The Official Propeller Driven CNC PCB Drilling And Laser Direct Imaging Machine
idbruce
Posts: 6,197
Hello Everyone
Over the past couple weeks, I have been keeping kind of a low profile in the forums, mostly because I have been busy. As you may have guessed from the title, I have been working on my CNC PCB Drilling Machine, and I must say that it is really starting to shape up. Since I just finished making the Z carriage spindle holder assembly, I figured it was about time that I start a dedicated thread for this new project. However, as the title suggests, this machine will also be used for laser direct imaging, but I have not completed any work on that part of the project. My intention is to create a spindle lock, and to create a laser assembly with an 1/8" shaft that can be locked into the spindle. Anyhow, this thread has been started a little early just to build the suspense and hype a little. By tomorrow evening, I will have taken a photo of the Z carriage spindle holder assembly, and I will upload it for your viewing pleasure, and there will be many more photos to come in the near future.
Until then, if you are interested in building machinery, you may want to view some of my other threads: Bruce
Over the past couple weeks, I have been keeping kind of a low profile in the forums, mostly because I have been busy. As you may have guessed from the title, I have been working on my CNC PCB Drilling Machine, and I must say that it is really starting to shape up. Since I just finished making the Z carriage spindle holder assembly, I figured it was about time that I start a dedicated thread for this new project. However, as the title suggests, this machine will also be used for laser direct imaging, but I have not completed any work on that part of the project. My intention is to create a spindle lock, and to create a laser assembly with an 1/8" shaft that can be locked into the spindle. Anyhow, this thread has been started a little early just to build the suspense and hype a little. By tomorrow evening, I will have taken a photo of the Z carriage spindle holder assembly, and I will upload it for your viewing pleasure, and there will be many more photos to come in the near future.
Until then, if you are interested in building machinery, you may want to view some of my other threads: Bruce
Comments
enjoy!
Mike
im have no cnc knowlege but by just looking at this thread im guessimg u connecct this to a cnc its not a standalone product. if someone were to want to build this and wjatever else is required to print pcbs with it what id a ball park price tag?
@rwgast_logicdesign -
Unfortunately, there is probably a minimum of a $1000 investment to be able to quickly and accurately reproduce PCBs. Of course there are many, many factors to be considered when talking about cost, such as: materials used, type of drivers, displays, power supply, accuracy, electronic enclosures, etc.... The list goes on.... Additionally, it would all depend upon the extent of reproduction. Does reproduction include plated through holes, silkscreening, and solder mask? If so, this would drive the cost up much further. However, I would imagine that if a lot of salvaged materials were used, and if speed, accuracy, and quality were not an issue, and that if plated through holes, silkscreening, and solder mask were not required, then I would imagine that a hobbyist could probably slap a drilling machine together for several hundred dollars. However there is no magic genie available for designing such machines and it all takes time.
This will be a stand alone CNC drilling machine machine. Boards ready for drilling will be placed upon the machine, and when the machine is powered up, it will search a micro SD card reader for a valid drill file. If a valid drill file is found, the machine will then permit a zeroing process and wait for the push of a start button. When the start button has been pushed, automatic drilling will commence.
Prior to drilling the PCBs, I will be using my Exposure Box/Cylinder to prepare the PCBS. To learn more about making PCBs, please take some time to review this thread: http://forums.parallax.com/showthread.php?135950-ATTN-PCBers-SUBJECT-New-website-and-exposure-box-design-for-your-creative-needs
Bruce
Without the proper equipment, prototyping machinery is a very time consuming process. When my patent sells, one of the first things that I am going to do, is build a nice shop, with some nice equipment. A lathe, a mill, a bandsaw, and a couple of welders are at the top of my list. Oh, and I certainly can't forget the foundry furnace, but most likely, I will probably build the one that I designed years ago. Proper equipment is essential for a productive work environment.
Anyhow....
I have three bearing holders to make for the lead screws, and then I should be able to bolt the main assemblies together. Theoretically speaking, by the end of today, I should have all the necessary components to start drilling PCBs, but in reality, there is still quite a bit more work that I want to do, before even attempting to do any drilling.
However, I will probably bolt all the main components to a piece of wood and give them a test run, just to see it all in action.
Bruce
Well I did not get everything bolted together yesterday as I had hoped. The bearing holders are a much bigger pain than I had anticipated, and in addition to that, I had forgotten that my motor mounts are not square.
