CNC Engaving Machines and Grinders to make the cutters....

[joenoci]

Well, its me again - still no engines made, and again, making tools and machines to make more of them....
I some previous posts I showed a 'large' CNC router that I use for, among other things, to isolation engrave printed circuit boards. I also showed a shop-made tool and cutter grinder, which, also among other things, I used to make the engraving bits with. Both machines served the purpose reasonably well, but were always plagued with setup time and effort, due to being used for many other jobs. In addition, the size of the machines meant that they were not optimum in terms of 'touchy-feeliness' ( now there'e a sense for you!) which meant that sometimes the accuracy of work suffered a little - like using a drill press designed for 20mm drilling in steel to hold a 0.5mm drill bit, drilling in fibreglass - the bit breaks 'cause you have no sense of touch on the quill downstroke..

So of we went on a mission and made two custom, rigid, downsized, accurate little machines to do just those jobs.  here are some pics of the various bits and pieces - I can add more if interest dictates. Hope you enjoy it at least a little - I certainly enjoyed the journey a lot!

Regards
Joe


The 'old large' CNC Router and T&C grinder:





Front and oblique views of the New CNC Engraver:
Table, 500mmx400mm, moves in Y and spindle moves in X and Z
Spindle motors is a RC World Brushless motor, shaft removed and replaced with an ER8 collet chuck with 8mm long shaft. A normal RC speed controller drives the motor
fed with a PWM speed signal from a custom bit of electronics, driven by MACH-3 for the G-Code controller. Spindle speeds are from 1000RPM up to 24000RPM





Various Parts of the Engraver:
The Floating spindle head now allows very accurate depth control when engraving through the 35um copper layer on the PCB substrate.
The floating nose is fitted with a teflon nose piece, and is screw adkustable ( micrometer style) up and down to allow the engraver tip to protrude the required amount
The nose adjustment thread is 0.6mm, ie, 1 turn moves up or down 0.6mm
Vacuum sucks up the debris through the floating nose piece, preventing the nose piece from riding over said debris.





PCB being engraved:



Here is a PCB with a Chip fitted - the 'pin' spacing of the device is 0.5mm, so the engraving has to be very accurate, lest adjacent tracks be cut away as well...
Fitting the chip, using a wide angle microscope, took 20 times longer than it took to engrave the PCB! ( engraving that PCB took 3 minutes)



A View of the completed little T&C grinder:
This was made targeted specifically at the shaping and sharpening of the carbide engraving D bits. Of course , as with most of these gadgets, it can be added to to sharpen most things...The grinding head stared in one of my previous posts - it was the larger of a pair made as Lathe Toolpost grinders.



And its making...

Engraving and indexing the handwheel dials-
And stamping the numbers, etc....









Before cleaning up the dials:



All Clean and nice:



And Sharpening a D Bit:





And a happy Chappy languishes among all the toys...

[RonGinger]

That is some very fine work. Hard to believe you could make a PCB with such fine traces and then solder a chip on it.

Can you show a bit more about that floating nose? I would like to try a PCB on my mill but the slight curve in the stock would be a problem.

[Roger B]

Those are two fine machines    You must be due to start an engine by now 

[joenoci]

Hi there Ron.

I will try to show some pics of how it all goes together. Basically the Z axis works normally, and fitted to the Z axis main plate are two slide receptacles. The slides ( two round polished steel rods) slide up and down in these two receptacles. These two rods are pressed into the base onto which the spindle is fitted, and this assy 'hangs' onto stops, and is pushed up as the Z axis moves down and the spindle nose impinges on the workpiece - A mouthfull at best!

Here the head is in the down position.



Here I pull the head up tween thumb and fingers...



Removing the floating nose from the floating head-



And the piece parts of the nose 'cone'





Nose cone with the indicating depthing jig for the engraving cutter:



The spindle business end, where the nose cone fits:





Nose cone back in place and ready for depthing the tool:



Setting the depthing jig:

Basically an LEd with a 3v3 lithium cell, a jumper lead to the spindle body, and the Ali disc has a 40um deep recess where the engraving tool will touch and light the led.
The you screw out the nose till the LED just dies, and lock the nose piece in place.





Here is a video of the floating head working:
(forgive the sideways view please...)

[youtube1]http://youtu.be/mhIhxr4S-uQ[/youtube1]

Hope that helps??

Joe

[petertha]

Joe, I am drooling over your machines! I really want to make something like your little T&C grinder. Is there some way to obtain construction drawings / parts used? I considered buying a used Sherline X-Y mill cross slide bed & adapting an adjustable Z column. Your setup looks really well thought out.

