Why not simplify ?

Hi,
I've just found this site (via instructables.com) and I really like the idea of prototyping my own projects, I'm currently building a cnc mill, but a reprap would surely update any workshop

okay enough intro stuff let's fly.

I've read here and there browsing your site and was wondering why things were done the way they were.
for example: why building nozzles for the thermoplast extruder using nuts and bolts and not just using readymade carburator-nozzles like that:
[www.dna-tutorials.de]
there are also injectornozzles with a fine tip if that is relevant.
They can be changed easily, and are available in all kinds of diameters.
(almost as cheap as a brass rod and nut, but very accurate)

Also there are lots of cnc controllers (_all_ kinds of prices) and freeware controllersoft out there, what will maybe improve builing time for a working reprap
(okay; it may be more expensive, but saving time is valuable too )

There are many things that could be work- and time- optimated by just adding a few pennies.
So why not choose the beaten track?

hand 'sid

If I'm done with my mill, I'm starting on reprap I guess

Well, the problem with that is that by the time you've added enough pennies to make things really easy you wind up with a CNC mill with an extruder for a toolhead which costs several thousand dollars. Pretty much anybody who wants can do that now and several people, notably Sebastien up in Canada, are doing just that.

The trick to RepRap once the Mk 2 extruder was invented by Dr. Bowyer is to break the back of the cost of the cartesian positioning system. We're trying a bunch of different technologies and design approaches to do that.

If you want to do the CNC route, be welcome! If you discover something on the journey to doing that which is useful that's better still! :-)

True, I understand what you trying to say...

BUT:
I just checked some prices on carb-tipped-nozzles 0,85 Euro each..
You need one, maybe a second or third in a different size.
How long does it take to build one or even three? One hour for three nozzles?

I do not talk about the whole project or such, only about small parts of it.
gearwheels nozzles and so on.

Okay, a readymade controllercard maybe costlier than a homebuild,
also I've seen one for about 60 Euros (emis, smc800) a programmable microcontroller is over 35 Euros, plus all the tiny little stuff that you need else for the controller... (will be a difference of about 5 Euros if I got it right, and thats without programmercard for the microcontroller)

I very much appreciate all "selfmade" things, but I do like to simplify the workprogress whereever it could be simplified,
and adding about 20-30 Euros (so far) for saving about three or four hours seems to be a fair deal

Nevermind, I was just thinking aloud

hand 'sid

Well, let's see.

Nozzle... For the extruder barrel I drill a hole in the tip of the extruder barrel that is made of a piece of 5/32 hard copper tube (costs about $1.50) and a piece of copper shim strip braised to it. A #76 wire gauge drill (0.5 mm) costs about $1.69 and, if you are careful, can be used many times.

I understand that people may want ot use wire pull dies like you describe. I'm a little worried about using them, though. My extruder barrel design emphasises low thermal mass and close proximity of the extruder orifice to the heating coil. Those tips look to have a considerable amount of metal in them and, as well, protrude away from the barrel. That makes it hard to get heat from the heating coil that I put around the barrel out to the tip. Unless I can keep that top quite hot it is going to be hard to extrude plastic out of it. I'd have to pump up the temperature in the barrel to keep the tip hot. Since the barrel is already heated to 200 degrees C. that gets a little difficult.

The total parts cost for my microcontroller board is about $25. $10 of that is for the board itself (Euroformat stripboard). The last 18F4610 microcontroller I bought cost me $6.85. The H-bridge chips cost $1.75 each. The next most expensive piece was the screw terminal connectors at about $0.50 each. I have about $7.50 worth of those on the board.

The axis limits detectors cost about $2.50 and the shaft encoders cost about $5.

I can understand your confusion about making things like gears by hand. I don't quite understand that one myself. I don't do it, either. I do know, however, that while gears are easy to buy it is sometimes very difficult to buy a gear that mates properly with the drive shaft you need to use. Looking at the work on Darwin I am wondering if that is the problem that Ed encountered that caused him to resort to molding his own drive gears.

thanks for the suggestions! after my failed attempts drilling out the nut, i've looked into alternative nozzles. i'm definitely interested in these... it seems like this would be an ideal alternative to doing it at home. especially if its so cheap.

as for the electronics... we're ordering printed boards and will be able to simply provide them for super cheap. then its just a simple soldering session and you're good to go. we're looking at around $60-80 for complete electronics. most controllers are double that. once the project really takes off, those costs will come down dramatically.

as for the controller software... well, we just decided it would be easiest to write our own, especially since we're interfacing with our custom hardware. it gives us more flexibility to change and add additional functionality to the machine (multiple heads, materials, etc.)

of course... none of this is set in stone. if you'd like to make your reprap based on commercial hardware, theres nothing really stopping you from taking the 3d files and running a STL -> gcode script on them to generate paths. actually, that would be really useful, but i dunno if theres any code out there that does that yet.

Hi again,
Depending of "where the nozzle comes from" those are between 85 cent
(scootercarb; the idle thingy.. i think mikuni or dellOrto I can't remember) and about 5 Euro for a new Weber idf main-nozzle incl shipping (I got me some used nozzles once for my weber 44 idf for 50 cent each at my carb service)
Used nozzles are available at almost any carb-service.
diameters I know are between 0.6 and 2.1 mm but I think there is a much greater range.

