Batman would be proud. Progress.

This is my second attempt at making a post, because firefox decided for me to crash. Thanks firefox! Anyways... onto the real stuff.

(Picture of grant owning this PCB)

Today, I went over to Grant's house and we started to build the hobbycnc control board. We put Grant's fancy soldering iron to work on the PCB. For the most part he was doing the soldering, and I was ensuring component placement based on the schematic and instructions from the kit. What was nice, was that the board was clearly and neatly ordered/organized. I did solder a few components just for good measure, and to remind myself how fun it can be. I see it as a miniature TIG welder without the annoying foot pedal.



(I'm soldering on a few parts just to make sure I can still do it; and I can yay!)

If I build another CNC I will seriously consider using this company again. The only two shortcomings that i forsee with this board is the lack of opto-isolation (although troy believes he has a fix for this), and secondly the lack of high voltage so we can't overdrive the stepper motors much. While this board can only go to 36V the Gecko Drivers can go to 80V, although the nice thing about this board is that it is almost all contained on this board.

So we built almost all the board, with exception to some of the logic chips. The reason is you want to test the board for solder bridges and continuity before applying current and voltage that could potentially damage the expensive logic chips. Once you are sure your board is functioning properly with the powersupply connected you can install the logic chips.

Given that we didn't purchase a powersupply yet we could not finish off the board completely. This is because Troy and Grant were going to explore the option of fabricating a powersupply from old parts from a microwave power supply. From the documentation it looks like we need an 18A powersupply; at this point I don't know if it's okay for us to exceed that, although my gut says its fine.

Here's a website with some good information about building or buying a powersupply http://www.solsylva.com/cnc/power_supply.shtml

Keep chugging along

Steppers ordered
While dealing with my Discrete math class (scored 100% on the midterm wooo), and keeping my airsoft business alive (shipping product); I managed to select and order some 425 oz-in stepper motors for all 3 axis on our CNC machine. All together with shipping it was 163.17$, We could have gone with a ~300 oz-in motor for 30$ less, but the price/performance difference was warranted an increase in power. We wanted to be sure we could meet our torque requirement. We are happy with rapid speeds of around 70 IPM, but we are confident with appropriate gearing/microstepping we can hit whatever targets speeds we need. Other than the previous research and calculations that were conducted a few days/weeks ago (with that program I wrote), I did additional research on cnczone. The interesting reads I found were:

http://www.cnczone.com/forums/forumdisplay.php?f=193&order=desc&page=17
http://www.cnczone.com/forums/showthread.php?t=27527&page=2
http://www.cnczone.com/forums/showthread.php?t=80245
http://www.cnczone.com/forums/showthread.php?t=75192&page=2
http://www.designworldonline.com/articles/3807/21/Improving-High-Speed-Performance-of-Hybrid-Stepper-Motors.aspx?utm_source=mcnl&utm_medium=newsletter&utm_content=feature&utm_campaign=MCNEWSLETTER
http://www.homeshopcnc.com/StepperMotorFAQ2.html

Also while inspecting specs on keling's website I needed to convert 2.15N-M to oz-in, a quick google solution was found at:

http://www.unitconversion.org/torque/newton-meters-to-ounce-force-inchs-conversion.html

It looks like our machine will be severaly limited on the Rapid speeds based on the controller board, with a max voltage of 36V. So we can't overdrive the steppers much. They run 6V stock. So we will only be able to run them about 2x their rated voltage even though we could go up to 20x their rated voltage. With the stepper order aside, I continued my frame research for our machine.
-----------------
Frame Research
For awhile we were seriously considering making the whole frame out of epoxy/granite composite. But after eyeing the costs, the difficulty in manufacturing and the time frame we are on, Troy and I decided it would be wise to follow the KISS principle. I had found several instances of aluminum Bosch framing for homebuilt CNC machines. Given the repeated success I had found, I pushed for what we know works, and with ease. Given our constraints on the project Troy, agreed heartily. Although, we both conceded it would be very cool to incorporate some E/G aspects into the project later to enhance rigidity and vibration absorption.