When building machine prototypes, I normally use architectual extruded aluminum to make my parts, unless of course the part is so unique that it requires casting or some serious milling. Using architectual aluminum for parts has both it's pros and cons. The upside is that ready made shapes are widely available, such as tubes and bars. The downside is that the tolerances are all over the board. Referring back to the unsquare motor mounts.... Extruded aluminum angle is never square, (at least not in my experience)
If anyone should ever end up in a similar situation and would like to see a pictorial drawing of the bending jig, just ask, and I will upload one.
Bruce
EDIT: I cannot imagine that I will need to save this drawing, so I have decided to upload it now. This is a side view of the motor mount straightner. Screws have been omitted from the drawing.
Referring to yesterdays discussion of unsquare motor mounts, I have decided to provide a little more information, with the hope that it will someday be useful to another person. To aid in this discussion, I have provided a modified drawing of the motor mount straightner. If you look at the drawing closely, you will notice that I have trimmed two edges of metal. The purpose of this trimming is to provide an interference free area to take a measurement with dial calipers.
Considering that I want these mounts to be as square as possible, instead of relying upon a store bought square, I have decided to use the Pythagorean Theorem. For those of you who are unaware, the Pythagorean Theorem simply states, "In a right triangle, the square of the length of the hypotenuse is equal to the sum of the squares of the lengths of the legs". In other words, A squared plus B squared equals C squared.
In my application, the NEMA 17 motor mounts are constructed from 2" X 2" X 1/8" aluminum angle. As mentioned in my previous post, the tolerances for architectual aluminum are not particularly accurate, and the measurements given should be considered nominal. In reality, the actual measurements are 1.991" X 1.994" X 0.123". For the sake of discussion and referring to the provided drawing, it will be assumed that A equals 1.991" and that B equals 1.994". To apply these measurements to the Pythagorean Theorem, 1.991" would be squared (3.964" rounded) and 1.994" would be squared (3.976" rounded), and the sum of adding these two together would equal 7.94", which is equivalent to C squared. By obtaining the square root of 7.94" (2.818" rounded), we know that the motor mount will be square when C equals 2.818". Actually this will be very very close to being square because of the rounding of the measurements and slight loss of length due to bending the metal. However, it should be close enough to suit my needs.
Once again, I hope this will be useful to someone in the future.
Bruce
EDIT: I just straightened my first mount and it took about five minutes to get it perfect. After the bending procedure, the angle tends to flex back, so incremental overbends became necessary. Just a little at a time. The tool worked very well in my opinion. Two more to go.
EDIT: If anyone ever decides to make this tool and use the Pythagorean Theorem to help determine when a mount is square, you must be very careful when measuring with the calipers to just catch the edges, otherwise you might overbend the mounts like I did. OOOPPPSSS. Now I will have to make an adapter plate to bend them in the opposite direction
EDIT: No adapter plate is needed to bend in the opposite direction, just rearrange the parts. Additionally, I was making this way too complicated with the Pythagorean Theorem, because obviously the bending jig forms a right angle. Simply bend the angle aluminum until both legs are parellel with the bending jig. DUH
I snapped a motor shaft on my wire feeder for the bender several years back, which was due to an uncentered drive roller and too much tension ($132 loss). Since then, I pay particular attention to having things center and in accurate alignment.
I must admit that there are some pretty nice couplers on the market, but they are a bit rich for my blood and especially for this particular machine. Even though this particular machine requires very high accuracy, I am trying to assemble it in the cheapest possible means available. Hopefully I won't regret it, but if I do, luckily these motors are about a third of the forementioned price.
In my situation, I am not coupling two shafts, but just a leadscrew to the motor. The main factor in my case would be the carriage binding, but that should not be a problem.
Thanks for your input.
Bruce
Your machines rock!
I thought about something similar, but eventually concluded to bend the mounts. I always weigh my options before proceding and then choose what I think is the best course of action. Not that this is the sole reason, but one reason to bend the mounts as opposed to leveling them is heat transfer. Stepper motors can get warm, and the more metal to metal contact there is, the better the heat dissipation, which will eventually make the motors last longer.
Bruce
As an emergency mount, I have cut tubing in half, and actually, extruded aluminum tubing is worse. When you finally cut through it, it tends to collapse inward, like it is under pressure.
Anyhow, all is good. My mounts are now in perfect shape.
Bruce
It has been a while since I have updated this thread, so here goes.
