The TangiBot 3D Printer, a clone of Makerbot Replicator

[nctnico]

Poor form, probably. Legal, certainly. But I think it's moot.

I'd bet that well before Christmas 2013 some "Shine Happy Lucky Joy Light up your Butt" Chinese company will have a total clone, with a nice molded plastic covers, and a better option than plywood for a chassis, on the market for well under $1K.

And it will 'not quite' work without extensive fiddling, and come with a piss-poor, unsafe power supply covered with fake approvals marks.

A quick look on Ebay tells me that there are already many 3D printers for sale for a low price. This thread is about getting something decent for an affordable price.

Edit: Look what I found on Ebay:
http://www.ebay.com/itm/Fully-Assembled-Thing-O-Matic-3D-Printer-thingomatic-Makerbot-Design-Reprap-/251126337294

[T4P]

Look on the back cover and you will know why it became so cheap ... "MADE IN CHINA"

So what ?

Poor quality ensues, planned obsolescence

I will not even try to argue about this, apparently you're one of those guys who still have not understood that "made in china" means nothing. Apple products are made in China, so are most cheapos low-end computers, and they are not the same thing. To reuse one of Dave's favorite expressions, this is really an attitude that ticks me off.

Which explains why apple and dell plus lenovo is low quality, your arguments?

Look on the back cover and you will know why it became so cheap ... "MADE IN CHINA"

So what ?

Poor quality ensues, planned obsolescence

Many countries started up like the Chinese. The Chinese only imitated them. In 1950's a Japanese product meant cheap, bad quality product. Then Korea followed the example, then Taiwan. Is a Japanese product a bad quality one today?

Japan was very quick at improving but china? What changed in the last 10 years?

[casinada]

http://store.solidoodle.com/index.php?route=common/home

$499  to $599

[gregariz]

Japan was very quick at improving but china? What changed in the last 10 years?

We've now had 35 years of capitalism in China. 35 years after the war we were generally getting good quality out of the others. It's on quite a different road. Overpopulation is a serious issue for them. We can't buy enough of their crap to make them all wealthy. In the meantime, manufacturing, engineering and hence employment growth around the world have been seriously hurt because they are attempting to sell their way out of poverty by always being the cheapest guy in the room. China version 2 doesn't appear to be on the horizon.

[nctnico]

Companies are moving to countries like Cambodia, Laos and Vietnam because China is getting too expensive and political stability has greatly impoved in those countries.

[caroper]

Japan was very quick at improving but china? What changed in the last 10 years?

RIGOL

[nctnico]

We are getting a bit offtopic here but the way this thread is going reminds me of a business trip many years ago. I had to go to Motorola in Malaysia for some tests with a radio system. While waiting I noticed a time-line on the wall. It appeared Motorola gave two Malaysian engineers a simple project in 1978. They did very well. From there on that particular Motorola branch evolved into a large manufacturing and R&D site. I've met some very clever engineers over there (most of them where from Chinese origin). And Motorola was not the only high tech company I spotted over there.

[T4P]

Japan was very quick at improving but china? What changed in the last 10 years?

RIGOL

In general, i know rigol and owon changed alot but the rest? Not really

[RJSC]

It's time someone offers an option move on past plywood! (Be it Acrylic, Aluminum, etc).
I've hated makerbot industries plywood design from the first second.
It's fine for an internal prototype or homebuilt clone, but for a commercial product?  WTF?

[amyk]

It's time someone offers an option move on past plywood! (Be it Acrylic, Aluminum, etc).
I've hated makerbot industries plywood design from the first second.
It's fine for an internal prototype or homebuilt clone, but for a commercial product?  WTF?

I agree. Depending on the process used, wood might even be more expensive than plastic or metal, and it's not as stable; not very suitable characteristics for the chassis of a machine that needs accurate, repeatable positioning.

[Psi]

One thing I would also like to see, (and im not sure if this has been discussed to death), is feedback on the stepper motors.

It seems to me that, no matter how good the current version is, motor feedback would make it even better.

[amyk]

One thing I would also like to see, (and im not sure if this has been discussed to death), is feedback on the stepper motors.

