A Router-Based Dev Board That Isn’t A Router

Here’s somethirouterng that be of interest to anyone looking to hack up a router for their own connected project or IoT implementation: hardware based on a fairly standard router, loaded up with OpenWRT, with a ton of I/O to connect to anything.

It’s called the DPT Board, and it’s basically an hugely improved version of the off-the-shelf routers you can pick up through the usual channels. On board are 20 GPIOs, USB host, 16MB Flash, 64MB RAM, two Ethernet ports, on-board 802.11n and a USB host port. This small system on board is pre-installed with OpenWRT, making it relatively easy to connect this small router-like device to LED strips, sensors, or whatever other project you have in mind.

The board was designed by [Daan Pape], and he’s also working on something he calls breakoutserver There’s a uHTTP server written specifically for the board that allows any Internet connected device to control everything on the board. There’s also an HTML5 app they’re developing which could be pretty interesting.

All in all, it’s a pretty cool little device that fits nicely in between the relatively simplistic ‘Arduino with an Ethernet shield’ and a Raspi or BeagleBone.

Comments

  1. vb says:

    Jist look at http://8devices.com/carambola-2 – same SoC, same functionality, cost only 33Euro

    • daanpape says:

      Hello,

      Yes indeed, it is the same SoC and functionality. But we are much cheaper at only $20 for the module and there is a built in antenna. Just like the Carambola2 we have CE/FCC certification.

      For about the same price as the Carambola2 you have the complete DPT-board and it’s custom software.

      Kind regards,
      Daan

  2. RooTer says:

    Equivalent can by already bought on http://8devices.com/carambola-2 . But it is more pricey. If you are looking for something dirty cheap (like ~13$ for module with external antenna) go http://v-solution.en.alibaba.com/ (25$ is for breakout board). They prefer to handle orders of >200 but I got some samples from them no problem. Quality is not the best and I would guess that “ar9331″ is not calibrated (worse wireless performance) – but the price;)! . The biggest drawback? no FCC/CE I know about. After that you have to recompile your own openwrt (provide good mtdlayout – not that hard).

    If you prefer ebay http://www.ebay.com.hk/usr/alfie.zhao sells some routers (less gpio on pads but integrated antenna and rj45). Around 16-20$ depending if you want enclosure or not. Again not sure about FCC/CE.

    • daanpape says:

      Hello,

      Our module (DPT-Module) also has CE/FCC approvement and is much cheaper than the carambola2. Unlike many of the other solutions we provide European customer support and specialized software.

      We are currently working on setting up a community forum so anybody can upload software for the DPT-board and ask questions about hardware and software.

      Kind regards,
      Daan, DPTechnics

  3. Bogdan says:

    I like routers a lot for hacking, I am really wondering how could we convince the manufacturers to break out some of the unused pins on a header…

    I am not sure what to think of this board… they are basically making a breakout board for that module on the HW side, there are plenty of these out there. If only they would have added a simple ADC on all that empty space on the board…

    • tekkieneet says:

      That’s a problem. You might want ADC, Joe user might want something else. May be a proto area in the free space with a matrix of 0.1″ grid pads so that the users can at least solder what they want on there. That layout could be squeeze tighter by about 1/3.

      Specs-wise my old $35 router with exception of FLASH (8MB) and lack of internal GPIO has better specs. Size of FLASH is no big deal as long as linux get sufficient space to load in USB FLASH drivers.

      • Bogdan says:

        yes, a proto space might be a better idea. I was suggesting ADC because that is the only thing missing from a typical micro board.

        Looking at the specs I see that there is definitively something else they should have added, a level translator: “5 normal 2.5 volt IO ports with software I2C, SPI, … capability”

        I would still prefer this one instead of their board: http://8devices.com/carambola-2

      • daanpape says:

        Hello,
        I am Daan Pape, the DPT Board developer. We are very open and listening to all the tips and wishes of our users.

        We offer the DPT Module at $20, it has about the same specs as the Carambola 2. The DPT Board you see on the picture is not the latest revision. I really like the tip for a proto space, and I will make this available.

        As for the level shifting we opted for 7 high power outputs (NPN) and 3 high voltage inputs, you can use 3V3, 5V or even 10V input on them. I will look at the possibilities to make the 2.5V IO shift to 5V.

        Thank you very much for all the tips.
        Kind regards,
        Daan

      • tekkieneet says:

        I am a big supporter for having some logic level translator (such as those 74CBxx parts) on GPIO and/or series termination instead of just breaking raw signals to headers. The level translation device would at least extend the GPIO input voltage range.

        Since users are unlikely to use coax cables to GPIO and such, all of the connections should be consider loose wires and without impedance control. A fast signal even at the on a long piece of wire even at the right I/O supply voltage can cause ringing exceeding
        the input limits. Series termination of about 100 ohms range would help a bit on damping and offers some protections.

    • Some of them actually break out GPIO, GL.iNET (more or less a wr703n clone) does this. I’ve written a few words about it here: http://blog.eikeland.se/2014/07/22/gl-inet-openwrt/ . The manufactorer offers instructions on how to build a custom openwrt for it here: http://www.gl-inet.com/w/?p=398&lang=en .. Really nice for small projects needing a few GPIO, USB, and/or Wifi/LAN.

      The router is available on eBay and Amazon. Costs about $25 + shipping.

    • RooTer says:

      arduino Yun seems like something you want – the same chip (ar9331) with atmega for communication with the physical world (including adc). In fact I bet you could put together something like that your self at half the price (cheap ar9331+arduino-compatible board). Yun is supposed to be opensource.