A simple way to do that would be to fill the cavities of the aluminum framing with E/G. Also we also conceded it would be very nice to have a solid head, so we may make that solid later. At the moment I am toiling with fasteners. I measured one of the THK rails we obtained from ebay. They appear to have 7mm holes for SHCS (Socket Head Cap Screws.) and 9mm countersunk holes. A quick google search is not promising for 7MM SHCS.

http://www.roymech.co.uk/Useful_Tables/Screws/cap_screws.htm

Shows some typical metric SHCS sizes. M7 is not on the list. I wonder if they are specialty screws for the rails, or if they simply use M6 screws.

In my research I began to learn that these so called aluminum frames, are sometimes nicknamed "80-20" extrusions. Which I guess is some sort of industry standard. Apparently there is even a whole section on cnczone dedicated to 80-20 frames. (i'll have to read into it more before making final design decisions.)

From an inspection of Bosch's website and manuals it looks like a 45x45H extrusion section should meet our demands for the X/Y tables. I am a little concerned because milling flat that extrusion will be a hastle on the bridgeport mill that hasn't been opened up yet at our shop. Not to mention we don't even know if the bridgeport's ways are true. Any error in that machine will be amplified in our own base if they are off. And from my research it is not uncommon for these extrusions to have about .008" of variability across a single piece. I also read some individuals using T-slot inserts that are premachined and have SHCS holes in it. I hope i can find more on this.

Some websites concerning 80-20 framing:
80-20 surplus
80-20-T-Slot-Aluminum-Extrusion-45-S-45-4545-
Bosch 3D models
Bosch Website
http://www.3dpublisher.net/SWDownloads1/2044316659-857082/Machining.PDF

I know I will be ordering some Mic6 precision ground plate for bolting some of these frame parts together, just don't know the physical dimensions yet. So that's on hold. I want to model the whole thing in solidworks first.

So to review:
I need to:

• M7 screws- do they exist for the THK rails?
• Precision ground t-slot rails for 80-20 t-slots to bolt the THK rails to - do they exist/where?
• 80-20 End machining options at bosch are confusing, figure it out!
  
• Model it!
  

The truth about building a CNC machine
Gosh everything about this project is Read, Read, Read, process, learn, discover how very little you actually know, rinse and repeat. And somewhere in there you feel confident enough that you aren't going to make a disastorously horrible mistake, and that it is about time to learn by doing.

Progress, but slow

I'm getting this lingering feeling that all the projects I'm doing are getting spread too thin. There is considerable progress when you look at the big picture, but I feel they are getting off track. This shotgun approach to picking tasks I don't think works for the longterm. I may need to review my project goals, and evaluate what would be the most worthwhile use of my time. Despite this I'm still having fun, and doing really cool things.

Like today I started and finished an End-mill box. I'm becoming more proficient at developing and converting designs to be used with the Epilogue laser cutter. Here are two things I made today:


An endmill box and a Metric/English Ruler. I accounted for the laser's thickness to be .004" this made a very snug fit. So snug in fact that when i first assembled it - it took me a full 10 minutes to pry it apart to put the ribs in. I'm going to have to come up with a solution for that. Possibly a bit of simple sanding. I'll have to make a second box as I apparently have about twice as many endmills as depicted.

The ruler, didn't come out that great because the driver interpreted some of the lines that I intended to be engraved, as cutting lines. So the ruler is fairly serrated as the image shows. I'll have to make more later, but hey at least its decently accurate and useful!

Airsoft Business on its feet

Airsoft Business
I visited the USPS post office today and figured out the shipping details for the airsoft products that I've developed. After figuring that, and some tests with the hopup arms that I developed; I released it to the public. After a few sales, I quickly discovered why secretary's are useful for this sort of work (it takes a surprisingly large chunk of time and effort.) I now have a newfound respect for them. It is so rewarding to see a product you've developed being distributed to help others with their goals. (Albiet, that is for playing airsoft lol)

CNC Progress
We fairly accurately priced out several of the future components we need. From this assessment the total cost of the CNC will be around 1733$. So split among our team that will be 577$ per person. Not bad.

