Author Topic: Development Board Recommendation Please  (Read 7814 times)

Offline sparky961

  • Hero Member
  • *****
  • Posts: 844
  • Country: ca
Development Board Recommendation Please
« on: June 07, 2019, 08:15:05 PM »
It's been a while since I've popped by, but this seemed like an apt question for this forum...

I'm wondering if anyone can give me some guidance on purchasing a development board for my next project.  I've ruled out anything Arduino/microcontroller based because I'd like to have a nifty UI with nice graphics and a bunch of other higher order stuff.  From my little bit of research, that leaves the microprocessor based Raspberry Pi and BeagleBone at the top of my list.  I've been favouring Beaglebone because of the 7 analog inputs that I would likely use up, but if it's easy to graft on an analog input board I'd be open to anything.

There will be lots of analog data acquisition in this project, but I haven't worked out how many inputs - plus things will evolve.  I don't think 20 is out of the question.  I'll also be doing some PWN output and simple digital on/off control.  12V power, or the availability of DC-DC power supplies is an asset, as this is an automotive type application.

I may be able to narrow it down a little if someone wants to ask some questions, but at this point I'm ok with getting something way overkill as long as it has LOTS of I/O.

Thanks!

Offline efrench

  • Full Member
  • ***
  • Posts: 213
Re: Development Board Recommendation Please
« Reply #1 on: June 08, 2019, 04:03:09 AM »
Well, I'm using a Teensy 3.5 along with a Nextion touchscreen for my rose engine project.  Plenty of UI and IO capabilities.

Offline Muzzerboy

  • Full Member
  • ***
  • Posts: 124
  • Country: gb
    • The Mighty Shiz
Re: Development Board Recommendation Please
« Reply #2 on: June 08, 2019, 09:08:31 AM »
There are literally hundreds of SBCs offering compatibility with the likes of the Arduino and RPi, so if you are looking for something a little off beat such as loads of ADCs, it may be worth having a look around. Mickmake provides a useful overview of the bewildering choice out there. https://www.youtube.com/channel/UC7GMT3ohvYEAJFDenzj9EMQ

From what I understood, the RPi was based on the sort of SoC used in smart phones, so has more video processing hardware on board, compared to the Arduino family. Having said that, you can branch out from either starting point into any number of different directions.

The Beaglebone Green is used as the basis for the machine controller used in the Centroid Acorn. It uses a 1GHz ARM Cortex A8, like the other Beaglebone SBCs. However, Centroid created a PC-based GUI connected to the Beaglebone BOB by Ethernet. That's another approach you might take.

Offline russ57

  • Sr. Member
  • ****
  • Posts: 278
Re: Development Board Recommendation Please
« Reply #3 on: June 10, 2019, 12:21:47 AM »
A gent I was talking to Saturday had an interesting view.

He is using a cheap ($50) smart phone*, with q python, as a control (=raspberry pi with screen) and (multiple as needed) Esp32 boards to provide io.

There is some sort of 'helper library' which facilitates comms between the 2. (his preference was udp for simplicity).

There are a few things I'm not sure about with this approach, but for some projects it seems a good idea.



Russ
* no sim card, just wifi. Thinking about this, a cheap tablet might be even better...

Offline sparky961

  • Hero Member
  • *****
  • Posts: 844
  • Country: ca
Re: Development Board Recommendation Please
« Reply #4 on: June 10, 2019, 08:33:25 AM »
Thanks for the tips so far.  I was really hoping to keep everything to one physical unit, but I can see the potential benefit to distributing it among smaller cheap boards.  The variety and selection are just so plentiful it's hard to make anything but a random choice!

If anyone has specific links to examples/tutorials/videos showing an interconnected systems that would help narrow down my options and get me going quickly it would be appreciated.

Offline russ57

  • Sr. Member
  • ****
  • Posts: 278
Re: Development Board Recommendation Please
« Reply #5 on: June 10, 2019, 09:03:48 AM »
What is the project, roughly?

There might be something you want to achieve that would make one a better choice.

Check out clough42 for his build of an electronic leadscrew . Using the ti launchpad because it has a hw floating point processor.

Andreas Speirs has some good stuff.

As does mick make.


The Esp32 offers 6or so capacitive sensors.


