity.com/catalogue/product_detail.asp?Tg=301-R8WSFBRB·The tool second from the right is designated·for circuit boards however, the taper makes it touchy and unreliable for close traces. A little too deep and you get the grand canyon and a little light and you etch the board without breaking the copper clad.
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."
That is excellent Sid. Even better than i thought possible. You have certainly progressed a long way in the last couple of weeks. I guess you should be able to apply the same type of theory to a diagonal line.
Eric, I use a .020 end mill and a .015 router or end mill·from Drill Bit City.
Kelvin, the circle program is working out quite well.· I can cut any size circle I want by multiplying the values in my Excel table, which were calculated using a radius of 1, ·by the radius of the new circle.
Quad 1 is lower right and Quad 4 is upper right.· The values for Quad 3 are the same as for Quad 1 except that X = -X and Y = -Y.· For Quads 2 and 4 the values of X and Y are reversed.· In Quad 2, X = -X and Y = Y.· In Quad 4 X = X and Y = -Y.
The program cuts counter-clockwise, so if you start with Quad 4 at X = 1000 and Y = 0, when you finish Quad 4 you will be at X = 0, Y = -1000, which is the precise position you need to be at for Quad 3, and so on for Quads 2 and 1.
Each quad takes 360 DATA lines, which is 1080 bytes, so I have to have a separate sub-program for each quad.· The sub-program that runs the etching operation, Quad_Rev_A.bse, is in Slot 3.· The DATA list, for instance, Quad2_Data.bse is in Slot 4.· It contains the DATA list and a very small program to read the DATA, then jump back to Slot 3:
start: ··· READ eePntr, Word f, dir ··· x = 1 ··· RUN 3
To cut a full circle I would put Quad4_Data in Slot 4, then Quad3_Data, and so on.
That should pretty well explain how my program goes about doing its thing.
Newzed said...
Here is a picture of the quad I just cut with my new program. Not too bad !! Comapare it to the picure of the quad I posted a few days ago.
Sid
I need to chime in on this one.
WOW!
That looks perfectly usable.
I concur, nice job.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - 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.
Thanks, PL.· Believe me, the whole project was quite an effort but I think it was worth the time.· Aside from the circle, I also learned how to improve my other etching operations.· Being able to create the .dxf files and converting them into a usable DATA list format saves me hours and hours of time, plus, since I never manually change any coordinates, it gives me an error-free program.· What more could you ask for?
Now I only have to figure out how to cut a 45 degree arc and I'll be finished.
Comments
Since i have only worked with big machines, and mostly milling aluminum, i guess some of my knowledge does not apply to a pcb. I suppose cutting fluid would be a problem here, but maybe a small air nozzle by the mill tip to blow away shaving might help to prolong the life of the bit. I understand the expense of the bits, some i use at work are 50 bucks a piece, and if there is a coolant problem, it will be toast in about 5 seconds. Anyway, i read for pcb drilling that a layer of material should be used on top of the board to get a cleaner cut. I am thinking it may secure the board better, thus reducing the vibration / pulling of the copper when being drilled. Now, what material that would be, i don't know yet, but should not be hard to find out. Steves' idea of moving the bit around in the hole is good, the side of the mill bit will actually deburr the hole some amount. A circular pattern would be perfect if it could be done.
kelvin
For the problem with the burr's on Sid's machine, try layering the whole PCB with Duct tape on the bottom of the board before running the mill's drill operation.
I know it sounds weird, but try it. It helps with woodworking and "chip out", and I've heard it works well with metal too.
Knight.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-
This message transmitted with 100% recycled electrons.
-=-=-=-=-=-
Gravity doesn't exist. The Earth sucks.
-=-=-=-=-=-
Make a man a fire, and he will be warm for the night.
Light the man on fire, and he will be warm for the rest of his life.
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-
I have no problems with the holes - my drill goes in rather slow compared to the speed of an Excellon machine.· The test hole pattern I drilled yesterday produced no burrs at all.· They all come from the etching process.· I usually deburr the boards with 400 grit wet-dry but it would be nice to be able to do it with the mill.
Knight, thanks for the duct tape suggestion but putting it on the bottom of the board that would accomplish nothing that I can see since I usually make only single-sided boards.· I use a piece of 3/8" oak for backup and normally the bottom is burr-free unless I happen to hit a hole that was drilled before.· I flip the backup over when it gets full of holes.
