RX-666: Listen to A Gamechanger!

Flying high: RX-666

Since Marco Polo, combinations of Italian and Chinese had proven fruitful. The Dragonfly RX-666 is such a combination or, at least, a very special “kind of”. It is a 16bit SDR, ranging 1 kHz to 30 MHz in a row with 16 bit resolution – plus some extra above 32MHz (1.5/1.7GHz) by the help of an R820T2-chip of RTL-SDR fame, but 8 bit only. Priced not much over US-$ 200 (if that) at some ebay sellers, it is a real gamechanger, offering for the first time 32 MHz streaming via USB3.0 at 16 bit resolution to ensure a competitive dynamic range.

32 MHz at a glance! And just 15% CPU load on an i7.

It has been built around LTC2208 chip and seems a clone, a twin or a pirate piece of a concept, literally layed/laid out by Oscar Steila, IK1XPV, an electronics engineer from Turin. Über-DXer Bjarne Mjelde has diven deep into the story and the receiver itself. I don’t want to repeat what he found out in only my own words. You simply must read his story here, and I can stress each and every word of it.

In this blog, I may add only some audio clips to give an impression of the reception quality in Northern Germany on the evening of August 11 and the morning of August 12, 2020. Antenna, as always, is the professional active dipole MD300DX, vertical with 2 x 5 m legs. Recordings were made with HDSDR software, but replay done with SDRC V3. Both, Bjarne and Oscar, helped me to get the SDR flying at all. Without their friendly hand, it would have been only another heavy paperweight on my desk …

Now for some twelve audio clips:

IRIB Ardabil/IRAN, 1512 kHz, 20:00 UTC in Farsi
RUssian Aero Net with control station like Samara and Ural, 4712 kHz, 20:05 UTC, each running 1kW with a low-hanging dipole; in Russian.
Echo of Hope, clandestine from Suwon/South Korea on 4890 kHz at 20:00 UTC in Korean.
BPM, Chinese time signal station in AM with CW-ID and voice ID in Chinese on 5000 kHz at 19:59 UTC. Please compare to reception during the same time on 10’000 kHz below.
Voice of America/Botswana on 6.080kHz duly doing for what they are payed for: Hail to the Chief, hail, hail, hail! 20:00 UTC
The other chief: TWR Botswana on 6130 kHz in Kimbundu language – Amen. 20:00 UTC
Singapore VOLMET on 6676 kHz with weather e.g. for Kota Kinabalu, Bali, Penang and Singapore/Changi, 19:50 UTC
BPM, Chinese time signal station in AM with CW-ID and voice ID in Chinese on 10’000 kHz at 19:59 UTC. Please compare to reception during the same time on 5000 kHz up.
World Christian Broadcast from Mahajanga/MALAGASY transmitter on 11’965kHz at 20:00 UTC
BBC via their Atlantic Relay/Ascension on 12’095kHz, 20:00 UTC
Radio Kuwait from their homeland on 15’530kHz in Arabic, 05:00 UTC
China Radio International from Kashi/Xinjiang on 17’720kHz in German

The range above 32 MHz is covered by an R820T2-chip at maximum streaming bandwidth of 8MHz only. Please see below a screenshot of a part of the FM broadcast band:

FM broadcast is covered with an R820T2-chip at a maximum streaming bandwidth of 8MHz only. Here you see a part of the FM broadcast band.


  • The worst name for a ‘toy’ . In a Facebook group advised that the nominal bandwidth is even smaller , around 20 MHZ while the voltage polarity is reversed . all the job is in the CPU and does not work with HDSDR
    more from Bjernes comment in a SDR email list https://sdr-radio.groups.io/g/main/message/59213

    • Hi – they use a very good A/D chip. And performance on HF is stunning in many aspects. I think, we will see this type of 16bit/32MHz SDRs for under US-$ 250 becoming the standard for usHF DXers in the next years to come. With my Winradio Sigma and the pre-production sample of FDM-S3, I am already using this type of technology for about one year and it leaves only minor things to wish (ok, GPS synchro among it). With such type of receiver, the development of receivers seems to have come to an end – to an end of what we DXers can use. It is now entirely up to us to use this technique and to to develop interesting applications … 73 Nils, DK8OK

  • You made me curious, hi hi. Can’t wait to see more details and your comments. In what price range is this receiver? Is it also 16 bits above HF which I guess it is?

    • Hi, Ron – Bjarne Mjelde has already prepared a no-fuss, hands-on test report of this beast. He will put this on his blog shortly, as addendum to his blog of August 9th. The 16 bit/32 MHz resolution/range is available only on HF. And the costs are around 200 US-$ from several sellers at ebay. Delivery takes just roughly a week. Keep an eye on Bjarne’s site, please: http://arcticdx.blogspot.com
      73 Nils, DK8OK

  • Hello Nils, your comments (and Bjarne’s) are very interesting on this new SDR development. It appears that the related RX-888 model does away with the unneeded and confusing 5VDC connector, includes a LNA, and has a higher precision TXCO. Besides the HDSDR functionality, the EXTIO should in theory let the RX-666/RX-888 work with Studio 1 and the EXTIO version of SDRuno.

    I would appreciate your observations on how this new SDR’s dynamic range may help it resolve weak DX signals on MW next to powerful signals, perhaps in comparison to other models you own.

    Best DX,


  • … thanks, Guy, we will see that. So far, I (we) have only the RX-666 here. As soon as there opens any opportunity to buy an RX-888, I will order it. Anyway – it seems that this concept (to say the least) will open a new chapter of SDRs: professional performance at a ridiculous low price. I am just doing some tests & comparisons which I hope to publish here in due time. But I fear, there is nothing substantial to add to what Bjarne already had found out. To be honest: without his & Oscar’s help, my Dragonfly even wouldn’t have taken off … 73 Nils

  • Hi Nils, thanks for adding those interesting audio clips! The demodulation seems overall very clean and low noise. There’s an excellent clarity to the voice reproduction.

