Makerbot 5th Generation: Slop and Wiggle

[brent]

http://youtu.be/8aWZ3zoyVpQ
Video says/shows it all.

Thoughts?

[Generic Default]

Did they make it out of rubber or something? My Rostock's magnetic arms aren't even rigidly connector to the effector, but the effector only deflects about 2mm before breaking contact with a few pounds of force.


I thought cartesian machines were supposed to be more rigid than that. All those fancy injection molded parts and they just had to make them out of a soft plastic. Maybe if they had used 60% glass filled PPA they would have a better machine. I guess this really sucks for everyone who pre-ordered them or is going to get one for thousands of dollars.

I think 3d printers need to be built more like conventional CNC mills. Any stepper motor can move the nozzle at 300 mm/s, which is what printer manufacturers advertise. But the acceleration and deflection are what show up in prints. Rigid machines are good machines.

Just as a reference, here are the relative modulus' of elasticity of several materials;

ABS 1.5 to 3
PLA 4
Nylon 2 to 4
60% Glass filled PPA (similar to nylon) 20
Aluminum 70
Brass 120
Steel 200

So for those of you who want to build your own printer, I recommend using metal for frame and mechanical parts.

[dtgriscom]

It's not clear to me how much that lack of rigidity will affect the final print quality. No, it's not great, but one job of a manufacturer is to figure out where they can cut costs without significantly affecting results. I'd want to see how it prints before I made any judgements (e.g. if the slicer approaches the same point from different directions, does the final printed point end up at the same point?).


Dan

[Captain Starfish]

Easy fixed.

1. Get a filastruder.
2. Feed the Makerbot to the filastruder.
3. Get a Rostock Max.

[barry99705]

It's not like this thing is a router. If the print head were to have lateral forces applied to it while printing, slop is the least of your worries. How much does the print head weigh? How fast would you have to move it so it would deflect under that weight? I'm pretty sure you could do this to any 3d printer.

[JohnStack]

I find on my Robo3D, there's some effector movement when I twist it a bit. The head doesn't move when I try to move it around, only the effector.

The effector plate is connected to two standard rail bearings on two rails. If I grab the front and the back and push in opposite directions, I can move the effector approximately .75mm. I think this is enough to create some nubs or uneven walls during movement at any speed.

Certainly speed has something to do with it's overall effect.

For me, the video was indicative of more serious problems for Makerbot - that's the ability of a large company to build a quality product.

[Eaglezsoar]

Easy fixed.

1. Get a filastruder.
2. Feed the Makerbot to the filastruder.
3. Get a Rostock Max.

I like your idea but we would have to chop up the Makerbot first, what would make a good
Makerbot chipper? I know, we could drop a feather on it, that should break it into a thousand pieces!

[Polygonhell]

Honestly I doubt any of what was shown on the video would dramatically impact print quality. There didn't appear to be any backlash in the mechanism, just a general lack of rigidity.
Applying forces that will never likely be applied while printing seems like a someway useless way to evaluate a product

[Eaglezsoar]

Honestly I doubt any of what was shown on the video would dramatically impact print quality. There didn't appear to be any backlash in the mechanism, just a general lack of rigidity.
Applying forces that will never likely be applied while printing seems like a someway useless way to evaluate a product

I agree but since it's a Makerbot we are just having a little fun.

[Generic Default]

Any machine will bend (deflect) when forces are applied to it, but the Makerbot in the video looked way too loose to me. 3d printers don't have to take a lot of force while operating, but the momentum of the moving parts will significantly affect print quality. To put it simply, more rigid machines make better prints.

Since the Makerbots cost several thousand dollars each, I expect excellent machine rigidity. For that much money they should be CNC machining or stamping metal for all of the critical parts.

Brian from Tricklaser on this forum made his own delta printer out of CNC'd aluminum. I don't know how much it cost him, but in general a large company like Makerbot should be able to manufacture high quality products with cost discounts coming from the quantity of machines being made rather than skimping on quality.

If I compare it to something like Taig or Sherline mini mills, there is a clear difference. The Makerbot costs ~3x as much but is probably cheaper to manufacture. I think they're just marking up their retail prices because they understand the massive demand for assembled 3d printers right now.

[RAMTechRob]

They charge that much and people buy them. I almost bought one, but decided on the Delta Orion. I am very happy with it. My buddy bought a 5th gen and already sent it back.

[jmpreuss]

I thought cartesian machines were supposed to be more rigid than that.

Here's the thing, it isn't a Cartesian, it's an H-bot (http://www.edn.com/electronics-blogs/me ... -an-H-bot- for an general explaination). I am not an expert in these sorts of things but from what I understand to get an H-bot style bot to work well you have to have very tight tolerances, which isn't the typical MBI way. They really should have gone with a coreXY.

Sad to think my rep 2 is vastly superior to the new machine.