Sid
I'd like to see if I can make a board that would be acceptable to someone besides me, even though I consider myself very picky.
Sid
I don't know if you have a solution for this, but thought i would bring the subject up.
I was looking into the idea of getting some output from the gcode to do a circle. Unless someone knows, or there is a resource available for the math to accomplish this, it looks like it would be futile to bother with. And even if there was, i don't know if the stamp could handle everything for it to run properly. But, i did find something interesting. Arcs and eclipses are handled differently by gcode, because there is no diameter input for the gcode to translate it to the I and J commands, as normally output with a closed circle. So, to make a long story short, it decodes the line to a series of small straight lines. This way the mill would still be able to accept and run the same data format to mill a circle. The pcb software i am playing with can place an eclipse, and then stretch it into a circle. Another way would be to have 2 or 4 arcs made into a circle. Either way, when changed to gcode, it outputs the xy moves. Obviously, it would make a big circle a little rough looking, but a small one would probably be okay. The xy axis would have to be run at the same time, like doing a diagonal cut.
By the way, i am using a trial version of Diptrace , very nice, lots to play with.
kelvin
What version of DipTrace are you using?· Will it generate .dxf files of the traces?· The Home Edition is only $75, but it is limited to 250 pins and I don't know if it will generate .dxf files.· I might try out the full-featured version with the 30-day limit to see what I could do with it.· I would be interested in any comments you have about the .dxf files.
I would probably not use the Autorouter since I'm sure it assumes you are making double-sided boards with PTH holes.· Being limited to one side I would have to do the routing manually, which is really no problem.
Sid
I downloaded a trial version of the program you referred to and I do not like it.· Here are a few of the reasons:
It is not near as user-friendly as Expresspcb.
Most import, when you are drawing a trace, it will not "snap to grid" as
·· Express does.· If you draw a trace to·X 200, the trace might end at
·· 196 or at 211.· The coordinate locator at bottom·right indicates in
·· increments of .010 so you neveer know exactly where you are.
With a few exceptions, the pattern names bear no resemblance to
·· what you are actually looking for.· For example, a 24-pin DIP .600
·· wide is a DIP-24/H310.· I found a TO-92 transistor in the Cylinder
·· library labelled BCY3 something.
The .dxf export function does not generate a .dxf file I can use.· As far
·· as I can tell, the closest you can come is to select Top Edge and
·· Bottom Edge, which I think traces a path that outlines a trace.· I
·· drew a small trace pattern and the resultant file I got was 937 XY
·· lines, which translates into 1874 DATA lines.
There were several other problems which made the program difficult
·· to use, such as resetting the layer colors each time you open the
·· program, when setting a pad there is only one size which has a
·· .035 hole and can not be modified, and so on.
I shall stick with Expresspcb - I have yet to find a program that is as easy to use and that generates a good .dxf file, even though I have to use the silkscreen layer to draw my etch path and hole patterns.
Sid
Since we are not here to promote / criticize anyones software, we will just leave it to whatever floats the boat. The circle / arc thing is a toughie, but i think it is possible. The floating point math is a major obstacle here, as the stamp will have to do some calculations to avoid a huge data entry. Maybe one of those math coprocessors with i2c would help? I will give it a closer look this weekend.
kelvin
Sid
Many smart people have been working on that for hundreds of years.
Many solutions.
The one I would recommend in your case is Bresenham's.
The words to search for are "Bresenham's line algorithm".
Here is a nice example for you. Even has some Basic code.
www.mandelbrot-dazibao.com/Bresen/Bresen.htm
Don't let all the formula-ize put you off.
It is really very simple.
There is an X loop and a Y loop.
How fine the circle is depends on the loop depth.
In other words a loop depth of 8 would render an octagon.
Hope that helps.
Jack
EDIT: Opps! Just noticed that I was not on the last page of this thread when I posted.
BTW: I believe there is an integer only version of this formula.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - 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.
Post Edited (PLJack) : 4/15/2006 10:02:29 PM GMT
Sid
I'm sure.
Looking forward to it.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - 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.
