Comparison between EasyMaker and other printers

Hi All,

MetrixCreateSpace in Seattle, the hackerspace where I’m a member, has a 3d printer night every thursday night.  I go every once in a while to keep track of developments in the space.  As you know, I’m not currently selling EasyMakers ( Mr. Market has spoken ), but I’m really surprised by the low quality of the other offerings, and don’t understand why people are willing to buy 3D printers that have such terrible problems.  I’ve seen the following robots up close, and talked to their owners over the last few months — these are their stories and my opinions:

Ultimaker, Ord Hadron, Rostock ( Johan himself ), Rostock clones ( Made with either OpenBeam or 80/20 ), and have come to some conclusions.

Rostock is good.  The clones of it are OK right now, but may be good soon.  The problems in the Rostck clone designs are getting solved, and are mostly around the slide carriage/bearing systems.  This is an easy fix, and I expect the Rostock derivatives to be good printers eventually — but they’re just OK right now.  The original Rostck isn’t as cool looking as the T-Slot rail derivatives, but it works better.  The 8mm rods and lm8uu bearings are more proven than any of the other bearing systems used in the clones.  If you’re building your own printer, a Rostock, but not any of the current derivatives, is a good choice.  I think in 6 moths, the derivatives will be pretty good as they fix the few middling issues they have left.  Id’ call it 90% ready — but not yet consumer ready.  The Original Rostock is a nice and simple robot.  The derivatives have too many parts that need too much skill to assemble, and unless you’ve already built a 3d printer, cannot recommend the derivatives — stick to Johan’s design.  It’s a little faster than EasyMaker, and has better X/Y space efficiency.  It has poor Z space efficiency — resulting in a good desktop footprint, but a very tall tower printer.  It cannot mill, and multiple extruders may be a problem for it.

The Ultimaker.  The owner loved it, and hated it.  She spent over a month trying to calibrate it, and managed to get one small, nice quality print.  The one part was really nice — but she couldn’t replicate the result.  It’s the same speed as EasyMaker, just harder to setup and get quality parts from.  At the same time, it’s also a one-trick pony.  OverHyped, but not very good.  If you’re highly skilled, you can make it a good printer that works well — but this printer is really a “user skill” robot, and I can’t recommend it at this time.

The Ord Hadron bot.  This thing is crap.  I don’t know why anyone would get one.  The motion system shakes like a scared Chihuahua.  The quality of parts used is very nice — it’s a very complete-looking, attractive robot.  It’s also entirely form over function.  Quality of output is terrible.  If you get an ORD, get the smaller one.  The motion system does not scale well.  I have video of the owner talking about how badly it shakes.  It shakes so badly, yo cannot see it in video — the camera’s image stabilization software thinks it’s my hand shaking, and compensates.  The problem is the 90 degree connectors used to mount Z are way too weak.  There’s not enough room in the design to put in proper strength connectors.  Hence, the robot has to be very small to work.  The Hadron version is larger than the normal ORD bot — and the Hadron is too large for the connectors.  Bigger/better Z beam connectors are needed to make this robot decent, and the room for that just isn’t there.  The other option is to make it a Bowden drive.  The owner had mounted a Greg’s extruder, and that was just too much moving mass for this weak a design.

MakerGear M1 — I didn’t get to see this running, but I was impressed with it.  The owner was having electronics problems, and it wouldn’t boot.  The design seemed nice, and I think it’ll make a good printer.  Small, portable, lightweight.  I’d consider it for a portable printing robot.  I think the M2 will likely be as nice.  If I wanted a portable, fully built unit, I would seriously consider this unit.  I would like to see it working to be sure, but I have a good opinion of it.