[Generic Default]

I think an H-Bot is a type of Cartesian machine. Both Deltas and H-Bots have the advantage of not moving printed parts sideways while printing, which is important if you want to print tall, narrow objects that would be tipped over from repetitive sideways acceleration.

Also, check out this quote from the article you linked to;

Paths with sharp corners, such as a square, are challenging to accurately reproduce with a machine. The challenge resides in accurately following sharp corners. Poor implementation of the calculation of the position command causes an overshoot on the corner, yielding imperfections in the product.

That's exactly what I'm talking about with print quality and rigidity. Traditional machine design for metal cutting mills and lathes focused on tight fits and extreme rigidity. Machinists were making accurate, precise parts because the machines didn't overshoot their target coordinates and wobble. 3d printers demand higher accelerations and lower platform inertia since maintaining constant velocity is important for holding tight tolerances.

Reprapers haven't had the luxury of access to machine tools in most cases, so reprap 3d printers developed to be small, light, and cheap to make. The trend right now is for companies to make ultra cheap 3d printers to get to as many customers as possible, and to do that, they're reducing the quality of their designs to cut costs.

I want to see more well-designed printers in the next few years. Printing junky little plastic trinkets is a waste of time and energy in my opinion. I want a machine that can print itself without quality loss.

[Eaglezsoar]

I think an H-Bot is a type of Cartesian machine. Both Deltas and H-Bots have the advantage of not moving printed parts sideways while printing, which is important if you want to print tall, narrow objects that would be tipped over from repetitive sideways acceleration.

Also, check out this quote from the article you linked to;

Paths with sharp corners, such as a square, are challenging to accurately reproduce with a machine. The challenge resides in accurately following sharp corners. Poor implementation of the calculation of the position command causes an overshoot on the corner, yielding imperfections in the product.

That's exactly what I'm talking about with print quality and rigidity. Traditional machine design for metal cutting mills and lathes focused on tight fits and extreme rigidity. Machinists were making accurate, precise parts because the machines didn't overshoot their target coordinates and wobble. 3d printers demand higher accelerations and lower platform inertia since maintaining constant velocity is important for holding tight tolerances.

Reprapers haven't had the luxury of access to machine tools in most cases, so reprap 3d printers developed to be small, light, and cheap to make. The trend right now is for companies to make ultra cheap 3d printers to get to as many customers as possible, and to do that, they're reducing the quality of their designs to cut costs.

I want to see more well-designed printers in the next few years.

Well said.

[Demolishun]

Are you serious? Several thousands dollars for that? I would be better off buying a Taig CNC Mill, putting my extruder and hot plate on it and calling it good. Which is one option I looked at before settling on the Rostock. I would end up with a nice CNC metal mill and a 3D printer.

[Generic Default]

If you have a CNC mill you can make a top of the line 3d printer. I bought the Rostock at about this time last year, and although stock it performs at or slightly above the level of other 3d printers, it was never good enough for me (no machine is ever good enough for me)! I upgraded it with cold ends, hot ends, smaller nozzles, magentic arms, stepper motor dampers, ect. Right now I can print 0.05 layers in nylon for parts that almost look injection molded, and I know by comparison that my Rostock makes better parts than the $50,000 Stratasys machine that I made my first prototypes with.

But I know that filament type printers aren't being pushed to their limits right now, so I'll keep designing and making stuff to improve printing.

Also, if you did buy a Taig mill, it would probably be better to just stick with milling metal and plastic rather than putting an extruder on it (although I plan on putting a multi-extruder on my Taig mill very soon). I don't think the mill can move or accelerate fast enough to print economically, plus you have to consider the repetitive motion of infill that will wear out your linear sliders.

[RAMTechRob]

I sold a Taig mill to buy a Delta. The Taig factory is 4 miles from my house, but I never though of making a 3D printer out of it.

[atoff]

I nearly bought the 5th gen, it was on my top 3 list at the very beginning when I was set on buying my first 3D printer. Then I did A LOT of research, and it fell off the list pretty quickly. After that, I was dead set on buying a CoreXY for $3K. I'm actually very very relieved I didn't... just for the fact that I saved $2K and learned a lot more than I would have purchasing a pre-built machine. In any case, the 5th gen is beautiful LOOKING, no doubt, but how much does that matter?

[Demolishun]

atoff,
I have been thinking about this a little bit. I am guessing it works well enough for the general customers that buy them. If it priced at a quality versus price break point and people are happy with the turn key system there is no issue. I have watched the slop on my own machine and realized this is NOT a metal cutting machine. It does the job well. I can improve it because of the open design. So while I would not buy a Makerbot as I have radically different goals than someone that wants to not learn the technical side, wants a warranty, and someone else to fix it if it breaks. That is what they are really paying for.

I was in a similar situation of looking at buying an already working one. However, I did not want to spend that much on a printer. I needed a tool for making stuff and getting a kit was perfect. I also have no qualms about warranties if I modify it for my purposes. I don't think the Makerbot is really looking to satisfy the technically inclined market.

