Liberty4Ever's Hadron Build
Posted: Thu May 31, 2012 8:20 am
I should have waited until I had an image that looked vaguely like a Hadron ORD Bot to start my build log, but I needed to get some of this obsession documented.
Briefly, I've been keeping an eye on the affordable end of 3D printing for years. Every few months, I'd take another peek. Nope. Still crummy print quality. That was OK for those who were doing it as a hobby, but I wanted to make real parts that weren't stringy and ripply, and blobby.
Finally! The ORD Bot! Thank you Bart! It was the 3D printer I had been wanting. Unfortunately, my timing stinks and I missed the leading edge of this wave. No nice Hadron kit for me. I've been struggling to source the parts. Even with the very nice BOM (thanks again Bart!) there is much to learn and many electronic parts and accessory hardware bits to find that aren't listed on the scavenger hunt shopping list.
Another unfortunate artifact of my late arrival was the disjointed order in which I'm stumbling upon the information. I've kept tabs on the 3D print quality and the general principles for years, but none of the gritty details. This is my first 3D printer, and there is MUCH to learn. I've spent a lot of time reading websites and watching YouTube videos, but my education is very random. I'm learning a lot of intricate specifics before learning a lot of the basics. It's a bit like walking in on a movie that's been playing for 55 minutes, walking out five minutes later, finding the book that was the basis for the movie and reading the last chapter, then watching the edited-for-TV version of the movie at 2:00 AM while repeatedly nodding off on the couch.
After a couple of weeks, I finally stumbled upon Bart's 37 page build log (http://www.buildlog.net/forum/viewtopic.php?f=16&t=1035) and I'll be reading that from start to finish after I get a few hours of sleep. It was a bit discouraging to read Bart say in the initial post, "The basic prototype was designed, fabricated and assembled in about 4-6 hours over 6 days." I have 5-6 times as many hours invested over a bit more than a week, and all I have to show for it is a few parts in a box with a lot more on the way, but no actual building.
The hardware, software and firmware are developing very rapidly, so this project is an exercise in hitting a fast moving target. On average, Slic3r has a new release more often than I change underwear. I shouldn't complain about fast free upgrades to open source software. The frantic pace does contribute to some inefficiency though. I bought a Mk 1 heated build platform and then saw the very nice looking Lava heated build platform in ORD Bot blue, so I ordered one of those tonight. I'll either sell the Mk 1 on eBay as a new unused part, or I'll keep it as a spare. Similarly, I'd like a fully integrated ORDuino controller, but I already bought the Arduino & RAMPS stack, and I'm a bit reluctant to buy into a brand new controller at this point. With all I need to learn, I'm trying my best to stay on the path and not go off blazing any new trails.
As interesting as a Hadron would be as a hobby, I don't need a hobby. I have plenty of hobbies that I'm ignoring. I'm running a small business and simultaneously retrofitting a 10X24 Clausing lathe in my basement for CNC, and retrofitting a mid 1980's vintage Powermatic gantry router (two 15 HP routers, a 7.5 HP radial arm saw and two air drills over a 4'X8' bed) for modern CNC control. Both are LinuxCNC projects, and I'm learning LinuxCNC as I go. So, I basically have three concurrent projects.
My ultimate goal for the ORD Bot is to run production parts for my home based internet business. The business model is predicated upon low overhead and efficient manufacturing, to offer value to the customer in niche markets. I have ideas for products that I'd like to have injection molded, but the $3500 to $6500 cost for an aluminum quick mold and the cost of the minimum initial run would have a lengthy payback period. My business model is a natural for 3D printing. If I can run an average of ten parts a day (about 1.5 ounces each), I could be initially profitable and my cash flows would stay well in the black.