I think what I’m seeing is an RC-type brushless outrunner motor? If so, are you driving it through a typical (RC) ESC speed control & DC power supply? Pardon the clumsy wording but does running it at ‘reduced throttle’ rpm for sustained periods result in controller heat & issues? Or do you gear it appropriately to some target motor rpm at best efficiency? What are we looking at for typical kv & watts etc?

Where I get stuck on the homebrew grinders is the spindle. Many folks make their own with special bearing types & arrangements, angular contact? anti-axial play?. I’ve also seen compact direct drive (motor+spindle end-to-end) units people adapted as toolpost grinders with ER collets… another project of strong interest. I believe you showed some beautiful creations in the past? Same question - any chance of getting construction details?
http://www.modelenginemaker.com/index.php/topic,3163.msg74419.html#msg74419

[Admiral_dk]

Very, very nice CNC PCB setup you got here  I really should do something similar with both my own CNC router and the one at work. On both I got problems with the Z-axis in regards to do PCB's - the optimum removal is 40um on the thin stuff and 75um on the high quality stuff.

 I'm also very curious about where you get pre-tinned raw PCB stock from - I haven't found any suppliers of this stuff anywhere 

Best wishes

Per

[joenoci]

Hi there petertha.
I fear that, as with most of the bits I make, there are no drawings...Generally it starts out as an idea of what I am trying to achieve and then begins a sort of iterative process - how big the machine will be depends on the work it must do, and on what the the scrap box can be cajoled into regurgitating...The accuracy and rigidity required then also sneaks in to further bias the idea towards something concrete. I do make a few rough sketches to get proportions right, and then start metal bashing!
This little grinder has had a revamp - The need for rigidity did not influence my ideas sufficiently on the first pass ( double speak for 'I screwed up'..). The vertical column, onto which the grinding head is fitted, was very sturdy, but not enough. It is now a 70mmx60mmx300mm tall chunck of steel.  Without this, the smallest out of balance of the drive motor, the drive pulleys, the belt, the grinding wheel, etc would be enough to cause a slight vibration, which in turn was enough to cause microscopic chipping of the cutting edge of the micrograin tungsten carbide I was grinding for the V engraving cutters. So, all you need are X, Y and X axis, with a very sturdy Z, with lots of inertial mass. And balance the rotating parts as best possible. Of course , the tool holder is another road down which you can spend half a lifetime..so many angles, and rotational moments - It was quite a story to get the V cutter motion to work properly, as well as achieving the relief angles. The 'fineness' of Feed of the axis was very important on this machine, as grinding the V cutters proved very challenging indeed. It is only the very tip of the cutter that actually cuts - the copper on the PCB is 35u thick, and the cutter penetrates about 38 to 40u to cut a very narrow slot in the copper. Therefore the tip must be chip free, with correct clearance angles, and end in a sharp point.  Generally 3 out of 10 cutters ground do not work properly...

The drive motor is an RC motor. I used a 285kv motor, running of a variable DC supply, from 12v to 24v. I also generate a variable PWM signal from 1ms to 2ms ( max speed) to drive the ESC. The motor is a 20amp continuous capability motor, but when sharpening the V cutter the motor is running at 1.8ms PWM, and 15volts supply, and 3amps. The motor to spindle speed has a step-up of 1.2 times. Max spindle speed is around 8000RPM. Grinding wheel is a diamond, 46grit ( very fine..) of 100mm diameter now.

The spindle is not fancy - I machined a spindle from stainless steel, ending in an ER11 collet fitting. The bearings are plain, simple ZZ ( double sealed) ball bearings, and I have a light pre-load for min side play. I am happy to replace bearings after X hours of running - bearing are a few dollars each. So far I have ground over 90 cutters, about 10minutes per cutter, and the bearings are still fine!

This spindle/motor combo is one of a pair - I made a smaller, much higher speed unit - both are/were intended for use as toolpost grinders ( internal and external)

Regards
Joe

[joenoci]

Hello there Admiral_dk,
Thanks for the kind words...I am not sure what problems you have with the Z axis in PCB engraving - most problems are due to trying to engrave PCB's without a floating head - that means the Z plane must be perfectly flat and co-planar. Generally that is very difficult to achieve - you can fit a sacrificial base ( supawood, or the like) and machine it down, it will then be almost perfect -do not move the machine after, however, any slight twist of the machine in its 'new' location will will transfer to lack of co-planarity. This still does not always work, as most often the PCB is also slightly warped, and so the engraving bit penetrates to varying depths all over. Since the tip is a V, that translates to varying widths as well. I tried all the aforementioned methods, including a vacuum hold-down table to suck the PCB flat, with some measure of success, but it was just never sufficiently repeatable. The floating head on the smaller machine cured all this. Then any Z axis inaccuracy in depth control also does not matter - the Z axis simply moves up and down 6 or 8mm, with the PCB at around midway of that travel, and the spindle rides - floats- on the PCB surface on a Teflon nose-piece whose protrusion below cutter tip level is finely adjustable. Depth is then very accurate and well controlled. However, the V cutter cutting edge must be very good - any burrs on the copper with make the floating nose piece ride high, lifting the cutter from the copper surface.

