I am toying with the Idea of an electronically re configurable driver board. It wouldn’t be cheap, my plans would be an reprogram-able, and Arduino compatible micro controller then various FET outputs, and A2D inputs (along with a bunch of digital I/O ports).
I want this to experiment with measurement systems for UFIDs, since at this point i am not sure what type of configuration would be useful i have not started picking things. Things that i would want to be considering when designing the circuit.
Digitally controlled op-amps (using variable pots)
Rotary encoders
Optical measurements (opacity/color)
Hot end driver
Heated Bed Driver
Thermistors
Thermocouples
Strain gages
steppers
DC motors,
Clip in sodium slot for the new RBPi
clearly i am not intending on having these all controlled at the same time, i just want to have a really flexible dev kit for mechanics. and i want to off load some of the real time control to the FPGA and higher-level control on the micro. Adding an FPGA most likely means that BGA soldering, so not at home.
I am not planning on making this board in the near future (i am swamped right not), but i wanted to gauge interest.
Might want to look at the Rambo board for some of your goals. At least as a starting point.
Look at the SmoothieBoard. It has more CPU power and is more open.
@Wylie_Hilliard care to qualify the quote “more open”. Both are open source, seems both supply schematics and all. So at that point it would seem they both hit the end of the open spectrum.
Not open as in open source, more i/o pins are available for use and the code is modular so that it is easier to add extra features.
Edit: It looks like Rambo also makes all the pins available.
Raspberry pi and one of the breakout development boards might cover it.
Ugghh on the Pi. Look at Beagle Bone Black, Radxa Rock, or any other of a host of more powerful boards that are fully open.
“UFID = University of Florida Identification” ?
@Steven_Critchfield funny that you forget that there’s several arm processor boards that can be configured for this.
So what about the Ardrino add on development boards.
And believe me its easier to program when all the source is there.
Some of those boards you mention still havnt got all the sources sorted.
Pi doesn’t have all it’s source sorted either, nor will it. Remember it runs on a Broadcom chip. The BBB has it’s source sorted. The Rock, well same as the Pi, but has a lot more grunt. ram, and storage.
Ah, sodium = SODIMM, that had me puzzled! I like the idea of a CPU board that plugs into to a base board, seems more flexible and potentially cheaper.
Is this for steppers or DC? I’ve been playing with similar ideas. FETs aren’t the best idea, though. I’ve been thinking of going with an L298N dual H-bridge chip because it has built-in current sensing so that you can still do current control (which you do not want to be driving steppers without). I’m hoping that it will even make it possible to implement stall detection. I’ve been looking at probably using the Attiny2313 as a cost-effective microcontroller with enough pins to do what I want.
@Steven_Critchfield your an idiot. The pi is there to monitor and program. The breakout boards are there to drive the requirements…look at the breakout boards even the beagle can be used as a breakout board.
@Whosa_whatsis My thoughs are a FET Array ran to the the FPGA, For generic usage (IE PWM power, maybe implement PID on the FPGA to offload processing time, or driving really beefy steppers) , I would need to put a couple of H-Bridges to make the DC motors worth while, Without that you are just pushing them in one direction. I am thinking that if you have an FPGA the number of pins needed on the micro is minimal, but the cost would go up in assembly and in BOM.
@bob_cousins Sorry, wrote that without spell checking, was thinking about it more on the way to work. UFID, is Universal Filament(or feed stock) Identification, it is a standard we are working on to get key print parameters to the end user in a compact and cost effective manor. This project i have in my mind is to develop a measurement system that quickly and automatically generates the data for a filament to help out filament suppliers.
@Steven_Critchfield To be honest, i like opensource project, they are nice and fun. but i am not going to argue with the cheapest possible solution for an arm on board, i also like the clip in module, not having to deal with a stack mount or cable is appealing to me.
Note, if i do build this it will have every pin on the FPGA connected to something (Direct to the RBPi, direct to the arduino, and direct to every and every remaining pin will go to a break out header just in case). All of the wiring would be handled by the FPGA. I would want everything that can be broken out broken out. I don’t envision this being cheap, or small. But super flexible and reconfigure-able. I do plan on basing it off of a existing open source hardware package, but more so i am curious if there are people in the community that would be interested in it.
I’m also in the ARM camp, I won’t be buying another 8-bit chip. There’s many good cheap ARM chips with much more Pferdestärke than their 8-bit counterparts.
Mbed for your IDE with code examples. USB mass storage programming.
If you like DIP, the LPC1114FN28 are less than 2 quid individually. Otherwise I’m a fan of freescale (K64F has an FPU) or energy micro (not mbed, but fantastic datasheets and hobbiest support).