|
Re: why no one use 2d printer technology for 3d printing? February 12, 2014 12:32PM |
Some reverse engineering will be needed, just to power the encoder and pick out the signal. My thinking is that making a custom driver circuit is easier and/or more useful than trying to reverse engineer the whole controller...
Give me a day or two and I will put up what I have on github. If people think there is any value in it, we can use it.
I think this approach only makes sense (outside of academic curiosity) if the controller can be made very cheaply:
Cost of the standard option = Arduino + Ramps (or whatever) + 4X Pololu stepper drivers (4X $8) + 5X stepper motors (5X $9)
Cost of the "salvaged dc motor + encoder" option = Arduino + Ramps (or whatever) + 4X Controller boards (4X $P) + 4X dc motors + encoders (free!)
So, the break-even point is when the cost of the controller board $P = (4X $8 + 5X $9)/4 = $19.25
These numbers are approximate of course!
VDX: Do you have a part number or link to the motor controller you mention above? I am curious about what is already available off-the-shelf.
Give me a day or two and I will put up what I have on github. If people think there is any value in it, we can use it.
I think this approach only makes sense (outside of academic curiosity) if the controller can be made very cheaply:
Cost of the standard option = Arduino + Ramps (or whatever) + 4X Pololu stepper drivers (4X $8) + 5X stepper motors (5X $9)
Cost of the "salvaged dc motor + encoder" option = Arduino + Ramps (or whatever) + 4X Controller boards (4X $P) + 4X dc motors + encoders (free!)
So, the break-even point is when the cost of the controller board $P = (4X $8 + 5X $9)/4 = $19.25
These numbers are approximate of course!
VDX: Do you have a part number or link to the motor controller you mention above? I am curious about what is already available off-the-shelf.
|
Re: why no one use 2d printer technology for 3d printing? February 14, 2014 11:08AM |
Totally agree
|
Re: why no one use 2d printer technology for 3d printing? February 16, 2014 04:45PM |
VDX, thanks for the links. That looks like a really nice system!
Here is the little controller I made (in collaboration with some others): [github.com]
It uses a LMD18200 and a PIC. Unfortunately the LMD18200 alone costs more than the entire budget for a "cheap" board. The PIC is only about $3 however, so that can probably stay...
I am thinking about trimming down the design and using something like a L293D for the motor driver. Remember we're trying for a small, cheap board to run the cheap DC motors dug out of a cheap printer (are you detecting the "cheap" theme yet
) so I think the L293D might be OK.
Does anyone know of any other motor drivers that might work?
Here is the little controller I made (in collaboration with some others): [github.com]
It uses a LMD18200 and a PIC. Unfortunately the LMD18200 alone costs more than the entire budget for a "cheap" board. The PIC is only about $3 however, so that can probably stay...
I am thinking about trimming down the design and using something like a L293D for the motor driver. Remember we're trying for a small, cheap board to run the cheap DC motors dug out of a cheap printer (are you detecting the "cheap" theme yet
) so I think the L293D might be OK. Does anyone know of any other motor drivers that might work?
|
Re: why no one use 2d printer technology for 3d printing? February 17, 2014 03:05AM |
In my experience, it is a bit of work to get a good solid MOSFET H-bridge design, especially if you want to drive it at a few kHz with a PWM signal. You have to worry about shoot-through current and snubber circuits and gate drivers and so on... and by the time all that is factored in, an "all in one" package starts looking better and better. However, if anyone would like to share some schematics I'd like to see what people are using... Maybe it doesn't have to be that hard.
It looks like the UHU controller is using a IR2184 gate driver. They are also using kind of a complicated network on the gate of the MOSFET that I must confess I do not understand completely... I think this is to introduce a delay when the MOSFET turns on, but allow it to turn off quickly - in order to prevent shoot-through during transitions.
A biploar H-bridge might be more forgiving than a MOSFET one, but they can be tricky to bias properly, especially if you want to use a wide range of supply voltages. Maybe some experts in this area could chime in and offer some advice?
There's a nice writeup on a biploar H-bridge design here. Something similar to that circuit might work...
It looks like the UHU controller is using a IR2184 gate driver. They are also using kind of a complicated network on the gate of the MOSFET that I must confess I do not understand completely... I think this is to introduce a delay when the MOSFET turns on, but allow it to turn off quickly - in order to prevent shoot-through during transitions.
A biploar H-bridge might be more forgiving than a MOSFET one, but they can be tricky to bias properly, especially if you want to use a wide range of supply voltages. Maybe some experts in this area could chime in and offer some advice?
There's a nice writeup on a biploar H-bridge design here. Something similar to that circuit might work...
|
Re: why no one use 2d printer technology for 3d printing? February 17, 2014 07:32AM |
|
Re: why no one use 2d printer technology for 3d printing? August 22, 2014 06:24PM |
It is possible though! I played around with a linear unit from a hp printer and made it act as a stepper motor using feedback from the optical encoder and the PID_v1 library. In my main loop i set
I'm trying to build a repstrap using stepper-motors on all axis but the X-axis where I'll use this linear unit (just to see if it can be done ). Next step is to merge it with the Teacup firmware.
I'm trying to build a repstrap using stepper-motors on all axis but the X-axis where I'll use this linear unit (just to see if it can be done
). Next step is to merge it with the Teacup firmware.
Sorry, only registered users may post in this forum.