This machine is being constructed with three custom built stepper/screw driven linear actuators. The motor mounts, lead screws, couplers, and pillow blocks have been constructed with nothing more than a drill press, milling vise, and a few miscellaneous tools. At this current point in time, all three linear actuators operate fairly smoothly, however there is approximately 0.015" runout with the carriages per revolution on each actuator, which is unacceptable to me. I believe the shaft couplings are in good shape (I could be wrong), so I believe the runout is due to the bearing journals on the end of the lead screws not being concentric.
It is times like these that I wish I had a lathe and and a dial indicator.
All in all, I am very pleased with the design, I just need to eliminate the runout.
Bruce
I decided to remachine both the leadscrews and couplings, and as it stands now, I currently have approximately 0.005" of runout per revolution. Considering that this will not be a accumulative error on an axis, but only a slight deviation resulting in a zigzag, I believe it is quite acceptacle for any PCB drilling I might encounter, especially if I make my pads slightly larger. However, when it pertains to the future laser direct imaging aspect of the machine, I believe I will have to further reduce this rounout, but of course I will do some trial runs before making any changes. If further reductions in runout should become an absolute necessity, I will probably first attempt to change the leadscrews and leadnuts for a better class of thread fit to eliminate the side to side slop of the low class threads.
Bruce
EDIT: As an after thought to this post, I may have to further reduce the runout before proceeding on the Z axis, because I have heard of tungsten carbide bit breakage due to runout.
Yes Phil, I know
I'm impressed by your dogged determinism to "roll your own" critical pieces of your machine. I suspect though that for the precision you seem to be pursuing, you will have numerous disappointments. As has indicated to you before when you were trivializing the ease machining something to "a few thou", this is very difficult - indeed, bordering on futile - without a proper mill and lathe. I'm aghast that you are attempting to make your own lead screws.... do yourself a favour and purchase a proper screw set with recirculating ball nuts from McMaster-Carr. It will be the best money you have spent!
Cheers,
Peter
I am a pretty resourceful guy
I'm sure you are resourceful and have demonstrated some pretty amazing results. However, as one who "has been there", let me reiterate a caution for the many pitfalls in your path, otherwise you had better turn your "frustation filter" WAAAAY UP!
Cheers,
Peter (pjv)
One of the main problems is that this machine is very small, which would make it difficult to find stock components. Additionally, this machine will introduce a new design of linear actuators that does not contain some of the most common components. Basically with this machine, I am striving for very low cost with minimal parts. I realize that for all the time I have invested, I could have purchased some of my crucial parts (providing I could find any that would fit), but then I would not have learned all the things I have learned along the way. I now have a nice jig for grinding drill bits and making couplings, and soon I will have a nice jig for grinding journals and threading some drill rod. I just need some time to think it through well enough to obtain some decent screws.
Bruce
Check out Post #10 of the following thread: http://forums.parallax.com/showthread.php?140164-How-To-Make-Shaft-Couplings-Backyard-Style
I believe this will resolve my problem.
Bruce
OK, so maybe you can get them to an acceptable average diameter. What will you do about non-uniform thread pitch along its length or non concentric threads? Really, rolled threaded rods are not uniform or smooth enough enough for your desired precision. At an accuracy of 10 thou, maybe. But you are looking for "a few thou". The contact surface in the thread is also typically horrible, and will wear your nut out more quickly. If you don't want to spend the cash for proper gound lead screws and ball nuts, then at least go for a commercial Acme screw and nut.
Cheers,
Peter (pjv)
My answer is two double spring loaded UHMWPE tensioning blocks on each side of the lead screw. I previously discussed cost with another person. ACME lead screws and nuts would drive up the cost of the machine by another $150. When I am done with my assemblies, they may wear out much quicker, but hey, they will only cost me a couple of bucks to replace.
As with my previous discussion with Phil, you are probably right, but I am just trying to make it much more affordable. I at least have to try my design and see what the outcome is.
Bruce
I know people have pretty good luck using Acetal as a nut material. If you split that nut and add a way to adjust the squeeze you can take out a lot of backlash at the expense of some extra friction. Acetal has pretty good wear properties too. Know a guy that gets just as good a performance from his small mill with Acetal split nuts (ACME screws) as most do from affordable ball screws.
I missed the "pitch" thing,,,, It went straight over my head and I should have been paying better attention. To be perfectly honest, I never really considered the accuracy of the pitch. I guess at this point, I can just hope for the best. If it fails miserably, at least the techniques that I am learning now of making accurate couplings and grinding lead screws, well I should be able to apply these techniques to ACME threaded rod as well.
Concerning backlash, I am using (2) silcone bronze nuts per carriage assembly.
Bruce