small printer, runs off of 5v, fits in a backpack.

ok thats fine then, I think lets stick with the small 3 dollar steppers that come with the driver circuits. backlash can be compensated for in the programming, and on my motors the backlash is consistant.

as far as the actual structure, i think we can use a method called corexy to make an ultimaker style printer. this is where the motors are all stationary and the build volume is much more efficient. for example my big printer that im nearly done with has a 470mm^3 build volume, the total volume is 580x580x660. so its much more efficient that a mendel type design or delta, and even more efficient than the ultimaker because of the corexy instead of having the two rods that move across with the gantry.

I can post a picture if you would like.

also yes a hot end can run on 5v, specifically a 40watt heater cartridge will run at 7 watts on a 5v supply, many hot ends use 3 watt resistors so it should be enough. and i will have a small hot end we can use but im not sure of the exact price yet.

Edited 1 time(s). Last edit at 06/04/2013 10:21PM by aduy.

CoreXY sure looks like an interesting idea. Were you thinking about using aluminium, plastic or wood for the CoreXY parts?

The more I check the video of CoreXY running, though, the more I wonder how to integrate fishing line in this design. It must not "pile up" or cross over itself on the pulleys but it's so thin that I don't know if it would be a problem for only 100mm travels. We'll also need to balance the speed and strength of those 5V 28BYJ-48 steppers with the diameter of the pulleys. The bigger the pulley, the faster the travels. But also the weaker the torque.

For a portable printer I would be more inclined to think about something like the FoldaRap, the Portabee or the 3D Printer Briefcase.

Edited 1 time(s). Last edit at 06/04/2013 11:49PM by Yvan256.

it would be mostly printed parts, it works fine with belts for the cross over so i dont see the problem with using fishing line.

aduy quote 'also yes a hot end can run on 5v, specifically a 40watt heater cartridge will run at 7 watts on a 5v supply, many hot ends use 3 watt resistors so it should be enough. And i will have a small hot end we can use but I’m not sure of the exact price yet.'

I'm only offer a correction here to provide help with the heater. Wattage rating of a resistor has nothing to do with its power output. It only is a specification on the power limits the resistor can withstand in open air.

Although 3w resistors are used lets say 5.6 or 6.8ohms at 12v the power thru them at 12v is 12v/5.6ohms=2.14amps*12v=25.68watts and with 6.8ohms it is 21watts.

There does need to be a power level in order to heat up heater block to temp with power to spare. It is between 20 -25watts, although heat up time will be about 6-10minutes if at 20watts. I don't know where that study came from but it is out there and has been done already.

7watts is nowhere near the needed power level.

Perhaps set a goal of 30watts. 30watts/5v=6amps required at 5v. To get 6amps take 5v/6a= resistance total of 0.83 ohms. Granted wiring will have resistance as well so the gauge of wire will be very important for your project.

aduy Wrote:
-------------------------------------------------------
> it would be mostly printed parts, it works fine
> with belts for the cross over so i dont see the
> problem with using fishing line.

I think CoreXY used an aluminium plate because of the tension required on multiple pulleys everywhere on the plate. I guess a printed frame would probably be too weak as a completely flat plate and would require walls/angles on the sides to strengthen it otherwise the torque would twist it.

I was talking about the fishing line being tangled on the pulleys, sorry for the mix-up.

Any progress on this one?

aduy: did you continue your work on your tiny Rostock-style printer you had already started or did you start one with the CoreXY platform?

Hanic: did you receive the other 28BYJ motors, the 12V and the 1/16 ones? If you did, have you tested them?

i have not done anything else, i plan on using small magnetic balls for the joints like the berry bot if I do it delta style.

for the corexy ive been thinking about what you said, the rails will be cheap 1/4" steel rod, this will keep it rigid so that should be fine.

ａｄｕｙ，　ｉ　ｔｈｉｎｋ　ｉ　ｊｕｓｔ　ｒｅｐｏｓｔｅｄ　ｔｈｅ　ｃｏｄｅ　ｆｏｒ　ｓｔｅｐｐｅｒｓ　ｔｈａｔ　ｄｉｒｅｃｔｌｙ　ｃｏｎｔｒｏｌ　ｍｏｔｏｒｓ　ｈｅｒｅ．　ｉｔ　ｉｓ　ｕｓｅｄ　ｆｏｒ　ｔｈｅ　ｑｕａｔｒｏ　ｂｏａｒｄ．　ｊｕｓｔ　ｐｏｓｔｅｄ　ａ　ｗｉｋｉ　ｆｏｒ　ｉｔ　[www.reprap.org]．　Ｉｔ　ｉｓ　ｔｈｅｒｅ　ｆｏｒ　ｅｘｐｅｒｉｍｅｎｔａｔｉｏｎ．Ｉ used this board on one of my printers while waiting for a ramps board to show up in the mail.

ok thanks again, ill take a look when im ready to start working on this build. im still finishing up the big one so until thats done im not working on this little one.