Not too bad, really.· It looks like that if I had plotted 180 points instead of 90 the path would have been smoother.· Also, if I could move the X and Y tables simultaneously it would have been smoother.· The program defaults to a 1" radius or 2" diameter circle.· I can't draw the left side of a complete circle because my mill travel on the Y axis is 1.7" max; however, I can draw the bottom half.· All I have to do now is to figure out how to rotate the path I have 90 degrees clockwise and move it 1" to the left.· I think if I wrote another program and swapped the X and Y lookup tables, then start/finish the tables at 180/0 that might do it.· Anyway. that's what I'm going to try first.
Sid
Post edited by Newzed.
Post Edited (Newzed) : 4/16/2006 2:51:09 PM GMT
Sid
Sid
I'll probably never use this - I just wanted to see if I could do it.
Sid·
Looks like your plotter for the arc works great, but with only one axis moving at time, the staircase effect is hard to get around. I was trying to find a way to alternate the pulses on the x and y axis to move to a point. But since the distance each axis has to move will not be equal, an increment ratio would have to be determined, through a comparison of the 2 values. eg. the x axis would have to step a certain amount in relation to the y axis. Then it all went to hell when i looked at the math. It looks like having to drive both motors at the same time is the only way for a smooth arc line. The more i look at this stuff, the better the Propeller looks. Still, you have over-achieved for only having one stamp to work with. Gotta like that " nothing is impossible" attitude.
kelvin
Can you "computer" a quarter of an arc at a time?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
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."
Sid
My program uses lookup tables for the next degree to plot, and it reads:
0,1,2,3,4...and so on for X.· I can't add degrees like 1.5 because the Stamp Editor will reject it.· However, I can write:
0, 1, 1*/$017F, 2,...and so on, so this gives me 1.5 degrees.·
The only problem is that the Stamp can't read decimals.· I sent Tracy Allen an e-mail telling him of my math problems, so if he has the time maybe he can come up with something that will let me plot 180 points instead of 90.· That would really improve the circle.· We'll see what happens.
Sid
how do you send your commands to the steppers? Serial commands? Could you send two variables separated by a "."?? Would the stepper serial dohickey understand anything other than an integer value?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
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."
I couldn't generate 180 points with the Stamp, so..............I went to Excel and created a special template.· I added all the intermediate points so now my table reads 1, 1.5,2.2.5 and so on.· This is in Excel, not my lookup tables.
Then I devised a formula in Excel:
degree =(SIN(C1*PI()/180))*1000
SIN in Excel is in radians to 5 decimal points so I had to convert it to degrees - *PI()/180.
This gave me 180 X points and 180 Y points - just what I wanted.· I have the program loaded out in the shop and I'm going to run it as soon as I finish lunch.· I'll post a picture of the results.
Sid
Sid
I bought the·kit from http://www.drillbitcity.com/catalogue/product_detail.asp?Tg=301-R8WSFBRB·The tool second from the right is designated·for circuit boards however, the taper makes it touchy and unreliable for close traces. A little too deep and you get the grand canyon and a little light and you etch the board without breaking the copper clad.
Post Edited (Eric R) : 4/18/2006 11:12:40 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
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."
kelvin
Kelvin, the circle program is working out quite well.· I can cut any size circle I want by multiplying the values in my Excel table, which were calculated using a radius of 1, ·by the radius of the new circle.
Quad 1 is lower right and Quad 4 is upper right.· The values for Quad 3 are the same as for Quad 1 except that X = -X and Y = -Y.· For Quads 2 and 4 the values of X and Y are reversed.· In Quad 2, X = -X and Y = Y.· In Quad 4 X = X and Y = -Y.
The program cuts counter-clockwise, so if you start with Quad 4 at X = 1000 and Y = 0, when you finish Quad 4 you will be at X = 0, Y = -1000, which is the precise position you need to be at for Quad 3, and so on for Quads 2 and 1.
Each quad takes 360 DATA lines, which is 1080 bytes, so I have to have a separate sub-program for each quad.· The sub-program that runs the etching operation, Quad_Rev_A.bse, is in Slot 3.· The DATA list, for instance, Quad2_Data.bse is in Slot 4.· It contains the DATA list and a very small program to read the DATA, then jump back to Slot 3:
start:
··· READ eePntr, Word f, dir
··· x = 1
··· RUN 3
To cut a full circle I would put Quad4_Data in Slot 4, then Quad3_Data, and so on.
That should pretty well explain how my program goes about doing its thing.
Sid
WOW!
That looks perfectly usable.
I concur, nice job.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - 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.
Now I only have to figure out how to cut a 45 degree arc and I'll be finished.
Sid