Group Actions
Engagement
  • Author Author: mrvolt
  • Date Created: Date Created
  • Views 4018 views
  • Likes 8 likes
  • Comments 6 comments
Related
Recommended

The Difference Between SBC’s and Dev Boards

mrvolt

Introduction

 

  While it might be tempting to lump Single Board Computers (SBC’s) and Dev Boards together, as many retailers do for the sake of convenience, SBCs and Dev Boards have distinct roles and functionalities which separate them. Here’s how to tell the two apart and decide which board, or combination of the two (there are no rules!), will be right for your next project or mass-market, whizzbang IoT gadget:

 

 

What is an SBC?
image

 

Lower cost, lower profile, and ready-to-go, single board computers are the svelte system option when you need more processing oomph than a microcontroller, yet can’t cant afford the space or power budget that a full size motherboard or server requires. SBC’s contain everything you need to run a proper “full” operating systems such as Windows 10 IoT and many flavors (insert list here) of Linux, although it’s important to note that the main processor architecture of the SBC, be it ARM, x86/x64, Allwinner etc. will largely limit what kernels and drivers are available without extensive low-level programming and configuration. However, you will find the following key features common to many SBCs:

 

Power:

  • Consumption in the 10s of watts or less
  • Fanless or minimal airflow with compact heatsink operation
  • Low power draw allows cheaper and readily available PSUs

 

Connectivity:

  • HDMI, DVI, MIPI, LVDS, VGA, COMPOSITE display connections
  • WiFi and Bluetooth
  • 3.5mm line in/out
  • Full size USB 2.0/3.x common for I/O
  • Direct GPIO (General Purpose Input/Output) access
  • RJ-45 connections for PoE, Gigabit networking
  • CSI (Camera Serial Interface)

 

Downsides:

  • Processor and RAM typically fixed
  • GPIO access isn’t typically capable of real-time control
  • Limited graphical processing capabilities and extensibility
  • Analog to digital conversion usually requires external hardware

 

Applications and Usage:

  • Suitable as embedded systems for use in connected security systems, computer vision, point-of-sale, digital signage devices etc.
  • Mass availability for 1000’s of units for reliable supply chain and predictable EoL
  • Capable of running software that requires high-level programming languages and scripts

image

What is a Development Board?

 

  Contrary to their larger single board cousins, Dev Boards are focused to provide an accessible PCB breakout for an MCU (microcontroller unit) and typically have the minimal passives, oscillators, and hardware communication ICs necessary to test out the feature-set of said microcontroller. Some boards may have more advanced features such as hardware debugging, or extensive, built-in IO, like arrays of switches, LEDs, IMUs (Inertial Measurement Units), displays, and potentiometers to allow yet more rapid setup with no fuss.

  At the end of the day though, most Dev Boards aren’t the end product the manufacturer is trying to sell, barring educational/hobby boards like the Arduino platform. The Dev Board is the showpiece with all the bells and whistles to entice you away from competing products, so that you (the engineer) feel it’s the best option for your next widget, of which you intend to make 50,000 units, and thus incorporate their core microcontroller into your design.  This is why they’re called “Development” Boards.

 

However some manufacturers bridge the gap into mass manufacturing and release boards with castellated edges, effectively turning them into a SoM (system on module), which can save on valuable design and manufacturing time.

 

Here are some of the key features by which Dev Boards are set apart: 

 

  Power:

  • Low power consumption on the scale of milliwatts
  • Often able to run off a single cable for data transfer and power

 

  Connectivity:

  • USB (typically micro or mini variants at 2.0 spec)
  • Wifi
  • Bluetooth
  • Cellular
  • GPS
  • LoRa
  • IR

 

  Downsides:

  • Limited bulk availability, thus not suitable for incorporating into finished products
  • unpredictable EoL, making long term supply chain management/curriculum development difficult

 

  Applications and Usage:

  • Prototyping/proof-of-concept
  • Serve as reference designs
  • Interactive art installations
  • Low volume (<1000) goods

 

Conclusion

    The distinction between a Dev Board with a high-end ARM processor capable of running Linux and a proper single board computer is certainly fuzzy. There’s certainly a gradient between an mcu programmed in raw assembly up to an 86x system running a “full” OS, but the tiers along the way certainly matter, so I hope the comparison above has clarified what separates Dev Boards and SBCs in a way you find useful. If not, politely let me know in the comments below which areas you’d like to see me expand upon!