MakerBot cupcake — This was seriously upgraded over the original, and used the nicest plastic.  It was older, but the print quality was very good.  Say what you will about MakerBot, their printers are very good, and the most user friendly.  I’ve seen the cupcake run for the longest periods of time, and consistently get good quality and long run times.  It’s nowhere as good as EasyMaker in final quality — but the cupcake is rock solid reliable.  MakerBot did a great job on extruder design, and on electronics.  These are two large problem areas for RepRap designs.  The Mk6 on is a great extruder, but prior to Mk6, extruder quality was spotty.  I’d recommend getting an Mk7 extruder ( or clone ) if you’re building your own printer.  It’s much better than any other extruder on the market right now.  Makerbot’s Gear and filament guide idea are wonderful ideas, and every extruder should use both from here on in.  The filament guide is crucial for both feed and water control.

I’ll also debunk some bad information I fell victim to — the Stratasys device has good output quality, far better than any MakerBot I have seen so far, and is fully supported( aka overhangs).  I’ve seen bad output quality and good output quality from the devices — the skill of the operator is a big deal here.  I had previously seen only terrible parts, be recently, got to see good parts,  Also — most 3D printers depend on the skill of the operator.  It’s possible that EasyMaker’s quality is due to me, not due to the machine.  Some of what I’ve seen — I wonder why/how people buy those robots.  On the low end, the original Rostock is a good choice.  On the medium end, the MakerGear printers are a good choice, and on the high end — MakerBot or EasyMaker are good choices, as of right now for the home market.  In the professional market, I think that the Stratasys device could be a good choice.    I don’t know the middle professional end, maybe an OBjet.  At the very high end of professional devices, EOS is the one to beat — their printers are amazing — but they also start at many hundred thousand dollars and go over a million for their top of the line.   I’ve never seen as much detail, density, and strength as I have from an EOS device.  They are the “wow” level for me.  If EOS made a home device in $2,500 or less price point, I would buy it instantly, sight unseen.

I’ve backed the EventrBot and the MakiBox.  I’m looking forward to receiving these robots.  Why so many Robots?   EasyMaker is an amazing printer, but I find I need multiple robots to design efficiently — they each have different roles.  Grover needs a robot to test his designs with, and we sometimes contend for use of the existing robot.  3D prints are slow — and I design large parts.  Right now, I’m lucky to design, print, and test a part a week.  At this rate, I won’t have the next robot done until next fall.  Some weeks, like this one, were lost entirely to extruder issues from being in the cold, wet garage.   So, I went a little nutty and decided to get multiple robots.  Each of these robots I got as basic kit — I have a 3d printer and a lot of skill — so I can build each robot for very little.  The first 3d printer you get and make will be very expensive — the second will be nearly free.  They’ll both take some time to arrive.  I backed the Makibox in February, 2012.  I hope to get it before next summer, if I’m lucky.  I have no idea when Eventrbot will ship.  Again, I hope to get it before next summer.  Maybe Santa will help these ship before christmas?  That would be wonderful!

So, the MakiBox will be the “on the desk, inside the house, ABS” robot used for testing small to midsized parts where ABS is the right material to work with.  ABS is terribly difficult to work with in the cold — it is not “garage” friendly at all.  The problem is curl.  I can heat the bed and keep the part smooth until completion with ABS.  Then, I turn off the heated bed and lift the part.  The ABS shrinks and curls right in front of my eyes.  The layers shrink/curl differently, and I watch the part deform once in the open air of the garage, and I watch stress fractures form.  It’s really interesting!  But, ABS can’t be in a cold garage and get good results.  I have great ABS results — from the summer time, printed on hot days, inside the house.  If it’s in the house, it has to be small — or the wife will kill me.  The Maki is the smallest robot — EM is far too large to be allowed in the house.The EventrBot is to replace my RepRap Prusa, which is a terrible robot, with something better, and will be an “in the garage, PLA” robot, used for slightly larger parts and for parts better made with PLA.  When I can. EasyMaker will be put into mill mode, and will primarily be used as a mill. This way, I can get some high-density, large parts.  I may keep EasyMaker a printer until the end of the experimenting phase.  Printers aren’t as useful as CNC mills, in my opinion.  But, they’re far more home friendly.  If a design works on when made on a 3d printer, and can be manufactured in a single setup on a 3 axis CNC mill, then it’s a good design.  The single setup mill operation is needed to get good costs at production scales.  The “works when 3d printed” indicates good tolerances for production scale.  If both are true, then you can go to scale cheaply/efficiently.

