Which color laser?

On Tuesday, October 25, 2016 at 2:45:37 PM UTC-4, Jim Wilkins wrote:


=================> >>

So, the foot is spring loaded towards the board and the router to foot dis tance is controlled by a solenoid? That seems pretty easy to rig up. I'm no t sure why you'd (at least I'd) need to vary the Z axis at all. The range o f widths in their bits seems to be only a couple of mils. Seems like if you 're (I'm) drawing outlines of copper, a cut's a cut. If I need a wider cut, why not just take two or more passes? The software for that would be prett y simple once the rest was done.
I doubt I'd be able to find it in my archives, and if I coupld, it may very well be on an 8" floppy, but actually, years ago when I used a layout serv ice in Canada to design boards, and all I had was a single-pen plotter, I w rote software to read Gerbers and draw the outlines of actual-width traces. Of course, I was younger and smarter then ;) Seriously, though, once you h ave the picture in your head, it's just a matter of drawing rectangles with rounded ends. Not a very big deal at all.
Now, I'm in danger of actually doing this. I have a couple of X-Y mechanism s kicking around and a 45Krpm handpiece...
Dammitall, I still have actual work to be doing!
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

So, the foot is spring loaded towards the board and the router to foot distance is controlled by a solenoid? That seems pretty easy to rig up. I'm not sure why you'd (at least I'd) need to vary the Z axis at all. The range of widths in their bits seems to be only a couple of mils. Seems like if you're (I'm) drawing outlines of copper, a cut's a cut. If I need a wider cut, why not just take two or more passes? The software for that would be pretty simple once the rest was done.
=====================I don't exactly remember how the head operated. IIRC a stop screw set how far the vee shaped cutter extended below the foot when the head was down, maybe pulled by the solenoid? It could also drill the holes with a longer stroke.
I planned jobs such that the board fab time didn't delay the project and didn't use the T-Tech any more than I had to. Since it only made 2 layer boards without plated vias it wasn't that much better for prototyping than perfboard, and tended to require a lot of inspection and hand deburring. I usually needed 4 or 6 layers with internal power planes, or RF microstrip on Duroid.
--jsw
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Tuesday, October 25, 2016 at 11:24:37 PM UTC-4, Jim Wilkins wrote:

Their hottest looking machine has a tool changer to do drilling & other wid th milling, and auto-zeros in the z-axis. The also have (not inexpensive) h ole-plating & multilayer stuff. I could do without the tool changer, and th e auto-zero would be pretty straightforward. Many of my boards are far less demanding than the 6x6mil I said before. even 12x12 would be good enough t o get a lot of work done. I don't do microwaves, I do mostly industrial con trols and that stuff tends to be bigger. many of the parts I use are availa ble on breakout boards (generally for robotics hobbyists) and sometimes as arduino-type shields. Given that, and the fact that I'm rarely constrained by small enclosures, I usually don't need the high precision. I'm starting to think that I really could do this with an x/y positioned dremel and a tw o-position z axis.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Wed, 26 Oct 2016 06:43:19 -0700 (PDT), rangerssuck

The problem with the Dremel is the way excessive runout. But if you can fiddle with it to get the runout down to just a couple tenths then it would work. The runout has to be tiny at speed, not just when turning by hand, and it needs to be pretty good over the length of the cutter. With a limber spindle runout can cause the spindle to describe a circle as it spins at speed so that your .002" dia. cutter mills a .004" wide slot even though the tip is perfectly concentric to the spindle axis. You can buy carbide engraving cutters that have very small points so getting small spaces between the traces would be no problem. If you can tame any runout. Eric
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload


Manually jog the cutter along the margin of the board and measure the cut width with a magnifier. I don't remember graduations on the Z adjustment of the T-Tech I used. --jsw
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I estimated that I could build the mechanical part from surplus Thompson rod etc without too much difficulty. In college I built a wooden animation stand with similar x, y and z motions for a film-maker. https://en.wikipedia.org/wiki/Animation_stand
The snag would likely be having to code design rule checking to ensure that the cutter never violated the overlapping exclusion zones of all nearby design and drafting features. You couldn't just outline each line segment independently, they are meant to be created by an additive process that allows them to overlap while milling is subtractive. It was hard enough to properly define all the widths and clearances in the pad stacks when the program did the checking, although I had more layers to attend to on a board with inner planes, solder masks and silkscreens.
--jsw
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload


I faintly remember another technician showing me a board on which the T-Tech had fully outlined each track segment, resulting in arcs and circles milled through the tracks at every pad, corner and junction. He may have drawn the tracks as overlapping graphic lines instead of named electrical nets and the compiler didn't recognize that the pads and track segments should be left connected.
I quit planning how to build one when I realized that the programming effort to check spacing and detect all overlap conditions would more than nullify the time saved not waiting for etched boards.
--jsw
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
CNC isn't a milling machine. LPKF has all sorts. I was certified with them.
CNC is computer numeral control. e.g. commands sent.
CNC laser driven to expose trace and then in the box or outside develop and etch.
I used to use Ruby and blue tape for IC's. Then PCB's then CAD came and early cad used a CNC laser to draw on the wall to make large film sheets. In those days we were into 18x24 many layer. Later we went large square and special material with fluid cooling on both sides...
The small LPKF I used was to make protype test boards to spec and one day after the Cad design was completed.
Martin
On 10/24/2016 1:19 PM, rangerssuck wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Wednesday, October 19, 2016 at 4:19:24 PM UTC-4, snipped-for-privacy@whidbey.com wrote :

Hi Eric. This doesn't directly answer your question but it may give you som e ideas. Cutting lasers for fabricating work are mostly of two types -- fir st, CO2 lasers, with a light wavelength of 10.6 micrometers, which don't wo rk well at all on highly reflective metals. The light is just reflected bac k, screwing up the mirror and other components.
Fiber lasers, with a wavelength of 1.064 micrometers, are effective on copp er, brass, stainless steel, and other reflective metals. YAG lasers produce the same wavelength and they're used mostly for engraving and marking, rat her than cutting.
You'll note that these wavelengths are at opposite ends of the infrared ran ge. In the course of writing about these lasers for the past five years or so I've often asked the experts if the light frequency has anything to do w ith absorption in different metals. The usual answer is "no." The reason th at fiber lasers are more effective on reflective metals, they tell me, is t hat the shorter wavelength allows a much more concentrated focusing spot -- power is much more concentrated -- and that fiber lasers have less trouble with reflections. I haven't dug into it more deeply than that, but the peo ple at Trumpf, Bystronic, LVD-Strippett, Amada, or others who make both typ es of lasers may be able to give you more info.
I'm skeptical that laser *color* has much influence on the heating. Reflect ive metals tend to reflect a very broad band of frequencies. But it's not a n issue I've investigated, so take this for what it's worth.
--
Ed Huntress

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.