basic questions

I've been looking into sourcing the materials for my own prussa mendel, going for maximum cheapness. All the prices for hardware are pretty reasonable, but i'm having problems convincing myself to shell over $200 minimum for even the most diy kits/bare components, which will not be self-replicating. It also kills me that i don't really understand what goes on on the boards, other than at best a vague 'i give it this signal and it magically makes thing go', 'black box' understanding. I mean, I can get a fully functional computer for all of $15, albeit crappy. That can run an os, crt monitor, periferals, and all that; why does it cos $200 to run 5 measly motors? can't we figure out a way to modify these inexpensive motherboards to suit out purposes? And the signals going to the motors are ultimately several wires worth of on and off, correct? That's a serial connection, why cant we just solder the wired to a USB or serial port cable and hook it up to a computer, then write some software to run the motors? At most, it seems like we could just take a basic arduino, which can do pwm, sensor input etc with one hand behind its back, and tell that to directly run the motors, with some code to tell it to turn on and off as stepper needs it to to turn. If the arduino can't handle the voltages needed, we can just consider the outputs as signals and give each wire a higher voltage transistor using is amplifying abilities to bet power from a power supply and run the motors. All of these would be way cheaper than the current setup, and are thus too good to be true or someone else would've thought of it before me. So, why do these alternatives not work? Why do i need such complicated and expensive electronics to control such simple functions?
tl;dr: looking for simpler, cheaper electronics; why wont these cheap and easy alternatives work?

Do you have a link to such a fully functional $15 computer? A new one, of course.

Other than that, quite some electronics development is going exactly the direction you describe. Less parts, less costs. Have a look at RAMPS or Gen7. If you still think there is way to much on these boards, I'd like to hear that.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I didn't mean a new computer, I meant craigslist.

I have looked at RAMPS and Gen7, and I didn't mean to say that there are alot of waste parts, only that I wish someone would explain to me what they were all for. Now that I think about it, that's probably far too vague a question to expect anybody to answer. Also i was really curious why those solutions I came up with wouldn't work, even though like I said before I know they won't or they would've been implemented already. Really I just crave to know the why and how.

Quote

And the signals going to the motors are ultimately several wires worth of on and off, correct? That's a serial connection, why cant we just solder the wired to a USB or serial port cable and hook it up to a computer, then write some software to run the motors?

No several wires is a parallel connection not a serial connection. You need a micro controller to receive the USB serial data and convert it into accurately timed step and direction signals. Then you need stepper driver chips to convert those into sinusoidal bipolar current waveforms to drive the coils. You also need a higher current drive for the heaters and some temperature sensors. That is what is on RAMPS and GEN7, nothing more, nothing less.

---

[www.hydraraptor.blogspot.com]

I'm curious as well as a cheap computer for many is actually free. Some people have thousands of resistors and caps around, others have tens of spare computers

I applaud all of the various electronics initiatives going on, just a bit surprised there is no computer based version. Clearly this is something out of the scope of my understanding.

Understanding how being able to self replicate electronics is a huge goal, recycling / re-purposing existing electronics is a worthy benefit.

You could use a PC with a parallel port but you would also need a real time operating system, or lose the OS altogether, because Windows and Linux do not allow accurate timing in the hundreds of micro seconds range without real time extensions. You would still need all the stepper motor drivers and power transistors. You would also need to be able to read analogue values for temperature control. Perhaps a sound card

But then all you would have done is replace the Atmel chip, which costs a few dollars, with a PC.

---

[www.hydraraptor.blogspot.com]

If you're thinking of something like the Mantis brute force driver (transistors switched by the parallel port), that doesn't work for a 3D printer because it needs all 12 available output pins to run 3 axes, leaving none for the extruder. (Being a unipolar driver, it also can't use the motors' full capability, so may need more expensive motors.)

Connecting a parallel port to step/direction controlled driver chips as in the current Mantis design would work, but wouldn't save much money over connecting the same drivers to a microcontroller.

nophead Wrote:
-------------------------------------------------------
> You could use a PC with a parallel port but you
> would also need a real time operating system, or
> lose the OS altogether, because Windows and Linux
> do not allow accurate timing in the hundreds of
> micro seconds range without real time extensions.
> You would still need all the stepper motor drivers
> and power transistors. You would also need to be
> able to read analogue values for temperature
> control. Perhaps a sound card
>
> But then all you would have done is replace the
> Atmel chip, which costs a few dollars, with a PC.

I figured that it was something like that - thank you for the info.

What you're describing is what "Mach 3" (windows, closed source) and "EMC2" (linux, GPL) do. And there are some people who use EMC2 to control the reprap. I've never heard of using Mach 3 though. It's in the wiki somewhere, and there are a couple of people doing it.

Like nophead noted, it just replaces the arduino, you still need the stepper drivers and so forth. And you actually still need the arduino for temperature control, so you don't save anything by going this route. EMC2 is a much more mature product as a g-code interpreter than any of the firmwares available, so it does have better path prediction and can get higher performance out of the same hardware. It's also (for some people) easier to use. Mostly for those who've already used it?

I'm actually quite impressed with mach 3. Anybody who can hack a functional real-time kernel into the heart of somebody else's operating system has got the title "real programmer" from me!

Not that I'd ever use it, I don't use microsoft products.

---

--
I'm building it with Baling Wire