Building the electronics, part I

Stage 1: Getting the parts

While we were planning the purchases for our mendels, we wanted to solder the electronics ourselves, because we thought it would be cheaper that way. But it turned out that the do-it-yourself kits from makerbot are much cheaper than the PCBs and components, even with import tax, and since the PCBs are only available from makerbot anyway, it was clear that we would get the kits.

When I made the cost calculations I noticed that the complete kit would only be 20$ more, so we decided to get them complete as soon as they were in stock.

But I wanted an extra extruder controller to be able to work on my granulate extruder or maybe a frostruder or whatever came to my mind, while still having a functional extruder. So when the electronic kits were back in stock I waited a little to see if the extruder controller would come in as well.

20 hours later ALL electronic kits were sold out. I was really pissed!

So we went back to plan A and ordered the PCBs from makerbot, the components from mouser (because they are included in the reprap parts lister, which makes shopping so much easier) and the chips from farnell (because I know them from work and they had those in stock). In the end the costs were roughly the same as the complete kits, but since I didn't expect makerbot to get them back in stock soon, there was nothing else to do.

But of course I was wrong. Even before the package from makerbot was here, they got a whole bunch of electronics back in stock.

So what, we had already ordered everything, there was no going back.

Stage 2: Soldering

Since my old soldering iron was broken and the new one not really suitable for smt soldering, I decided to try the hot plate reflow technique.

I started with the extruder controller, because we have one of those extra. I applied the solder paste and set the components. That was not so difficult, even if I don't have the best equipment.

The only problem was the 180 ohm terminal resistor, the documentation says this is needed for one extruder in the system, but it is not listed. However it is listed for the mainboard, where it is optional, so it's not so bad. But I still think it should be the other way around.

Another problem was because I had to order other LEDs, the green ones I got are a little confusingly marked, so I put them on the wrong way.

For the hot plate I prepared my multimeter with a thermistor and monitored the temperature of the plate, which was a good thing. It should not get over 200°C to save the epoxy of the PCB, but more than 180°C are needed to melt the solder paste. I think by now I know the correct setting, but the first try should not be attempted without a thermometer.

After soldering in the through hole components the controller was finished and ready for a test.

Stage 3: Testing

I had to dig a little in the wiki to get the procedure for programming, but after I got the arduino software set up I was ready.

I connected the USBtinyISP to the board and burned the bootloader, which seemed to work well. (This was the point where I noticed the LEDs were not working.)

I tried to connect the serial cable, but I could not programm the controller. I tried the USBtiny again just to be sure the chip was ok, and it did not work any more. The interfaces are rather simple, and we tested the circuits and could not find anything amiss, but I must have shorted something somwhere so the chip was fried when I connected the serial cable.

I'll try a stepper driver next and I hope that works out better.

Granulate Extruder

The possibility to use plastic granulate instead of filament would be a great addition to the reprap project. Granulate is the usual form in which plastic is available anywhere, and it is much cheaper than filament.

Because of that several approaches have been discussed, for example in the wiki, blog and forum.

(picture from reprap wiki)

The solution presented in the wiki looks most promising to me (especialy because it is confirmed that is basically works).

I want to make some experiments on this, and the problems I see now are the following:

1. The auger

This design uses an auger to transport the granulate to the nozzle.

Plastic granulate is available in various sizes, but the original ware should have a size of 3-4mm, so I need an auger wich is able to transport this size. Assuming the core is one third of the diameter, that would be 12mm. To be on the safe side, I would say a diameter of 16-20 mm should be ok.

I had no luck finding an auger of that size available anywhere, so I fear this would have to be a custom job, which would not really fit into the philosophy of the reprap project.

For now I'm going to try my luck with the second best I could find to replace the auger: A drill, "Schlangenbohrer" in german, I think it's "auger bit" in english.

I got three versions, 12, 16 and 20mm. It seems the 16mm drill should be big enough, maybe even the 12mm one.

Of course the drill must be rotated counter-clockwise, but I don't expect any problem there. I don't have any knowledge of the dynamics of liquids and augers, but the tip of the drill doesn't looks like it will be ideal to generate the pressure needed to extrude the plastic. Well, we'll see.

2. The barrel

The auger sits in a barrel, which has the nozzle at the end. That end is heated to melt the granulate, so the upper end must be cooled to prevent the melting of granulate in the reservoir, I hope a passive element is sufficient there.

I would like to have a screw thread at the ends of the barrel, so I can fasten the nozzle easily to the barrel as well as the barrel to the reservoir.

Until now I have not been able to find a threaded tube of steel or aluminium which fits one of the drills, or even a tube that is thick enough to be threaded.

3. The reservoir

The current filament extruders have virtualy unlimited material, because the filament can be received from a spindel next to the reprap. A granulate reservoir would have only limited storage of granulate. Also it can't be very large because it would become too heavy and bulky.

So we might need a way to refill the reservoir continuously, or to stop the reprap when the reservoir is empty and continue when the operater refilled it.

For the first experiments I'll settle for small parts, which don't need any refill. It should not be impossible to solve this problem elegantly later.

This is how far I have thought until now. In any case it will take a while before I can even start to experiment, I still need to build my reprap before that.

another Reprap addict...

Hi, I'm Alex, m, 28 , 'Informatik' student (computer science), from Germany
When telepath first told me about '3d printing@home' I wasn't sure 'bout it... it looked interesting but rather complex for a 'normal' person ( ... let's better not discuss normality). I read the articles about it and went on.
As he presented me the project a 2nd time, I was rather impressed, not only the project itself had matured, but there were also lots of interesting side-uses for some of the components (maybe later more about that, I have to prioritize). And since you have to start somewhere if you dont want an everlasting 'chicken-egg-problem' one of the possible side uses becomes a main use: A small PCB-milling machine driven by the Gen3 electronics as a RepStrap.
First we'll focus to use it as a RepStrap only and after we've build our own mendels maybe I'll convert it into a small pcb-mill.

Introduction from telepath

I've been fascinated by 3D printing ever since I read about Fab@Home some years ago. Just a few weeks ago some impulse made me look up the topic again, and I noticed that the project was as cool as I remembered, and by now you can even buy a fully assambled Fab@Home, but it's still much too expensive to make one myself.

(picture from fab@home wiki)

But I also found an article (german) which mentioned another project which they called "not so advanced", the reprap.

So I had a look around the wiki and was astonished, it's SO much simpler and cheaper than the Fab@Home. And since the article was from 2006, the "not so advanced" project was much more mature by now, with the mendel just being released.

(picture from reprap wiki)

Since then I was certain that I would follow the project, but as any hobby it's boring to do it alone, so I told my friends about it and one of them, Alex, was just as enthusiastic as me (well, maybe not quite as much), and so I got a good friend on board who has experience with atmel chips and stepper motors and who has been tinkering with a small mill for wood and circuit boards.

By now we have ordered all the parts for the electronics and mechanics for two mendels and are basically just missing the reprapped parts. Since it's a major problem finding someone who is willing to print the parts and has not already a bunch of people waiting, we have been thinking of a repstrap design, and I think we are on a good way. We have many diferrent opinions about which way to do something or how to solve a problem, but I think that will help us very much to find a good solution in the end.