[atoff]

atoff,
I have been thinking about this a little bit. I am guessing it works well enough for the general customers that buy them. If it priced at a quality versus price break point and people are happy with the turn key system there is no issue. I have watched the slop on my own machine and realized this is NOT a metal cutting machine. It does the job well. I can improve it because of the open design. So while I would not buy a Makerbot as I have radically different goals than someone that wants to not learn the technical side, wants a warranty, and someone else to fix it if it breaks. That is what they are really paying for.

I was in a similar situation of looking at buying an already working one. However, I did not want to spend that much on a printer. I needed a tool for making stuff and getting a kit was perfect. I also have no qualms about warranties if I modify it for my purposes. I don't think the Makerbot is really looking to satisfy the technically inclined market.

I understand the want for a pre-assembled, ready to use printer, definitely... but even then, I'd probably choose an Ultimaker over a Makerbot. With the Makerbot you're paying for the name and a pretty box. For the price, I feel like there are better options out there.

[Demolishun]

You are right. This market is going to get really competitive as well too. Some of the first companies may not make it.

[notarat]

I haven't regretted my purchase of a MBI product for a nanosecond. (I have a Replicator 2)

Personally, I wouldn't purchase a 5th Gen. Not because of anything in the video, or this thread, but because it has a lot of features I don't need.

Camera? I already have 2 options for recording or streaming for each of my printers. First up is the Go Pro mounts on my Rep2 and FFCX. In addition to the GoPro, I have 2 spare cameras from my Home Surveillance system that I can view via any computer anywhere, or even on my phone.

Wireless? I hate wireless. No one is sniffing my packets. If I were to ever go wireless, I'd use the new Wireless SD Cards from Toshiba since it turns ANY device it's plugged into, into a Wireless device.

My Replicator 2 has been the most reliable, and consistently best printing unit, so far, of my 2 active printers and it "just works". When I have a 17+hr print to do, I click "Make it" and this guy appears and tells me to grab a beer.

[img]http://blog.poplabs.com/wp-content/uplo ... on_pic.jpg[/img]

[Eaglezsoar]

I haven't regretted my purchase of a MBI product for a nanosecond. (I have a Replicator 2)

Personally, I wouldn't purchase a 5th Gen. Not because of anything in the video, or this thread, but because it has a lot of features I don't need.

Camera? I already have 2 options for recording or streaming for each of my printers. First up is the Go Pro mounts on my Rep2 and FFCX. In addition to the GoPro, I have 2 spare cameras from my Home Surveillance system that I can view via any computer anywhere, or even on my phone.

Wireless? I hate wireless. No one is sniffing my packets. If I were to ever go wireless, I'd use the new Wireless SD Cards from Toshiba since it turns ANY device it's plugged into, into a Wireless device.

My Replicator 2 has been the most reliable, and consistently best printing unit, so far, of my 2 active printers and it "just works". When I have a 17+hr print to do, I click "Make it" and this guy appears and tells me to grab a beer.

[img]http://blog.poplabs.com/wp-content/uplo ... on_pic.jpg[/img]

Of all the people to pick and you had to choose HIM?
Glad that your replicator 2 works so well for you. Dependability is nice thing to have.

[barry99705]

I haven't regretted my purchase of a MBI product for a nanosecond. (I have a Replicator 2)

Personally, I wouldn't purchase a 5th Gen. Not because of anything in the video, or this thread, but because it has a lot of features I don't need.

Camera? I already have 2 options for recording or streaming for each of my printers. First up is the Go Pro mounts on my Rep2 and FFCX. In addition to the GoPro, I have 2 spare cameras from my Home Surveillance system that I can view via any computer anywhere, or even on my phone.

Wireless? I hate wireless. No one is sniffing my packets. If I were to ever go wireless, I'd use the new Wireless SD Cards from Toshiba since it turns ANY device it's plugged into, into a Wireless device.

My Replicator 2 has been the most reliable, and consistently best printing unit, so far, of my 2 active printers and it "just works". When I have a 17+hr print to do, I click "Make it" and this guy appears and tells me to grab a beer.

[img]http://blog.poplabs.com/wp-content/uplo ... on_pic.jpg[/img]

Of all the people to pick and you had to choose HIM?
Glad that your replicator 2 works so well for you. Dependability is nice thing to have.

When I was a kid I had his fishing pole!!
[img]http://www.rodmob.com/storage/post-imag ... 9871642136[/img]

[youtube]http://www.youtube.com/watch?v=NMzun_Qyrvs[/youtube]

[Flateric]

I dont really see the slop and play demonstrated in that video to be of any concern due to the lack of stresses that a 3d print head experiences.

If the head get's deflected at all during printing, especially that much, requiring that much force there is something very wrong.