Pricing the frame was interesting, because we decided to make it almost entirely of a aggregate/resin mix. We will use what is called "play sand" found at your local hardware store for about 5$/50 lbs. We estimated a "log" geometry where it will be formed from a series of legs that are connected together with a log geometry of 4"x4"x35" we overestimated a required volume of about 840 in^3 for the composite frame. With about 20% resin usage that puts us at a required resin volume of 4 gallons. (924 in^3) The cost of resin is about 200$ for that much. If we get lucky and can get away with only 2 gallons which would be ideal because that would insure a good dampening effect from the frame.

I think a good mixing idea would be to mix the resin without hardener in the aggregate mix before adding the hardener, that way we can ensure the resin is fully distributed through the aggregate.

What a wonderful Day

I finished making a batch of airsoft parts that I will be distributing. I had do prototype tests to ensure they fit properly. I had some issues with localized heat in the acrylic which was causing deformation to occur. A few tweaks to the geometry of the part fixed that. I also finished making the "wooden box" except I used what i had available - which was acrylic. I may make some more boxes later out of wood. Here are some pictures.



Some of the airsoft parts i manufactured (in clear) compared to the original stock factory part. I made all of these parts with a 45 watt CO2 Laser.

Wooden box

Not much happened with engineering today. I had discrete mathematics today, came back from class and received a letter from the MECOP program saying that I had been accepted! http://mecop.ous.edu/

I finished modeling a wooden box to cut out of .25" thick wood on the Epilogue laser cutter at the tech shop. It was surprisingly more difficult to model then the rotary table... This is because the geometry of the lid required me to do some math. Also the corners of the boxes required some interesting thought which I had not performed before.

I'm going to have to convert it to a DXF file, or make a drawing file from the model because the Epilogue 45 Watt CO2 laser cutter needs lines defined as hairlines and converting between file types sometimes you lose that vector based geometry in the image.

Reference:
http://www.woodworkingcorner.com/kschest.php

Rotary table calculations

Image from: http://upload.wikimedia.org/wikipedia/commons/archive/2/2c/20080518200048!Terms_involute_gear_engagement.PNG

Rotary Table

For the rotary table we will be making a gear cutter, worm, and worm wheel. I just "figured out" an equation for calculating TPI, and worm gear teeth. Starting with S=r(theta) I derived:

60/TPI = r(2*pi)

Where 60 is the gear ratio of the worm gear to the worm wheel. Referenced below (in the previous post) is what lets you calculate the resolution that will give you on your table. The 'r' in the equation i suspect is the diametrical pitch, or namely the diameter where the force of the gear is being transmitted, not the OD of the gear. Anyways with this equation I can adjust my TPI for the cutter, and worm gear till I find a TPI that matches what the lathe is capable of cutting. Now in my research i did find that 40' was common for this application no the 60' degrees you typically see. So I need to do some more research or simply "test" it out with an ACME cutter that i already have. (I'd prefer to take the simple route)

An example calculation yielded:
8 TPI, would yield a 1.1936 radius worm wheel.

CNC update + rotary table start

I received our first CNC package of THK linear rails, they are about 30" long were used, but in great condition. My only gripe is the wipers on some of the carriages have a tendency to "catch" on the rail a little bit. I simply hand bent it up a marginal bit and it rides smooth as butter. You can see the ground sheen off them. I put electrical tape on the ends of the rails to ensure the carriages don't slide off. Pictures of these rails will come later.

I had done some research on the spindle, and found several examples of folks using treadmill motors to power their mini-cnc's typically in the 2-3 HP range, (DC motor) with an operational range of 0-6750 RPM. I didn't know much about the controller, and Troy instantly thought that was a great idea, because the VFD we were looking at for the 3 Phase 3HP motor he has would cost 300$. Where we found a treadmill on craigslist for 20$. Hopefully we can pick it up tommorow for that price. In any event he found a DC controller for the motor for less than 100$ on ebay on a 'buy it now' auction which puts at a much more cost effective solution for the motor control system.

There was also a discussion about frame materials, cost, rigidity, ease of manufacturing, and ease of transportation. So things like extruded high grade aluminum (6061), welding steel vs. warpage, and granite composite were discussed. No conclusions were drawn up with that except that we want our solution to incorporate all the attributes listed above.