I have a beaglebone running my solar battery.
(not my work, it's a Victron product.)

Raspberry pi has great community support. But with a screen gets to be not so cheap.
The arduino is great as an edge device but anything past serial comms gets harder (or $).
The esp8266 has wifi built in.

There are also a number of others- orange pi,

And a huge range of options for screens.


Russ


Offline sparky961

  • Hero Member
  • *****
  • Posts: 844
  • Country: ca
Re: Development Board Recommendation Please
« Reply #6 on: June 10, 2019, 11:38:13 AM »
The project will evolve, which makes it difficult to define.  But I understand the importance of trying to define it before beginning, so here goes:

The entire system will reside in a space about 15' x 8' x 7', with plenty of nominal 12V (7V - 15V) power available.  Sensors, switches, and control will be distributed throughout this space.  The control components will reside in a ventilated, heated, but not cooled environment.  I expect temperatures between 5°C and 60°C are possible.

Phase I
- Completely customizable touch screen interface maybe 6-14", stupidly high definition not required
- Main controller attached to touch screen, capable of talking to everything else in the system
- 3-6 current sensors (0-75 mV shunt, best possible accuracy
- 4-6 point analog voltage monitoring (0.005 V accuracy)
- 10 analog temperature monitoring (0.5 °C accuracy) (could also be digital single wire if cheap sensors exist)
- 10 digital inputs
- 4-8 PWM LED outputs
- 2-4 PWM outputs to control variable speed fans
- 10+ miscellaneous digital outputs

Phase II
- Add second touch interface, capable of controlling anything the main interface can
- GPS logging
- GPS navigation (OSMAND) on second touch interface
- Read CAN bus interface
- Add 4 humidity sensor inputs
- 4-6 PWM servo outputs

Phase III
- 4-6 HD video and audio recording 10 hour loop (if possible... haven't done the math for storage requirements)
- Interface with speakers (of some sort) for audio playback, don't need to be stupidly high quality
- Write CAN bus interface
- Remote server I/O through cellular data interface


So, as you can see if I get all the way to Phase III I'm beyond the scope of a simple microcontroller development board...

Offline russ57

  • Sr. Member
  • ****
  • Posts: 278
Re: Development Board Recommendation Please
« Reply #7 on: June 10, 2019, 11:32:44 PM »
Ok then... Not your typical starter project..

I'd be thinking for that scale, you need a central 'controller', fairly capable, but I'd be looking to divide and conquer, by separating and offloading discrete functions either by say 'all the current sensors' on a smaller system, or split in some other way that makes sense - maybe a subsystem has its own current, temp, humidity sensors, and feeds into the master system.

Otherwise, you are looking for up to 80 odd io lines, which can be done of course by using things like i2c io expanders but I'd be concerned about scalability. (and stability).

And I assume many of the sensors are external to the main system so you will need to remote them anyway..

Sounds like a great project. Good luck with it.


Russ


Offline sparky961

  • Hero Member
  • *****
  • Posts: 844
  • Country: ca
Re: Development Board Recommendation Please
« Reply #8 on: June 11, 2019, 09:21:41 AM »
Well, I don't know if there's a more logical way to make the decision, but it seems that since I want to be communicating with CAN in later stages that it makes sense to make this the backbone of my smaller subnet.  Having a separate controller just to get a bunch of inputs seems overkill, but maybe I'm just dating myself with that paradigm.  The biggest advantage I can see is subdivision by location so that I don't have to run as much wiring and not as far.  Only the communication needs to come back to the main control, and if I'm not mistaken that's only two wires with CAN.

It certainly isn't a "blink the LED" project.  But of course, that's the first thing I'll do when I finally do pick the boards I'll use.

Having the CAN requirement may help narrow things down to a bunch of inexpensive microcontroller boards ("Arduino" style) that support CAN, plus one or two more powerful SBC (a la Tarte Framboise).

Offline sparky961

  • Hero Member
  • *****
  • Posts: 844
  • Country: ca
Re: Development Board Recommendation Please
« Reply #9 on: June 13, 2019, 05:23:15 PM »
I have a selection of boards on their way. Main controll is likely to be a Raspberry Pi , and a bunch of slave Arduino Pro Mini clones. I'm not worrying about CAN at this point. I'll just use a simple RS-232 link between boards.