    I have a RX-888 in transit now, so I’ll be joining the early adopters to see if we’ve been wise with our purchases, or not :^)

  • Hi Nils, very interesting news this is! Question on the comparison pic between RX-666 and FDM-S3: I see that the SNR is 77,6 dB for both receivers, but the peaks seem the same and the FDM-S3 has a higher noise floor. How come? It also looks as if the FDM-S3 has some artifacts the RX-666 hasn’t, like the two peaks on the right of the data signal? Can you elaborate? 73, Cornel

  • Hi, Cornel – thanks, and I can only show “as is” for now. Will come back to this with the RX-888 at hand in September … 73 Nils, DK8OK

  • Dear Nils,

    many thanks for testing this unit. I saw it some time ago and was very curious what the first serious test would reveal – here we are.
    I took a close look on your side-by-side screenshot and the noise floor seems lower but interestingly enough each signals S/N isn’t really better, because they where taken at a different time, it kind evens out. So it could theoretically be the same RX with the visual gain slightly turned down and captured at a moment where the main signals fade is e.g. 10dB up, matching the other shot.
    My question towards the above 30MHz RX is, are you sure they don’t use their fancy 16 Bit A/Ds behind the R820T2? That would be really dumb, because then you are better off running a 20$ TCXO RTL-SDR completely independent in parallel on a separate port WHILE receiving HF. Look at the Airspy R2 – they have a 12Bit A/D behind it and it is a great receiver, probably even better with 16 Bit though the R820T2 may not be able to produce the full dynamic the back-end can handle.
    Are you really sure the VHF/UHF part is running in 8 Bit mode, that would be a ridiculous design flaw in my opinion?

    Regarding the RX-888 it seems to have a GPS-Receiver on board as well – very curious – if that thing could at some point support TDOA, what a gift would that be!


    • … thanks, Jörg – and oooops, I published a wrong screenshot. The recording was indeed made in parallel, as the screenshot should have shown, but I failed to do this … If I find time, I will correct this. RX-888 underway, More information in due time. 73 Nils, DK8OK

    • I saw that “GPS” on the silver shielding box, too. I couldn’t believe there is actual GPS functionality on board, so I assumed it is just a portion of the part number of the TXCO. I hope I am wrong though :^)

      • It could very well also just have been some lettering with a different meaning because there is just a tiny internal antenna connector and a fixed external antenna connector would obviously be needed. While the internal might just be a test point for troubleshooting the oscillator unit.
        The future will tell.

  • … if I interpret it right, I see a small MPX(?)-socket for connecting an active GPS antenna just above the shielding … It makes only minor sense with not cable running out of the shielded case with its impressive cooling fins, but I also saw it at some other SDR … only inspection will tell ..

  • Ah, hopefully that is a correct guess, Nils! When I noticed that part I assumed it was a trimmer cap, or trimmer resistor. It does look like one of those tiny antenna connectors too.

  • Ladislav OK1UNL

    On ebay or aliexpress are now copies KIWI SDR but with 16bit ADC, Raspberry Pi3B+ and range up to 62MHz
    I am afraid there is not FW support. What do Think about this one marvellous offer??
    VY 73! Ladislav OK1UNL

    • … interesting, Ladislav. But first I want to look at the RX888 with, hopefully, the same A/D converter. Already ordered a week ago, it should arrive each day … 73 Nils, DK8OK

      • Ladislav OK1UNL

        Niels, in China must be LTC2208 16bit A/D really cheap or they have some copy of original Analog Device.This one A/D also using copy of KIWI SDR. Sure, the RX-666 or 888 for HF and VHF/UHF needs suitable LNA with MMIC
        like GALI -5 or INA02186. So if you need any usefull design we can send email to Siniša YU1LM/QRP. He is really experienced RF designer. Greetings to all VY 73! Ladislav OK1UNL

  • Whats the secret for the usb driver for the rx 666. Seems like i have all the files in place for HDSDR but it still cant find the device?

    • John – IT things are always secrets (for me, at least …) 😉
      The RX-666 is an USB device. It appears at your PC under the name of “WestBridge”. You have to locate this entry and renew the driver with the driver provided by the seller of the RX666. Then it works. A re-boot is always a good idea. 73 Nils, DK8OK

      • What’s the file name for the driver? I copied all the files they sent into the hdsdr folder.

      • John – it rests in the Driver folder. You may either let Windows search for the matching driver (activate something like „search also in sub-folders“, or pick the matching one manually – with me: Drivers -> Win10 -> x64. Let me know the result. 73 Nils, DK8OK

  • Moritz Schweizer

    Hello Nils
    Yesterday I go a Dragonfly RX103 which is claimed by the vendor to be the successor of the 888 and 666. Unfortunately I haven’t got any software with it. So I installed HDSDR but the problem is that I did not find an appropriate .dll. Have you got an idea?
    73 Maurice HB9CDO

    • … thanks, Moritz – I fear you have to ask the vendor … most of them do have a very fast return.
      What had worked for the RX888, might work also for the RX103:
      The receiver announces itself as “WestBridge” in Window’s Device Manager. For this, you have to install the matching “Cypress” dirver which must/should be supplied by the vendor. It maybe available also elsewhere “in the web”.
      After installing this one, my RX888 showed up as “CypressFX3 USB Streamer Example Device”. And this one will talk to HDSDR or, in my case, SDRC V3. Good luck & 73 Nils, DK8OK

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 )

Connecting to %s