Future Feature Idea: No PC printing.

An interesting feature would be to create an *optional* circuit card with another microcontroller and some memory (ram, flash, eeprom, USB flashdrive). It would connect to the Powercomms PWA where the PC currently does. Lets call it a Autonomous Operator PWA (AOP).

1. Connect the PC and play the serial data stream out into the memory of the AOP. *
2. Disconnect PC and power down to save power.
3. Press the start button on the AOP to start the print job.
4. Wake up in the morning or come back from work to your newly created object.

*could also allow the PC to be in a different location than the reprap or left connected all the time.

Thoughts?

Disclaimer: This is in no way meant to be a near term idea but rather an add-on once working Darwins exist.

Problem is that when you expand your controller enough to take in memory, you have spent just about enough to buy a cheap PC.

OTOH, if you don't mind throwing in a few hundred dollars for a low energy drain byte bucket you might want to look at the Gumstix boards.

[gumstix.com]

What would be your rough guess as to "enough" memory?

I like the idea and have been working lately on a usb/ pic device to run G code files using a host/ slave controller such that it can connect direct to a PC or read data from a USB stick giving it up to 4GB of RAM.

If i get this going I'll put up the information as a small firmware change should make it able to run RepRap but don't hold your breath!!!! it's a little way of yet

I think that

19200 baud
1920 bytes per second
115200 bytes per minute
6912000 bytes per hour
165888000 bytes per day
162000 kB per day
158.2 MB per day

Unless it takes longer than 6 days to print a given job a 1GB flash storage solution should do it. This is assuming 19.2 kbaud and 100% utilization. Of course this means you need to have a different way of loading the memory than 19.2kbaud serial .

Hopefully my calculations have no glaring errors.

I used to have a bad concience for using two old PC's as a firewall and server at home but they probably use less juice than the new ones...

Anyway I think off-line printing is a great idea. Once the first weeks reprap thrill has worn away, you might want to stove your printer away somewhere not in the middle of the livingroom floor.

I have been working some with a set of small micro-controllers from a company called Z-world. If you just buy the core module you can get well below $100 for some 512K-1M of memory. Check out www.rabbitsemiconductor.com for ideas.

The instruction set is similar to that of Z80 controllers but with enhanced instructions and a lot more Megahertz than you had back in the dark ages.

You can also check out www.z88dk.org for a free and open-source C-compiler that now (1.7 release) supports the Rabbit 2000 & 3000 architecture!

There is also a commercial compiler called Dynamic-C.

For the future i see a non PC reliant RepRap as essential to being able to be taken up in the developing world, cheap USB sticks could be loaded at full USB speed with files that would then be read by the onboard chips this would be a processed file i.e. not a native STL so the computer would still do the hard number crunching and the USB is simply a storage device. These processed files could even be download directly online, such that the final manufacturer could load their USB at a internet cafe etc and require no local computing

I think some kind of primitive user interface would be needed. Say the USB stick contains the receipt for one hundred different parts that are needed in the village-whatever-wherever equipment.

Assuming a small malfunction and the whole thing is reset.

It would be an absoulte necessity to be able to instruct the machine to start printing on item, say 63.

... why not use PDA's or smart-phones?

Then you need only an USB- or Bluetooth-port and the software runs as native java or C on the handheld ...

The actual smartphones or pda's are 'real' capable Hardware and should be much faster with more RAM in the future.

Ciao, VDX

I've actually always thought of it more as a network device... they've got networked printers... why not networked 3d printers? fill the memory over the network and schedule/monitor the jobs over that network as well... add wireless for convenience...

yeah, i agree with lordcat. the best way to go would be some sort of networked print system. the computer is still an important part as an interface to the machine, but having the reprap machine have its own internal buffer where it stores the data to be printed would be great.

something like this might be cool:

* your reprap printer is connected to computer via USB.
* you send the data to the printer, and it spools it to disk, just like a normal printer queues jobs.
* the reprap machine then writes that to some internal memory (perhaps we could have a SD / USB slot where you could plug in the memory of your choice.
* it then commences printing, and you are free to do other things with your computer, or even shut it down.

that would be awesome!!

I was thinking even more in third-world terms.

If we really get reprap going then the computer might become the larger cost, we consider the computer as a commodity but maybe the usb sticks could be used as carriers of "manufacturing receipies" so reprap:ing could take place at sites which lack computers...

Hi mimarob,

... but you need a computer in your reprap, which can read the USB-stick or SD-dard ...

I asked previous, why not use PDA's or smart-phones?

Maybe in third-world countries the smartphones are more distributed, than capable PC's?

Then a reprap without an own computer, but with an USB- or Bluetooth-Interface, which goes directly to the motor-drivers and a smartphone with the software on it would go better?

Ciao, Viktor

You can read a USB stick with a micro controller that has a host USB port, e.g. Freescale MCF5373 is one I used recently.

---

[www.hydraraptor.blogspot.com]

... on one of the sides you need a computer capable of slicing the STL-data in path-informations for the XYZ-stages and the dispenser.

Either you convert on a desktop-pc the STL-file in ncp- or G-code (or whatever the reprap needs), then you receive for a 'typically' 3D-object with 0,1 mm resolution many hundred megabytes of code ...

Or you transfer only the STL-code and/or the corresponding information how to build and which material-/timing-data, and so on - then the reprap has an expensive computer which calculates the path-data from the STL-file ...

In both alternatives you need an expensive PC, which isn't really cheap enough for the third-worlds ...

I think, a smartphone with the capability of receiving, calculating and outputting the data is more realistic, than the distribution of desktop-pc's.

My Acer n30 navi-PDA has an USB-port and with a 512MB SD-card i have enough capacity for the software and processing of big files ...

Some of the actual smartphones have faster cpu's and more ram, so it's a logically consequence for me, that the handhelds should be on focus for an third-world-seeding.

Ciao, Viktor