It seems to me that, no matter how good the current version is, motor feedback would make it even better.

Yes, that came up in Dave's review thread too. I also suggested not using steppers but regular brushed DC motors + feedback like they do in inkjet printers now.

[Psi]

One thing I would also like to see, (and im not sure if this has been discussed to death), is feedback on the stepper motors.

It seems to me that, no matter how good the current version is, motor feedback would make it even better.

Yes, that came up in Dave's review thread too. I also suggested not using steppers but regular brushed DC motors + feedback like they do in inkjet printers now.

Maybe even 3phase RC motors, as then you'd have better control of the cycles/speed.
And they can handle quite a bit of speed/power.
You'd just need to insure that the down gearing doesnt introduce play in the system.

[RJSC]

One thing I would also like to see, (and im not sure if this has been discussed to death), is feedback on the stepper motors.

It seems to me that, no matter how good the current version is, motor feedback would make it even better.

Yes, that came up in Dave's review thread too. I also suggested not using steppers but regular brushed DC motors + feedback like they do in inkjet printers now.

Do not forget hat takes the software part to a whole new level of difficulty!

You have to do complex calculations to consider the mechanical part of the system: the inertia of the axes, the friction, etc, or else you get a positioning system that takes to long to reach its intended point, or overshoots and and has to come back a little, which leads to disastrous results on an extrusion printer.

Inkjet printers get away with it because they just run the motor, read the encoder tape and fire the ink droplets when the print head is passing the right spot. In no way they stop the motor at a precise position and go back, like a plastic extrusion printer needs to do.

[Rick]

Japan was very quick at improving but china? What changed in the last 10 years?

RIGOL

In general, i know rigol and owon changed alot but the rest? Not really

OK but where they want to improve they do improve.
If we do not see a world war or some other global cataclysm, they will catch up with USA in technology. They already launch more space vehicles per year then Russia. They have their space station, by the way who else has? We have to face the truth one day...

[free_electron]

You know what would be an i provement to this class of machines ?
One that actually works....

-does not run its head into the plate

-Has a self leveling build platform

-Has an installer that works without asking you to download 3 other packages with no info on what version you need or where to put it ( see my other posts as well as a post on makerbot forum about this. You need to download python but no info if you need 2.xx or 3.xx , installer doesn't work , unsigned drivers etc. make an installer that is self contained and does everything as opposed to having to 'figure it out')

- does not shake its guts loose and destroys itself when running the demo file ( like dave did )

- is made from a stamped metal chassis instead of stupid plywood or acrylic. That is good for covers only. The guts need to be solid and inflexible and rigid

- has feedback on the steppers. Ive seen too many builds that somewhere in the middle skipped a detent and go all wonky...

- doesnt use 'printed bits'

I was at GoEngineer a few days ago to look at an industrial 3D printer (stratasys) .. Those don't use heated platforms... The whole build chamber is heated ! They don't rely on botched together scripts and finnickey broken-source but have proper CAD software drive them.

If you put those machines as the reference then the makerbot is of the 'dirt-under-the-carpet' level...

So, if the tangibot can improve on the points i listed above, i'd call that innovation and gladly give them my money.

[Erwin Ried]

You know what would be an i provement to this class of machines ?
One that actually works....

-does not run its head into the plate

-Has a self leveling build platform

-Has an installer that works without asking you to download 3 other packages with no info on what version you need or where to put it ( see my other posts as well as a post on makerbot forum about this. You need to download python but no info if you need 2.xx or 3.xx , installer doesn't work , unsigned drivers etc. make an installer that is self contained and does everything as opposed to having to 'figure it out')

- does not shake its guts loose and destroys itself when running the demo file ( like dave did )

- is made from a stamped metal chassis instead of stupid plywood or acrylic. That is good for covers only. The guts need to be solid and inflexible and rigid

- has feedback on the steppers. Ive seen too many builds that somewhere in the middle skipped a detent and go all wonky...

- doesnt use 'printed bits'

I was at GoEngineer a few days ago to look at an industrial 3D printer (stratasys) .. Those don't use heated platforms... The whole build chamber is heated ! They don't rely on botched together scripts and finnickey broken-source but have proper CAD software drive them.