Yvan256 Wrote:
-------------------------------------------------------
> Any progress on this one?
>
> Hanic: did you receive the other 28BYJ motors, the
> 12V and the 1/16 ones? If you did, have you tested
> them?

My apologies, I've been a little preoccupied with some other stuff and have only got back to this project over the last few days. I did receive all motors and here are my findings:

1. The 1/16 motor has virtually no torque. Whilst it does spin faster, even at rated speeds it barely has enough torque to do anything. The slightest tap and it stops spinning even unloaded.

2. The 12V variant seems to be the best performer (as expected). I could get it to move a significant mass on my contraption up to around 20rpm. Beyond that speed it's hit and miss. The torque is sufficient I believe for a delta style printer but as I'm not familiar with the kinematics of that system so I'm not sure what actual print speeds that will translate to. On a linear system I'd imagine a maximum x/y speed of about 15-20mm/s is about it, balancing pulley diameter/torque/speed. I'll try to post up a video when I get a chance of how i was testing them.

3. The standard 5V variant also seems to be quite usable but you'd have to reduce max speed a touch.

I wrote a couple of quick drivers to see if I could get a little more torque on the 5V variant by increasing the voltage and using a chopper/PWM type drive but it didn't make much difference as I believe the inductance of these motors is quite small to start with so no real advantage can be gained.

While I was at it I wrote a micro-stepping driver (which wasn't as complex as first thought). I haven't tested this yet as I don't have a suitable motor but it could make the 7.5 deg motor a real option.

I tested Teacup firmware a little more as well. I gave it a chance and tried to tweak it to get the best performance but I just don't like the way it prints. Very choppy and not smooth at all due to a lack of a good lookahead algorithm. This got me to to do a little more research into the electronics and I think I've found a couple of other options:

1. The motor drivers. TI makes a 1/8 microstepping driver that's capable of supporting 2 motors at up to 1A each in a single IC package. So you'd need 2 or these for a full printer. It looks quite affordable too at about $2-3 a unit. It's p/n is DRV8821.

2. The controller. Marlin appears to support the AT90USB variants of microcontroller. These have built in USB support so no FTDI USB/UART bridge is needed. I believe printrboard already uses the at90usb1287. The USB/UART bridge would have still been needed if using the atmega328 so would have added a few dollars anyway to the end cost. This seems like a good compromise between cost and performance.

I hope to get a little more done on it this week. I saw richraps delta variant and it looks nice and got me a little motivated to continue on this. Check it out on his youtube.

Also FWIW, my initial concept was similar to the printrbot simple type design PB Simple not build out of wood pieces but more plastic parts and maybe a couple of acyrlic sheets.. I believe this would require the least amount of material to build it simplifying the build process. A delta type offers the advantage of reduce mass of the moving parts reducing the torque requirements of the motors.

HaniC Wrote:
> 1. The 1/16 motor has virtually no torque. Whilst
> it does spin faster, even at rated speeds it
> barely has enough torque to do anything. The
> slightest tap and it stops spinning even unloaded.

I was afraid that it lacked torque, thanks for confirming my suspicions.


> 2. The 12V variant seems to be the best performer
> (as expected). I could get it to move a
> significant mass on my contraption up to around
> 20rpm. Beyond that speed it's hit and miss. The
> torque is sufficient I believe for a delta style
> printer but as I'm not familiar with the
> kinematics of that system so I'm not sure what
> actual print speeds that will translate to. On a
> linear system I'd imagine a maximum x/y speed of
> about 15-20mm/s is about it, balancing pulley
> diameter/torque/speed. I'll try to post up a video
> when I get a chance of how i was testing them.

Did you test if the 12V model is 4096 steps per turn? I know the specifications say that but in reality some models have a fractional value of steps for a full rotation.


> 3. The standard 5V variant also seems to be quite
> usable but you'd have to reduce max speed a touch.

With full step drive, it's quite usable, I think a rack and pinion system would deliver enough speed.


> I wrote a couple of quick drivers to see if I
> could get a little more torque on the 5V variant
> by increasing the voltage and using a chopper/PWM
> type drive but it didn't make much difference as I
> believe the inductance of these motors is quite
> small to start with so no real advantage can be
> gained.

I'll have to read about that trick, if we can get away with a pure software solution instead of costly current-limiting drivers it would help to lower the cost.