And yes if you hit the $80 mark for a complete electronics-kit I'm very impressed and didn't said a word about it
Hmm, yeah I didn't thought of your custom hardware that's why I mentioned controller software, you're right, writing your own does makes sense... sorry.

Oh, just one last idea was popping out of my mind yesterday...
the "selfmade" MagServo..
isn't it possible to build a selfmade mechanical stepper?
basically a gearwheel attached to the shaft switching two microswitches on and off (OR setup) sounds like an easy setup to me; will have to try i think.
depending on number of cogs you can adjust steps e.g: 90 cogs == 180 steps (2

Very interesting idea for nozzles... thanks.

on the magservo part... check out our wiki... theres a couple writeups in there on turning a cheap dc motor into a stepper. we're planning on using those in reprap v2.0 due to cheapness, etc, but decided to opt for steppers just because they are essentially plug and play.

forrest is actually using them on his and apparently they work very well... magnetic shaft encoders with 1024 step resolution. very nice and also pretty darn cheap. it makes alot of sense to hold off until v2.0 since we'll be able to print off custom holders, etc for the chips and such.

The only downside to the Austriamicrosystems AS5035 magnetic shaft encoder is that it is about the size of a showgirl's sequin and has surface mount pins about as big as the legs on a tick. It's a real pest to work with without a surface mount workstation.

are there any larger magnetic encoder chips out there? its hard to imagine that NOBODY would produce one

Imagine it. I've googled for hours looking for an alternative. I think that perhaps Austriamicrosystems has a patent or something of that.

hmm. i'll have to get some and play around with them after v1.0. did you hand solder them, or use the toaster oven trick? i bet a syringe + solder paste + toaster oven would work pretty well.

One of the problems is that the price has been skyrocketing due to appreciation of the Euro. They're running about $10 now. The nearest packaged shaft encoder costs about $30. We're getting pretty close to a tipping point.

Hi again..

what about optical or mechanicla encoders like the ones from cui.com

I'm still searching for some old R/C car in my cardboardboxes for trying out that gearwheel thingy...

A cheap stepper (e.g. SM 42051 from emisgmbh.de) starts at 18 Euros (1.8

Why not cannibalize scanners or plotters for most of the mechanical assemblies and drivers boards? My school district has literally dozens of scanners lying around, all HP, all working except for PC driver software. eBay has such models for well under $50 each.

I think there should be a balance between the ideologies of self-replication and "social replication" -- how can we best cause rapid uptake of this technology by millions of people who want such a gadget (see the Sears exploration of this market) but cannot build everything by themselves.

How about an open source hardware protocol that requires everyone licensing the technology to make additional components for sale or trade via the reprap community? That way, those who want to break even can do so, those who want to support themselves by supporting reprap can do so, and those who want to buy components and assemble it themselves can do so as well. Look at the many flavors of Linux, from geek-only to easy-for-newby.

Hi again...

I was just fixing my old varaflame (wasn't gastight anymore n stuff.. nevermind)
The gasnozzle is tiny, So I searched my trahcan for an empty "oneway"-lighter to rip it apart.
In fact, If you remove the rubbergasket from the inside you'll have a very tiny threaded nozzle.
This is way cheaper than any carbnozzle (it costs NOTHING!)

Oh... and regarding my "mechanical stepper" I've tried... works... not long
problem is the lifespan of the microswitch I guess 1mio times lifespan....
(assuming a low resolution : ) 60 times per revolution and let's say 2.5mm travel per rev, the switch will last for about 6.66 meters... that's awful! why didn't you warned me about that?
So.. forget about it...

'sid

Edited 1 time(s). Last edit at 04/11/2007 08:13PM by sid.

If you had asked many of us personally, we might have told you switches don't last very long for any number of reasons. But if all of us listened to conventional wisdom all the time, we'd be saying .. reprap what's that.. no.. the cheapest FDM is around 15000. Definitely not for the hobbyists. I like hearing what others find out, good, bad or ugly. And some of it is darned nice Keep up the experiments and nice find with the nozzle. How about a non mechanical switch like an opto isolator like we use on the limit switches, just an idea

thought of that "mechanical" switch because it holds the gearwheel in place (with a minimal force though) But I thought it could help reducing overshoots.

Everything I've seen so far replacing a stepper is not acceptable (effort accuracy or price)
So I'd like to know what you think of a dualmotor setup..

One end powered by a lowcost lowtorque stepper (about $10-15) with a gearbox to get acceptable torques (what slows that thing down again) and an extremely cheap standard motor on the other end just for long distance traveling
should be allinall less than $30 and if you already have those mountable digital measuring slides you can easily use the dataoutput to start and stop the motor and use the stepper just to adjust positions.
If you do not have those really expensive callipers you can catch and count the stepper turns.....

Oh man the walking theory is back again.

I'm sorry, I just stfu

'sid

if you want to play with motors, you should look into the servomotors that forrest hacked up. its basically a cheap, powerful DC motor with a magnetic encoder... it means you KNOW where the motor is, where its going, etc. plus it seems to be cheaper or easier than stepper motors.

theres also a great writeup that adrian did in the wiki:

[reprap.org]