Nueva Pieza: Circuito electrónico
La maquina tiene suficiente precisión como para hacer este tipo de trabajos. La fresa hay que comprarla adrede para estos trabajos, pues la fibra de la que esta hecha la placa es bastante abrasiva y un acero rápido lo desgasta en seguida y pierde el filo. Hay que modificar el código G para que haga todo en una sola pasada, no hacen falta varias. Y lo mas importante, el resultado es electrónicamente valido, es decir, se han creado las pistas conductoras sin cortes ni solapes, y seguramente funcionará. Cuando este en marcha con los componentes montados, colgaré una nueva entrada.
Yes, I wanted to broaden horizons, and to check if the machine has enough accuracy to make delicate little things, such as an electronic circuit.
A good friend , who helped me in all the electronic stuff of the machine (connections, wiring , etc …) had to get a little PCB and ask me for one. So we must to try. He give me directly the G code that had generated the program you used to design the board and check if it was compatible with the LinuxCNC, and except for some details, It worked very well.
As an introduction ( although I am not electronic, so, sorry for any mistake), I will comment that a PCB Board consist in a plate of glass fiber with a layer of a certain thickness of copper, which is milled to remove that layer leaving only lanes as you want it to be conductive . Of course,you will need a thin, sharp cutter , and a machine with enough precision is required .
About the mill, I had no idea, so I bought at Leroy Merlin, a Dremmel tool, conical sharp enough, but for wood. Total 12 €
We put the board in place , and after define the origin and the Z coordinate, in order to machined respect to the copper surface we want to go down (about 3 tenths of a millimeter ) , we give the Play.
We realize that the program makes a first pass around the edge and then two more displaced a certain distance , so that the wide groove, but leaving some imperfections.
Sure, this configuration comes standard with the program itself, and as we didn’t knew yet using it, we saw that we didn’t like. Why not make all the path at the same time, getting down more with the tool and removing more material?
We tried more with the program and discover how to change these variables, and finally returned to position another plate, with a new cutter (we lost the sharp of the tool after the first plate) and the result was more than acceptable.
At least the opinion of my friend, said it was fine, and for accuracy we were both pleasantly surprised. Neither Backlash or anything, mills moves and can be positioned in an error of 2 to 3 tenths of a millimeter.
Then we took the same tool to mark where the holes will be drilled to mount the various components, mainly because we forget the bit.
The machine has enough precision to do this type of work. The cutter has to be bought on purpose for this work, as the board fiber is quite abrasive and wearing a high speed steel acan make we loss the edge. We must change the G code to do everything in one path, not need several. And most importantly, the result is valid in electronic point of view, ie, have created the conductive tracks without breaks or overlaps, and it certainly works.