I use mostly 35u ( 2oz I think ?) copper, and penetrate about 48u, with a 30 degree include angle V bit - this gives a very nice, narrow slot in the copper, allowing easy engraving of tracks and pads with 0.5mm spacing.
The 'pre-tinned' stock - I purchase from a PCB manufacturer, asking them to take double sided FR4 and to tin/lead plate ( oh Dear, did I say the horrible -Lead- word??) and then reflow. The result is a nice shiny , silver layer, easily soldered. Do not try that with single sided stock - it warps something horrible!

[Admiral_dk]

Thank you very much for your reply.

I'm very aware of the fact that the PCB's are not all flat in themselves  and I'm certainly aware that your solution solves this problem and that's why I like it   

And your last reply explains why I can't find any tinned stock and how you get it anyway ....

[GailinNM]

Joe,
Thanks very much for posting all the details on your floating head engraver and cutter grinder.

I have been wanting to make a floating head for my Bridgeport for some time. I have been making PCBs by painting the PC board and then burning away the paint with a laser engraver. Then I etched the board in a conventional fashion. This produces low-quality but acceptable results.

So, following your lead and using many of your ideas I designed and built an engraving head to fit my Bridgeport CNC milling machine. It is not finished yet, but I was able to do some test runs with it today. It is mounted on a NMTB 30 tool holder to fit the Bridgeport spindle. Anti-rotation is accomplished with an 8 mm linear bearing following a shaft which replaces the anti-rotation block provided on the Bridgeport for use with tapping heads.

The spindle is an 300 W integrated engraving spindle motor from China with an ER 11 collet. It is rated at 12,000 RPM but it appears that with the light load is running at about 13,000 RPM. A higher RPM would be desirable, but it was available. The mount for the motor is carried on two 12 mm linear slides. Instead of using a vacuum to clear debris, I pressurize the head so low pressure air surrounds the cutting tool to blow debris away.

A photo of the engraving head installed and the first test printed circuit board are shown below. The test board shown was engraved with a 20° included angle cutter with a 0.3 mm tip. I was running the motor from a 30 V bench supply so was getting less than 8000 RPM and did not want to use a finer tip cutter had that low RPM. I have a 48 volt power supply to mount inside the Bridgeport control cabinet. It will be turned on with the same signal I normally use to turn on the air for a mist coolant system and that will also turn on the air to the cutting head. I will get all that done in the next week or so.

I still need to tidy up the software that I wrote to generate the G code from a DXF file. It's working okay but needs to be optimized to reduce extra movements.

If anyone is interested, I will create a new thread, so as to not dilute yours, detailing this head and/or the software I have written.

Thanks again for the detailed post.
Gail in NM

[joenoci]

Hi there GailinNM.

Nice going with that spindle! I make quite a lot of PCB's - electronics is a major part of my hobbies - Autopilots, Animal tracking tags, etc..That drove me to do a small engraver dedicated to the task as it was always a pain to set up the big engraver - it always had some or other jig or fixture on the table, for the next job in line, and setup is always the big pain! Anyway, the little machine does a new board nearly every day!

I chose vacuum to remove the chips , mainly motivated by trying to keep the fibreglass out of all the slides and leadscrews - rather suck it up out of the way. There is not much fibreglass in the chippings, since engraving depth is slight, but still. Also, it keeps it out of the air I breath!

These days so many of the slightly more complex electronic components are moving to very fine pin spacing surface mount types - you can see in my post the uProcessor with 0.5mm pin spacing. That leaves little space for some copper track, and a gap between tracks and pins! So I grind my cutters to a point, no flattened tip at all, and use a 30deg included angle and engrave to around 0.05mm depth - I vary that for the best cutting width as needed.  What is the width of the void in your sample? It seems around 0.5 to 0.7mm? ( assuming those square pads are 0.1inch apart?)

Also, the higher the RPM the better. I find mediocre results below 15k RPM and try to work at 22k to 25k ( the max I can get out of my spindle..)

When trying to achieve such fine tracks and spacings, there is a very fine balance between cutter RPM and X/Y feed speeds as well. To slow feed shows strong evidence of 'smearing' with resultant burrs, which lift the floating head, of course. I find the sweet spot lies between 10 and 20mm per second.

I guess my spindle is also 'Chinese' - it is after all a 3 phase brushless model plane motor , made in China!

Have fun!
Regards
Joe.