If you put those machines as the reference then the makerbot is of the 'dirt-under-the-carpet' level...

So, if the tangibot can improve on the points i listed above, i'd call that innovation and gladly give them my money.

The purpose of the RepRap/Makerbot project is to allow users to have 3d printers in their house. Can you imagine how much money you will need to have the improvements you say about stratasys, probably 10 times the price of a Tangibot.

[EEVblog]

So, if the tangibot can improve on the points i listed above, i'd call that innovation and gladly give them my money.

If Tangibot did only one of those things, no one would have complained.

I think the two biggest improvements would have to be a nozzle distance sensor, and some way to ensure the platform is always flat.
That would likely come easily with a more rigid and closer tolerance frame.
The absolute positional feedback would be great, but is a massive can of worms. Simply having the unit know how high the head is would be a big improvement.
And then there is all the firmware and software improvements that would make it a much better tool, without changing any of the hardware.

Dave.

[nctnico]

I don't think it is very complicated to add a optical feedback system to a stepper motor. You don't need full resolution. An indicator telling the motor has missed a step is enough. When using stepper motors the current through the motor should be more than enough to ensure the motor doesn't miss a step in de first place.

I agree about the casing. How much more expensive would an aluminum or steel frame be? It doesn't need to be thick as long as its stiff.

[merser]

I agree with those who remarked about the plywood chassis. It is so "look what pop made in the back shed". Offering it in acrylic is a slight improvement but still hobbyist looking. Only one of the many reasons I don't see a 3d printer as a worthwhile investment yet.
If makerbot can innovate then they should be able to stay ahead of the competition or clones. If you can't innovate then pick a market that's less competitive.

[EEVblog]

I agree with those who remarked about the plywood chassis. It is so "look what pop made in the back shed".
Offering it in acrylic is a slight improvement but still hobbyist looking. Only one of the many reasons I don't see a 3d printer as a worthwhile investment yet.

I think the plywood kinda looks funky and gives it a unique look, but certainly understand the "home made" feel people get.
Acrylic is probably a very bad idea. It it's clear then you can see all the guts and ugly wiring etc (just look at the Thing-O-Matic). And if it's opaque colour then it's going to look pretty gimmicky and "plasticy". I'll reserve final judgement until I actually see one, but I think it's most likely going to look pretty silly.

Dave.

[merser]

Yeah, I agree acrylic probably isn't the solution. I reckon if they just gave the ply a nice 2 pak finish it would make a better impression and cover up that burnt smell, lol.

[joseph.anand]

The only thing keeping me from buying any of the existing 3D printers or CNC milling machines is the absence of a feedback mechanism. For the same price, I reckon I should be able to make one with a feedback mechanism. For the time being, I am stuck with deciding on a rotary encoder with suitable gear ratio or a linear encoder for feedback. Any pointers anyone?

[Galenbo]

Why not print a 3-D printer?
So you have a free one :-)

[robrenz]

The only thing keeping me from buying any of the existing 3D printers or CNC milling machines is the absence of a feedback mechanism. For the same price, I reckon I should be able to make one with a feedback mechanism. For the time being, I am stuck with deciding on a rotary encoder with suitable gear ratio or a linear encoder for feedback. Any pointers anyone?

If all you care about is knowing if your stepper missed a step,  ignore all of the following and put a encoder on your stepper;

Assuming you are going to do a true servo control loop on the axis and you care about bidirectional accuracy and  repeatbility:

1. Rotary encoder with gearing is the worst. the backlash in the geartrain will add a large amount of position error on top of #2 below. Also will make servo tuning to a fine level next to impossible.

2. Rotary encoder directly on the leadscrew or pulley of a belt system.  Still has the play or reversal error of the leadscrew/nut or pulley/belt combination.

3. linear encoder is the best. It measures the position directly and bypasses most all mechanical innacuracies of the axis. (there are still a huge number of things to consider to get the most out of the linear encoder feedback).  It does not mean you can have a sloppy mechanical system. You will never get the servo loop tuned if the mechanics are sloppy. You may also need to have a separate motor encoder for closing the velocity loop of your control system independent of the positional control loop that is closed by the linear encoder.