Most people want a 3D printer to print little one-off designs or prototypes, but I want to develop additive machining short run production in the hobby end of the market. I think that has huge potential for many small businesses. I expect inexpensive 3D printing to be a transformational technology. I also expect to benefit by iterating 3-5 prototypes in a day for new designs, but my primary focus is on fairly continuous short run production. I expect to need a second Hadron very soon. Being able to debug the 3D printing process for each part as I'm making changes to the prototypes and then go into production on the same machine offers huge quality benefits and time savings compared to starting over with a different manufacturing technique after the final prototype design is approved.
I initially thought about scaling up the Hadron for a larger build area to accommodate short run production, but I don't think that would help me much, at least for the parts I plan on making. In fact, a better strategy might be the smaller Quantum 3D printer with a new & improved continuous build platform that would allow each part to cool and then eject it so a new part could be built. That may be one of the contributions I make to the open source 3D printing community.
I'm also very interested in the surface finish of the printed parts. I don't expect the finish of injection molded parts, but I plan on selling these parts so looks are important. I've manually finished black Delrin with a random orbital sander and produced a very nice velvet matte finish that was fingerprint resistant. I'd like to achieve a similar result if possible with the 3D printed parts. The first thing I'll try is aqueous tumbling with tiny stainless pins. This produces a gorgeous polish on brass. It might reduce the layered effect on 3D printed parts and result in a more uniform surface finish. If not, I'll experiment with abrasive tumbling. I have both rotating drum and vibrating bowl tumblers. Other possibilities for improving the surface finish without substantially altering the dimensions include IR and hot air to reflow the outer surface. Whatever I develop for improving the surface finish will be production friendly. Tossing 50-100 parts in a drum and tumbling four hours is acceptable. Hand sanding is not.
I bought most of the parts to build a benchtop injection molding machine but never got around to it. I may finish that project, not as an injection molding machine, but rather as a filament maker. I can buy ABS resin in pellets, as used by injection molders, for about 33 cents a pound in reasonably small quantities. That's about a third of the cost of ABS filament, so there is some cost savings there. I might be tempted to pursue that project... AFTER my 3D printer is debugged and making good parts and lots of them.
I'm about to be inundated by Hadron parts. I promise that my future build log posts will be much more technical and much less philosophical.
Briefly, I've been keeping an eye on the affordable end of 3D printing for years. Every few months, I'd take another peek. Nope. Still crummy print quality. That was OK for those who were doing it as a hobby, but I wanted to make real parts that weren't stringy and ripply, and blobby.
Finally! The ORD Bot! Thank you Bart! It was the 3D printer I had been wanting. Unfortunately, my timing stinks and I missed the leading edge of this wave. No nice Hadron kit for me. I've been struggling to source the parts. Even with the very nice BOM (thanks again Bart!) there is much to learn and many electronic parts and accessory hardware bits to find that aren't listed on the scavenger hunt shopping list.
Another unfortunate artifact of my late arrival was the disjointed order in which I'm stumbling upon the information. I've kept tabs on the 3D print quality and the general principles for years, but none of the gritty details. This is my first 3D printer, and there is MUCH to learn. I've spent a lot of time reading websites and watching YouTube videos, but my education is very random. I'm learning a lot of intricate specifics before learning a lot of the basics. It's a bit like walking in on a movie that's been playing for 55 minutes, walking out five minutes later, finding the book that was the basis for the movie and reading the last chapter, then watching the edited-for-TV version of the movie at 2:00 AM while repeatedly nodding off on the couch.
After a couple of weeks, I finally stumbled upon Bart's 37 page build log (http://www.buildlog.net/forum/viewtopic.php?f=16&t=1035) and I'll be reading that from start to finish after I get a few hours of sleep. It was a bit discouraging to read Bart say in the initial post, "The basic prototype was designed, fabricated and assembled in about 4-6 hours over 6 days." I have 5-6 times as many hours invested over a bit more than a week, and all I have to show for it is a few parts in a box with a lot more on the way, but no actual building.