When I met Troy and Bryan at the techshop. They showed me how to do assembly's in SW. This made me tremendously happy because it's like doors are being opened up for my imagination. Since we finished the fly cutter, we are now going to move onto other bigger and better projects, (which involve the use of our previously made fly cutter tehehe.) This time it is a rotary table. So I started working on modeling the rotary table on solidworks. Here is what i've got so far:



One of the problems i'm encountering is I don't know the dimensions for a standard t-slot so i went with an arbitrary depth of 1 inch and 0.25" slots. The beauty of parametric modeling, is I can go back and fix the dimensions later, Yay! I've also been playing with screw placement, clearance issues with mounting bolts and the bolts that hold the rotary table together.

The problem is I'm trying to make use of material that I already have (0.38" thick steel stock.) Which I have plenty of so I'm going to have to carefully play with the dimensions and design internal blocks to bolt everything together since 0.38" steel is not thick enough for me to comfortably drill and tap strong bolts into it. While I could calculate the yield load and fatigue life of bolts that would fit into the stock, I simply am going to use the KISS principle and oversize the bolts, and use internal mounting blocks.

References:
http://www.technologystudent.com/gears1/worm1.htm
http://www.helicron.net/workshop/gearcutting/cutting_gears/
http://www.astronomiainumbria.org/advanced_internet_files/meccanica/easyweb.easynet.co.uk/_chrish/worms.htm
http://www.cnczone.com/forums/showthread.php?t=63992

Fly cutter test

The fly cutter was tested today! Admittedly, I still have to use these silly bolts that "bump" into each other because that is all i could find at Lowe's/Home Depot. Apparently it's not wise to make something 1/4-28 NF thread. I did see quite a bit of 10-32 NF though.



I hand ground a HSS-cobalt blank on the cheap Ryobi 6" grinder at the shop. I noticed the wheels were pretty dirty from dust/aluminum. So I located the dresser, and dressed both wheels (they were in bad condition.) From there I found a bucket of water and got to work using cnccookbook.com as a guide. The rounded edge for cutting was the hardest part. I think i took a little too much off so it likes like 10-20 thousands wide, where the guide i read said you should go for 3-6 thousands. Oh well. It worked!

Here is a v-block i used earlier, and i faced one side with the fly cutter in a jiffy, hand feeding around 4ipm at 660 rpm. (High speed i know - but there is no lower gear on that machine.)

CNC update

I have successfully purchased all the linear rails i'll need for the CNC. They all are THK brand, the Z-axis has HSR20 carriage blocks for the moment load. The x and y axis are on SR20 and SR25's. We'll use the SR25's on the X-axis to take the weight of the Y-axis. The lengths for each axis are:

x-axis = 32.25"
y-axis = 29-7/8"
z-axis = 20"

Now these linear rail length's are the total length not the travel length, because the carriages when bolted up to the table mounting plate will be taking something like 8" off the total travel. So a rough guess of our work envelope will be something like:

x-axis = 24"
y-axis = 22"
z-axis = 12"

We now are working on locating screws and purchasing screws. I also purchased http://www.hobbycnc.com/products/hobbycnc-ez-driver-board-kit/ for 64$. We are holding off on the steppers for just a bit to finalize our torque requirements. Troy has been busy on the spindle design. He sucessfully modeled it up in SW, and he is performing FEA calculations on it. While I did find him some angular contact bearings that dimensionally work, and financially work, they did not have specifications on the tolerances. Troy has been working on some equations he can apply to ensure everything falls into our safety factors.



I have been keeping a spreadsheet on expenditures... Oh man it's starting to add up. I alone have purchased 450$ worth of parts specifically for our CNC alone. Troy special ordered some 4340 for his spindle for 50$. He also needs to make a tool post grinder, take a safety milling class, and buy angular contact bearings, and a VFD (300$) all for the spindle. This is going to hurt our pockets. This is really going to start being the "test" on us, we are on a tight budget (our own pocket money no less.) Our wits, our ability to scrounge, our courage, and our vigor are all going to be necessary to get through this project. The last project that i have taken part in that is of this magnitude was the 2008 HPV project. That said back to research!