  4. Sven says:

    It’s true that most modern ATMs run Windows, every model i have seen with a modern full color screen has been running windows, and i know this because they frequently crash and hang at the windows background, or pop up a windows error notice that steals focus from the ATM software, thus making input impossible since none of the accessible buttons are space, enter or escape…

  5. GotNoTime says:

    The router module that they’re using is the Gainstrong Oolite aka EasyLink EL-M150. You can buy it from the usual places in China. It is actually just the Atheros AP-121 reference design so there shouldn’t be any surprises when upgrading OpenWRT. One thing is that the boards aren’t calibrated at the factory so the WiFi performance may be bad. The three EL-M150s I’ve got have exactly the same ART calibration data so I expect that they’re just using the same for all.

    • daanpape says:

      Hello,

      Yes we have worked with Gainstrong and have tested the modules before the current CE/FCC version. This way we can assure the design is solid and that we can deliver what we say on our campaign. This has taken us several months and many revisions.

      The DPT Board is designed around the module and provides all the extra functions such as the high power IO, switching power supply, USB with protectionisme, dual ethernet port,…

      We really focus on supplying great software tuned for the hardware. This way you van make IoT project without much hassle.

      Kind regards,
      Daan

  6. StinkySteve says:

    “and it’s basically an hugely improved version ” what??

    • daanpape says:

      Hello,

      It is hugely improved on IO. You have SPI, UART, 7 high power outputs and protected inputs. On top of that we have a high performance server to control the board.

      Kind regards,
      Daan

  7. DeKay says:

    Anyone know of an open hardware version of the IPAC?

  8. omegatotal says:

    and dont forget about the USB host port.

    ;-)

  9. Emeryth says:

    Searching on taobao for ar9331 gives a couple of already existing similar boards, like this one:

    http://item.taobao.com/item.htm?spm=a230r.1.14.40.b72MtR&id=36447327606&ns=1#detail

    • daanpape says:

      Hello,

      We are aware of some similar boards. But none of them provide similar software as on our board. Also we provide service and try to answer all the questions we receive.

      On top of that, our module has CE/FCC approvement.

      Kind regards,
      Daan Pape

      • Squonk42 says:

        And what about the board?

        • daanpape says:

          Hello,

          We have plans to CE/FCC certify the board. As it is a development board the first batches won’t be. This because we are still listening to the public for wishes and making small changes to the design.

          If we would reach a point were volumes become very big we will look for certification if needed/wished by our clients.

          Kind regards,
          Daan

  10. Erik Johnson says:

    Not sure I understand the “specially modified uhttpd for GPIO control” I’ve been using stock uhttpd to communicate with my driver program thus GPIO for years now. uhttpd supports CGI (executing binaries/scripts) natively, and the CGI/script can toggle the GPIO directory or tell a background task to via pipe file or other various means…

    • daanpape says:

      Hello,

      We have started with the µHTTP code but removed all the CGI scripting. We have implemented a REST interface in a single process. This way we reach maximum performance, better than with any scripting language. It is even faster than Lua and our software will get the same, if not more functions than LuCi.

      Kind regards,
      Daan Pape

      • Bogdan says:

        What was normally not high performance enough? You cannot expect to bit bang a pwm in an LED through the internet…. so what was the limit?
        In my experience with such a system most of the delay is caused by the communication itelsf, eg: it takes more time to go: click button on web interface – send packet – router redirects it – your board than it takes for the board to execute a script turning on the LED. The communication time becomes even more significant when you do this across the internet.

        So, what was wrong with CGI ?

        • daanpape says:

          Hello,

          We had a slow experience when using multiple clients and we noticed high CPU usage from LuCI from time to time. Using less CPU to do things means lower power consumption.

          So when you occasionally switch on and of a LED you wont notice the difference. But when using multiple clients or much traffic (like high speed temp logging) you really notice the difference.

          We have chosen the C implementation because we want to get the most out of the board with the least of resources so the user has plenty of resources left for his own software.

          Kind regards,
          Daan

          • Bogdan says:

            “Using less CPU to do things means lower power consumption.” Are you sure? I did notice any difference in power consumption between idle and 100% CPU usage.
            The only thing that seems to make a difference is how much you are transmitting via wifi.

            The problems i noticed with LuCi was that by default in the openwrt implementation the connection number is low (maybe 3?) and this causes the interface to become non responsive quite easily. In my tests I have increased this number a lot (maybe 30 or so). I tried 10 updates/s per page(just data) and opened the page on 3 devices a total number of 10 times and got CPU usage under 15%. You can check the last post on my website to understand what kind of page I was talking about.

  11. surfingtheether says:

    Ordered one, this looks like a great project, i have used openwrt and dd-wrt for a long time in a lot of thing i have built.

  12. richms says:

    Is there a dual band wifi version of this possible? 2.4GHz is bordering on useless for me in the evenings now.

    • daanpape says:

      Hello,

      I’m afraid we only support 2.4GHz, but you can select any channel you want. We will make a solder pad available for an external antenna.

      Kind regards,
      Daan Pape

  13. Frank Marien says:

    Daan, can you give us an idea of the power consumption (min, avg, max), and: can the Wifi be entirely disabled (e.g. not radiating ANY energy)? Also: are the Ethernet port 100 of 1000 Mbps?

    • daanpape says:

      Hello,

      The power consumption is 0.36Watt average and 0,8~1Watt maximum. We are working on software optimizations to get a minimum of 0.2Watt.

      The internal switch is 1000Mbps, but the ethernetports have a bandwidth of 100Mbps.

      Kind regards,
      Daan

  14. KS says:

    This sounds very exciting. I am getting more and more into OpenWRT and a low cost platform to experiment on is right up my alley. Also, I like this guy’s attitude. He answered all these questions without getting flustered.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 96,459 other followers