The hardware, software and firmware are developing very rapidly, so this project is an exercise in hitting a fast moving target. On average, Slic3r has a new release more often than I change underwear. I shouldn't complain about fast free upgrades to open source software. The frantic pace does contribute to some inefficiency though. I bought a Mk 1 heated build platform and then saw the very nice looking Lava heated build platform in ORD Bot blue, so I ordered one of those tonight. I'll either sell the Mk 1 on eBay as a new unused part, or I'll keep it as a spare. Similarly, I'd like a fully integrated ORDuino controller, but I already bought the Arduino & RAMPS stack, and I'm a bit reluctant to buy into a brand new controller at this point. With all I need to learn, I'm trying my best to stay on the path and not go off blazing any new trails.
As interesting as a Hadron would be as a hobby, I don't need a hobby. I have plenty of hobbies that I'm ignoring. I'm running a small business and simultaneously retrofitting a 10X24 Clausing lathe in my basement for CNC, and retrofitting a mid 1980's vintage Powermatic gantry router (two 15 HP routers, a 7.5 HP radial arm saw and two air drills over a 4'X8' bed) for modern CNC control. Both are LinuxCNC projects, and I'm learning LinuxCNC as I go. So, I basically have three concurrent projects.
My ultimate goal for the ORD Bot is to run production parts for my home based internet business. The business model is predicated upon low overhead and efficient manufacturing, to offer value to the customer in niche markets. I have ideas for products that I'd like to have injection molded, but the $3500 to $6500 cost for an aluminum quick mold and the cost of the minimum initial run would have a lengthy payback period. My business model is a natural for 3D printing. If I can run an average of ten parts a day (about 1.5 ounces each), I could be initially profitable and my cash flows would stay well in the black.
Most people want a 3D printer to print little one-off designs or prototypes, but I want to develop additive machining short run production in the hobby end of the market. I think that has huge potential for many small businesses. I expect inexpensive 3D printing to be a transformational technology. I also expect to benefit by iterating 3-5 prototypes in a day for new designs, but my primary focus is on fairly continuous short run production. I expect to need a second Hadron very soon. Being able to debug the 3D printing process for each part as I'm making changes to the prototypes and then go into production on the same machine offers huge quality benefits and time savings compared to starting over with a different manufacturing technique after the final prototype design is approved.
I initially thought about scaling up the Hadron for a larger build area to accommodate short run production, but I don't think that would help me much, at least for the parts I plan on making. In fact, a better strategy might be the smaller Quantum 3D printer with a new & improved continuous build platform that would allow each part to cool and then eject it so a new part could be built. That may be one of the contributions I make to the open source 3D printing community.
I'm also very interested in the surface finish of the printed parts. I don't expect the finish of injection molded parts, but I plan on selling these parts so looks are important. I've manually finished black Delrin with a random orbital sander and produced a very nice velvet matte finish that was fingerprint resistant. I'd like to achieve a similar result if possible with the 3D printed parts. The first thing I'll try is aqueous tumbling with tiny stainless pins. This produces a gorgeous polish on brass. It might reduce the layered effect on 3D printed parts and result in a more uniform surface finish. If not, I'll experiment with abrasive tumbling. I have both rotating drum and vibrating bowl tumblers. Other possibilities for improving the surface finish without substantially altering the dimensions include IR and hot air to reflow the outer surface. Whatever I develop for improving the surface finish will be production friendly. Tossing 50-100 parts in a drum and tumbling four hours is acceptable. Hand sanding is not.
I bought most of the parts to build a benchtop injection molding machine but never got around to it. I may finish that project, not as an injection molding machine, but rather as a filament maker. I can buy ABS resin in pellets, as used by injection molders, for about 33 cents a pound in reasonably small quantities. That's about a third of the cost of ABS filament, so there is some cost savings there. I might be tempted to pursue that project... AFTER my 3D printer is debugged and making good parts and lots of them.
I'm about to be inundated by Hadron parts. I promise that my future build log posts will be much more technical and much less philosophical.