Seeed KiwiSDR metal enclosure finally available!


  • Looks fine, are they EMC tested?

    Thinking about shielding the VHF (145MHz) spurs




  • jksjks
    edited December 2017
    I did some preliminary tests which showed a big improvement using the enclosure. But I don't have a "typical" 2m rx setup here to observe the real-world effect (i.e. with an actual 2m antenna and feed line). All I can do is relative E-field measurements with the spectrum analyzer. A lot of the problem was observed to be RF conducted on the DC power cable and re-radiated, despite the filtering on the Kiwi board.

    I still need to try things like putting ferrite beads on the SPI clock and data lines. But I believe the problem is more complicated than that. I think it has to do with the fact the ground return path between the Kiwi and Beagle travels over just a few pins of the header connector between the boards. This is an unfortunate limitation of the design where the Kiwi is an add-on board to the Beagle.

    The two boards themselves have good multiple ground planes. Once you get past the Ethernet noise problem (not easy it seems) the Kiwi received spectrum is very quiet. There is little self-interference from the Kiwi or Beagle. There are several examples on of excellent RFI/EMI-free reception. So except for the SPI spur problem the Kiwi works as planned. Remember that you can move the spurs away from 2m by changing the SPI clock frequency setting on the admin page.

  • I should have taken a before and after, but the 8074 KiwiSDR using a 210' top dipole lost most of the noise after installing the enclosure. The 8073 KiwiSDR using the W6LVP loop did not show as much noise reduction, although it did drop most noise by about 6dB, as I observed it. I do not know the exact frequency where the 2 meter spur is located, but I do not find it on an FM radio at 145.000MHz on an outdoor antenna. Perhaps if I set an antenna in the shack I'd find it?

    Anyway very nice enclosure.

  • edited December 2017
    OK on 144 MHz. I track it roughly with an FM transceiver and an outdoor antenna, breaking a tight squelch from 144.000-144.010 MHz. It would never matter in my use of 2 meters. If it became a problem a 1/4 wave coax stub on a T connector, trimmed to attenuate the spur might greatly reduce the signal going to the antenna.
  • Collected my enclosure today and will assemble tomorrow. It is now out of stock at Seeed. 

    Bjarne Mjelde
  • edited May 2018
    Back on 14 April 2018,  Amazon in the US listed the enclosure for $18.50 US.  I saved shipping costs by purchasing from Amazon US.   Currently, (May of 2018) the listing in Amazon is showing out of stock.  Search Amazon for 

    Seeedstudio KiwiSDR Enclosure  

    which should find the listing to see if stock is available again.

    Larry W0AY

  • FWIW, Mouser USA is currently showing 4 enclosures in stock. I don't recall seeing this before. Shipping will be faster and less expensive than buying from Seeed for those of you in the USA:
  • Thanks John but I just cleaned out their stock buying 3 cases for my latest batch of Massdrop Kiwis which arrived this week. I hope that encourages Mouser to stock more.
  • I purchased one of the metal cases and installed it today. It definitely helps to eliminate RFI. Very happy with it!
  • edited June 2023

    Hi. I found that KiwiSDR in such metal case emits interference at frequencies above 65 MHz. The culprit of the interference is the LMR10530Y pulse converter. But the main problem is that the SMA connector thread does not have physical contact with the metal case. I replaced the end plate and there was much less interference. You can probably use the original end plate and large washers and nuts to ensure contact between the SMA connector and the main metal case.

    Short video:

  • My KiWi metal case modifications were illustrated in this previous thread.



  • Martin, thanks! Very interesting article!

  • I suggest being a bit cautious about simply grounding the SMA's, HF antenna and GPS, to the enclosure. In so doing you are creating another shunt path for common mode current through the entire structure, That is, for example, current coming in on the LAN going throught the Kiwi's ground plane and back out the PS.

    Whether the overall result is better than leaving the path out the SMA common choked may depend upon the characteristics of the injected current and the resulting path.

    In the past I've made PCB end plates to replace the stock ones, some having GPS chokes, some flux-coupled transformers and some shorted. I didn't find a clear winner for all situations.

    Glenn n6gn

  • In addition to KiwiSDR, I have several other receivers with remote access. And these receivers work on VHF. The installation of this end plate with a physical SMA contact on the shield completely freed the VHF receivers from interference. And by the way, these common-mode currents (in which I don't understand anything) most likely explain the absence of these interferences on HF.

  • Hi Glen,

    If you take a look at the spectrum analyser plots in the previous thread I linked to, you will see that even with the standard metal case, there is a significant amount of radiated noise at <150MHz , 450MHz, 750MHz & 1.5GHz.

    My concern was also that the KiWi was on a remote site with other VHF / UHF receivers and I wanted to avoid causing interference to them. Connecting the SMA connectors to the case made a significant difference in this respect. However I appreciate that doing this may also cause other problems with the KiWi and reception <30MHz, such as the common mode issues you have mentioned.

    You are definitely correct in saying that there is no clear winner, and I have found something that helps at one site often makes things worse at another. Experimentation is nearly always required to determine the best compromise.

    The only sure fire thing is that adding any type of metal case as a first step, rather than just using the supplied plastic plates and metal sides, is likely to significantly reduce the level of radiated emissions.

    To be honest I don't know how many of the manufacturers of single board computers get away with the level of emissions their products produce. If they were a domestic item they would not pass the various EMC standards that would be required to allow them to be sold commercially. I can only assume that because they are a partial assembly or kit, that is intended to be used in something else, this provides a loophole for exemption. Or maybe they just don't care...



Sign In or Register to comment.