Alright, thanks for reading such a long post.

–Imran

Printing below 50 degrees

Hi All,

My Robot lives in the garage.  An unheated garage in Seattle.  This is the first winter where I’ve kept the robot in the Garage, and man, is it a painful thing!  Once the temperature drops below 50F ( say 12C or so ), the printer is pretty much useless.

Problem — The filament can no longer be fed through the hot end.  The filament becomes cold and brittle, and the Wade’s hobbed bolt just strips it when trying to feed it.  It also has much higher friction characteristics — it needs a lot more pressure and heat to flow and output anything.  I have a Ron’s extruder as well — the Ron is a direct-drive extruder, using a planetary gear for torque.  It doesn’t work in these conditions, either.  While the MakerBot style spur gear does a much better job of gripping filament, and provides the needed grip to feed without stripping, the motor isn’t powerful enough to feed in those conditions.  So, the Wade’s doesn’t have the grip, and the Ron’s doesn’t have the power.

So, what to do?  I can either move it back in the house ( something the wife hates ), or I can design a new extruder ( which is a waste of time — I’m not selling EasyMaker, so designing new parts for EasyMaker won’t yield a return. ).  If I must design a new extruder, then I’ll have to make it compatible with the next design ( I’m not going to do the work, only to have a design I can’t run ).  I’m leaning towards a design that hybrids Ron’s and Wade’s design, and is meant for vertical mounting.  This would allow all three extruders ( yes, three.  I want a triple extruder system ) to fit in the same line.   Mass will become a problem with three motors mounted on a moving carriage.  I think a Bowden may be required at this point, if I want a triple-extruder design — regardless of how much power I put behind it.

Thanks,

Imran

Ron’s Compact Extruder Review

Hi All,

I recently replaced my Wade’s extruder with a Ron’s extruder, and have been putting it through its paces. It’s a decent extruder, but the motor is at the edge of speed and power requirements, making it a tough extruder to get working all the time.

The good:

Very small, amazingly lightweight, very cheap.  Super easy to load filament, mount, clean.

The bad:

slow retract speed, needs a recompile to make it work.  Motor is at the edge of power, cannot always feed plastic in all conditions.

 

Overall, Ron’s extruder is a really cool design.  I’ve talked to people on the IRC channel who have it or use the same motor in a wade’s. It doesn’t work for me.  I print in an unheated garage, and it is winter here.  The cold weather “freezes” the PLA filament I use.  Ron’s design doesn’t have the torque to push the plastic, and the motor fails badly when it fails.  So, it feeds for a while, until the pressure in the hot end hits the motor’s torque limit, which causes the motor to “unroll” due to the stall, then begin feeding correctly again.

It also is a very low current motor, and it’s easy to over-current.  If you’re feeding in a warm room, relatively soft plastic ( like ABS ), then Ron’s is a great extruder.  Don’t use it with PLA in a cold garage — it’ll go past the limit the extruder can handle, and you’ll have a lot of trouble with it.  Stick with a Wade’s and torque it down heavily in cold conditions.

 

Thanks,

Imran

General Update

Hi All,

Peerbhai Robotics was formed when I designed a CNC mill/3d printer hybrid robot I called EasyMaker. This blog was forked off my main ipeerbhai.wordpress.com blog, as I found that the volume of people looking for the robot was high enough to separate out and, with some luck, make the robot available for sale. I then did some additional development of the robot after forking the blog, and eventually, reached a point where I felt it was complete enough to try a Kickstarter. The KickStarter failed, and I’m now designing a new robot, and have reconfigured the site as a blog for the new robot’s development.

