Posted on Sunday, March 10th, 2013 in Arduino, AVR, open source, wireless by the machinegeek

Emanuele has informed us of a new open source project known as OpenduinoGSM. This is a standalone wireless remote control board which communicates via GSM.

With OpenduinoGSM you have the option of using the Arduino development environment, so anyone, even those who are unfamiliar electronics and circuits, can try to realize the automation they want by writing simple code in C ++ or by using one of the many open source projects published and modifying it according to their needs.

For advanced users, the OpenduinoGSM also has a JTAG interface for use with AVR Studio.

Their first project is a GSM gate opener to control an automated gate. When the device receives a call from one of the numbers in the SIM phonebook, the OpenduinoGSM relay closes the contact for 3 seconds, activating the button that controls the opening of the gate. This project saves money by avoiding the need for a separate remote control other than a cell phone for the gate (or other connected device of your choice), and even the call is free, since after receiving the call the device only reads the caller ID and hangs up without answering thus avoiding consuming phone units!

You can find the schematic and code download links at

Via the contact form.

This entry was posted on Sunday, March 10th, 2013 at 4:44 am and is filed under Arduino, AVR, open source, wireless. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.

11 Responses to “OpenduinoGSM”

  1. erdabyz says:

    The board layout is pretty bad… fails in too many levels. Power distribution to the SIM900 looks like it’s badly done on pourpose and the RF trace design is just wrong. Their design is neither behaving like a microstrip nor like a coplanary waveguide, so the trace thickness is probably off and there are also discontinuities in the ground plane. The impedance of that trace could only be close to 50ohm by pure chance, and you don’t want reflections in a GSM system where the peak output power can reach 2W. And there are a lot more fails too.

    I would love if that thing catastrophically failed or didn’t work at all so the designers would be forced to learn how to do things properly in RF designs. But as always happens with these things it just won’t fail and it’ll keep working reliabily for long, and several people will copy the design. All by pure chance.

  2. Emanuele says:

    Hi erdabyz,
    the power supply of SIM900 it’s good. We have designed and tested it for the purpose. The antenna works fine also, and the power of signal is strenght compared to some cellphones.

    The theory of RF signals is wonderful, but is only theory and in the real world there are more factor to be considered, like dimensions, cost and mass production optimization etc.

    Today we have more than 10 OpenduinoGSM installated, in continuosly function…..

    No chance, not IF not MAYBE….. simply it’s works fine ;)

    • erdabyz says:

      The power supply might be good, but the power distribution isn’t. The SIM900 uses a quite high power RF amplifier, and power distribution for it is quite tricky. It draws about 100mA IIRC while idle but when it starts transmitting it can draw up to 2A peaks at 244Hz, switching rapidly from taking 100mA to taking 244mA. A good transient reponse of the supply system is crucial for optimal performance. You have added a couple of large reservoir caps near the PA supply, and that’s good, but then the ground trace branches (for what I can see in the picture) to other ground pins, and it does so with T branchings which are not good because they promote effects such as common ground impedance coupling, which in a system like this with high currents at relatively high frequency might be catastrophic. I’m actually surprised that it does in fact work reliabily with that design. For designs like that a proper ground plane with low impedance connections (i.e. multiple small vias per pin) is probably a must. You also have a ground loop surrounding the big decoupling capacitors.

      The RF trace isn’t well designed because it’s not even a standard transmission line. Your layout doesn’t behave like a microstrip (the main trace is surrounded by ground) but isn’t behaving as a coplanary waveguide either (ground surroundings are too thin and they aren’t stitched to the ground plane). So any trace dimensions calculation done basing on standard models will just don’t work. Your trace will have an uncontrolled impedance and reflections (probably large reflections) will occur. That will make the power amplifier work in suboptimal conditions stressing it. The trace also has unneccesary bendings and discontinuities and the ground plane also has discontinuities that could have been easily avoided. I agree that RF theory is RF theory and that in most designs the final tuning is done by empirical measurements because there are too many variables to take into account for simulations and calculations in real environments, but RF theory is there for something and it does work, and there’s a way to do things based on that teory and a way not to do them. That design follows the later. Also for what I see in your design, improving the RF part should be very easy and wouldn’t add any cost, size or anything.
      Maybe you’ve been lucky with this design but it’s not a good design and I think you should be aware of it. It’s the only way to improve and to stop doing things that work by pure chance.

      • Agree with all the points. The unconnected GND traces around the RF line at the sim900 side are very bad. Your GND traces are NOT at a zero volt potential, because the wavelength at 900 MHz is quite short. So you just have something like a hairpin filter /stripline coupler with unknown properties!

        A design that “happens to work” is not a good design. Specially about RF. Do not mess with high frequencies. Your USB routing is not good eirher, with multiple unnecessary layer changes and a wide space between D+/D- traces.

        I had problems with a board that had USB traces routed as hell, and I had problems.

        Months after the board went into production.

        So we had to use kludges (rework, additional ferrites on cable, etc). I had to say that the offending USb traces were not long either, a few centimeters at most. But I had RF in the same design. Bam!

      • (sent too fast : I wanted to say:

        I had to work on a board that had USB traces routed as hell, and we had problems)

  3. Emanuele says:

    The analysis you have done is quite correct, but you have not taken care about two factors:

    1) The board is a prototype. We did a lot of testing. Surely there is a lot of improvement to do and that is why we are in crowdfunding. However, once again, the board works fine and is not a coincidence that 10/10 different installations running well.

    2) Our intention to start from hobbists quality and try to improve it, as long as you keep the costs low. Time to market (time for prototype) was crucial for us.

    In conclusion, thanks for your suggestions, but it is a prototype, it works in many places and we have realized it in a very short time.

  4. fuzzzy says:

    I agree with comments above and I’m expecting improvements on this device in future releases. As i can see it’s not worse than Arduino GSM SIM900 Shield you can find on market. It can be a good chance to start with GSM and its applications, cheap and easy to use. Don’t forget it’s completely OPEN SOURCE and everyone can contribute to develope and improve it. I hope developer will take in mind everything suggested in this page in order to make a good device. OPEN SOURCE – OPEN MIND

    • I would say far better !

      The split RF module with very short RF traces is a clear improvement over your previous design.

      But I did not have a look at power distribution :) so erdabyz will have to post another comment :)

      Good job!

  5. alenoise says:

    What do you think about this one (Dromo-1)?

Leave a Reply

Notify me of followup comments via e-mail. You can also subscribe without commenting.

Recent Comments

  • KH: Really? They are digital inputs driven by an MCU. The MCU GPIO probably has diode clamps and ESD protection. Have you experienced such failures? Please...
  • KH: Neither. Google "Marco Schweighauser’s 8-bit virtual Javascript CPU" and see the first result. Sounds more like a very limited feature toy CPU.
  • Alan: "8-bit x86 like CPU" - isn't that an 8080, or Z80?
  • Geoff: Is it still Sunday somewhere? If so, yes please.
  • hli: Sunday++