> While I was at it I wrote a micro-stepping driver
> (which wasn't as complex as first thought). I
> haven't tested this yet as I don't have a suitable
> motor but it could make the 7.5 deg motor a real
> option.

Software micro-stepping? Have you seen Linistepper?


> 1. The motor drivers. TI makes a 1/8 microstepping
> driver that's capable of supporting 2 motors at up
> to 1A each in a single IC package. So you'd need 2
> or these for a full printer. It looks quite
> affordable too at about $2-3 a unit. It's p/n is
> DRV8821.

DRV8821 is only available in 48HTSSOP… not the best package for a DIY solution.


> 2. The controller. Marlin appears to support the
> AT90USB variants of microcontroller. These have
> built in USB support so no FTDI USB/UART bridge is
> needed. I believe printrboard already uses the
> at90usb1287. The USB/UART bridge would have still
> been needed if using the atmega328 so would have
> added a few dollars anyway to the end cost. This
> seems like a good compromise between cost and
> performance.

AT90USB is only available in 64QFN or 64TQFP. Even PLCC would be nicer since you can get though-hole sockets for PLCC. Given the cost of the AT90USB and the FT232, I think we could look at alternatives for the USB interface. How about using V-USB on an ATtiny85 instead of the FT232? It would keep the cost at 4.98$CAD (3.54$CAD + 1.44$CAD for the ATmega328P+ATtiny85), which is well below half the cost of the AT90USB which is listed at 12-14$.

But do we even really need USB anyway?

I'd rather add an SD card, a 20x4 display and a navigation keypad. There's plenty of 20x4 displays on eBay for under 7$, it would make the printer more portable if it doesn't need a computer attached to it. It would probably only add around 10$ to the total cost for the display, SD socket and navigation controls.

Also, given that full-color TFT 128x160 LCDs are now under 7$ too, including an SD socket on the back, I think firmwares should start supporting those instead of fixed 20x4 characters displays.


> I hope to get a little more done on it this week.
> I saw richraps delta variant and it looks nice and
> got me a little motivated to continue on this.
> Check it out on his youtube.

I just checked it, and while the idea of using printed parts is nice, I still think a better option would be to only have flat parts, that way you have more options to build your own machine: CNC cut parts, lasercut parts, 3D printed parts or even hand-cut parts using printed paper overlays glued to the sheet material.

Edited 1 time(s). Last edit at 06/27/2013 12:26PM by Yvan256.

yeah i think you guys are taking this in a direction that i dont want to go in, so im out.

aduy Wrote:
-------------------------------------------------------
> yeah i think you guys are taking this in a
> direction that i dont want to go in, so im out.

I think we're still talking about portable and low-cost, is there a particular item or feature you don't like?

I only have experience with CNCs, so I might make wrong assumptions on things like firmwares and hot ends for 3D printers, otherwise the mechanical aspects are pretty similar for both types of machines.

It's your thread and your idea, if anything we should be the ones to be "out" if there's a problem. Me and HaniC probably took on the "low-cost" idea a bit too far, maybe we should start our own thread and let you be with your portable idea, I think that's where the directions differ.

Edited 1 time(s). Last edit at 06/30/2013 11:07AM by Yvan256.

I am currently looking into what has been described in this thread. I also would like to build a low cost small scale extruder type. I know this thread is old but I haven’t found anything similar. I am interested in using rack and pinion gearing as well. I have always wanted to implement this with 3d printing as it is so prominent in CNC design and seems very rigid. In that same regard I favor using aluminum extrusion though it probably is not best in this situation. I want to build something cheap to get started with understanding the design process of 3d printers. The last thing mentioned was starting a new thread. I was unable to find it. Has this concept ever been implemented? Is anyone still interested in this?

Someone finally made a cute delta printer using these 28byj-48 tin-can motors at $1 apiece.;-)

Quote
aduy
you know i was also thinking about cheap servos, but the issue is that they can only rotate so much, but if you took the pot inside the motor and then replaced it with a long slider, then it could accurately position itself on each axis. well the only problem is finding long cheap sliders.

What about the encoder strip that inkjet printers use?
[2.bp.blogspot.com]

Hello, I realize this forum post is two years old (May 28, 2013 11:37AM), but did this project go on to completion... Is it available? I have a high interest in this area. I teach interfacing with Arduinos and am playing around with the idea of building a Prusa i3 rework a piece at a time (working on a teachers budget @grin@) -profharris

As vn mentioned, the only thing that seems remotely similar in concept seems to be the Tiko 3D printer.

I'm still playing around with geared stepper motors with a rack/pinion but only for a miniature briefcase-CNC now.

just posting code for canned motors. here is the link

[reprap.org]