The roadmap for the new robot is as follows:
1. Design a new linear motion system. Existing linear motions systems are cheap and have good availability — in China. I have a Chinese supplier, and can get the bearings here in the US for a good price, and the rail comes in many lengths right now. So, why design a new liner motion system, when good, cheap ones already exist? Lead time and production risk. It takes me 60-90 days to freight ship parts to me. I can air ship, but then, the cost and quantity have to go down, and it still takes 4 weeks to get through customs and the like. If I want to control production risk, I need a bearings system I can source in the US — from a large, reputable supplier. I’ve chosen Grainger and VXB as my bearing suppliers, and McMaster as my shaft provider — so, am building my linear motion system around what I can get from these 3 suppliers. The new linear motion system has different designs for X, Y, and Z. Right now, Grover and I are designing and testing bearing elements in different configurations.

2. Design alignment systems. There are essentially 2 major types of error in a cartesian system — they are positional error and square error. Positional error is solved by the linear motion system, and square error by accurate and precise positioning of parts. EasyMaker taught me how to design a system with little positional error — but square error was more difficult to solve. I could do it, but I wasn’t positive users could. The new robot will be designed around easy squaring and low positional error.

3. Cost reduction. EasyMaker has a very large BOM, and thus, it costs a lot. The new robot has fewer parts, more integrated together.

4. Designed for milling. I don’t mean functionality — I mean in producing the parts. I had designed EM to be produced by Laser cutter and 3D printer. This made EM a true RepRap, in that it could make its own parts. But no one technology could make the parts — so I couldn’t outsource to a single supplier, as I’d have to use different job shops to get the best price quotes. This made EM expensive. The new robot, while produced via 3d printing, is designed to be produced via 3 axis milling.

5. Design a secret feature. This is something I won’t blog about — other than the new robot has a secret feature that no other robot has.

6. Put it all together, launch a new KickStarter, and see how it goes.  EM didn’t launch — but it did show me that I’m close to the right point.  The new robot will, hopefully, align better with what the market wants.  EM is a truly great robot.  Not just good — Great.  I’ve got a Prusa and a mill right now, and EM is a better robot than either of them.  Hence, I’m using it in the design of the new robot, and you can see it being my main workhorse robot in some of the videos in the blog.  The new robot will take everything I learned from EasyMaker, and improve upon it.

Market vision

I wanted to also share my market vision, for those of you in the same field.  3d printer/CNC mill hybrids have existed for some time.  I know of 5 robots designed around this idea:

1. Zen toolworks F3 robots.

2. Hydra-MMM

3. QU-BD’s hybrid

4. SumPod

5. EasyMaker.

More entries are coming into the market, and Hydra-MMM isn’t actually available ( it was a student’s design, and never released to the public ).  At the same time, more people are entering the market, and more convergence is happening in designs.  I fully expect the “Machine Shop in a Box” concept to reach fruition in multiple offerings over the next 3-5 years, with some strong robots in the space.  All of the designs work as claimed.  I haven’t seen them all in action, but I’ve seen some of them working, and I designed one of them ( EasyMaker! ).  I know what the laws of physics says — and the problems in the space aren’t that hard to solve.  A single person can, for a little while longer anyway, solve them all  now.  I fully expect the robots that can do this to be commonplace, and many conventional ideas to change.  I also expect more machines to be converged into the concept.  We’re almost at the point of a new singularity here — a revolution in home production.  It’ll be exciting to participate in its unfolding!

624 u-groove wheel

Wow, I didn’t expect this part to print!
It’s a u-groove wheel for 624 bearings to ride on a linear shaft. It is essentially a huge over-hang. I expected the print to fail.

But, the garage is cold. PLA layers are cooling so fast, that the arch is enough support.

Easymaker is able to print perfect circles. The Z radius is too large, but the part works!!! Rolling is perfect!

20121103-095255.jpg