  • OMG! This topic is getting too close to call. I have seen development boards reach many capabilities like a Linux machine without the OS/ An example i Max32650 with LCD in a kit. Although only a M4 running at 120Mhz, it can process data without CPU involvement using SmartDMA... On the other hand   Zynq processor is a complete A9 with PL to build your own ARM/MicroBlaze that encroaches on the development boards. Expect the future to further blur the rules...

    Clem

  • in reply to mralpro

      wrote:

     

    I think is not totally accurate, the uC dev boards are not a little piece of junk as you describe, ...

    I re-read the original post. This is not what the poster wrote at all.

  • I think is not totally accurate, the uC dev boards are not a little piece of junk as you describe, it had a very specific function, that is a tool to develop a future hardware implementation. Indeed the earliers ARM boards were more like dev boards than SBC, because the real difference between two concepts is the audience that are leaded. uC dev boards are usually leads to hardware guys, and the SBC for software guys. The arrive of software engineers to the embedded world impulse to a new concept, boards ready to use for programming, testing, and deploy, by this way the developers don't need to thing about hardware, more than connect some headers or cables, because the real complexity is on software. Actually in the industry the uC are a very mature technology, because they don't need a little board that may to run a lot of things, they just need a simple thing that runs their software, each time is power up, with no hangs. To be more clear, if your a hobbyist, student, or just don't have restrictions about hardware or energy on your project, or your project is so complex that must to run a complete OS, then SBC is for you. If you're an electronic engineer that wants to learn how to do a customized system, dealing with restrictions as on real environments, a very low power application, or you're working on a subsystem part of a more complex system, then uC is for you. uC can to run very complex software, as ANN, fuzzy logic, digital signal processing, manage multiple I/O, and real time characteristics, if you looking more in deep you can find a uC DSP. On the other hands FPGA technology is a more sophisticated solution, with a totally different approach. I used these 4 technologies for differents applications, each one have their charm, but on my experience, many of the SBC solutions can fit perfectly on a uC, that is cheaper and more efficient, but a bit tricky to debug and deploy.

  • in reply to jomoenginer

    True. There's definitely a smooth gradient from a development board through to something that can and should be permanently embedded in a product. No hard and fast semantic rules really.

     

    I think most FPGA dev boards are just that. Then again, if you configure the logic to run as a processor, like a Z80, and if it has the necessary memory hardware, is it not then a computer on a single board? I think it's ultimately a philosophical argument where these products are in superposition you know? A board can both be a dev platform and and SBC until it is observed and running an application. Then you can clearly say based on the usage which way it should be called.

  • In my opinion development boards are boards with all the design files and source code open to the public. The idea behind the development board is to give a detailed development environment for the main chipset featured in the board.

    On the other hand SBCs has to be considered as a general purpose computing device with very limited design files and source code available. Raspberry Pi is an example.

     

    Earlier days only semiconductor manufactures/vendors only made development boards. ( Only to showcase and promote their products )

     

    Development boards mainly focus on one IC, It can be a microcontroller, microprocessor. a sensor or any analog or digital IC. But in an SBC there is no particular chip to focus.

    For example a sensor development board may have a microcontroller to collect data, process and send to a host machine but the focus will be mainly on the sensor and its modes of operation. Designer of the development board has to make sure that his design is capable of exposing all the possible features of that sensor to the user. But if an SBC is using a sensor it will be only for serving a purpose, which is mainly taking some reading at certain intervals.

     

    Regards,

    Aswin