At first glance, the ColibriNANO SDR looks like another cheap SDR dongle. But after watching [Mile Kokotov’s] review (see video below), you can see that it was built specifically for software defined radio service. When [Mile] takes the case off, you notice the heavy metal body which you don’t see on the typical cheap dongle. Of course, a low-end RTL-SDR is around $20. The ColibriNANO costs about $300–so you’d hope you get what you pay for.
The frequency range is nominally 10 kHz to 55 MHz, although if you use external filters and preamps you can get to 500 MHz. In addition to a 14-bit 122.88 megasample per second A/D converter, the device sports an Altera MAX10 FPGA.
In addition to interfaces to different software packages, the dongle works with remote software. The idea is to put the dongle and an antenna somewhere advantageous (that is, high and radio-quiet) and then use a Raspberry Pi or similar computer to pipe signal over the Internet.
If you don’t want a dongle, we can endorse [Lukas’] build from scratch. If you are looking more for a getting started resource, check out what [Richard Baguley] had to say about SDR.
>14-bit 122.88 megasample per second A/D converter
and 3 MHz bandwidth, its junk
I haven’t watched the video, don’t know anything about the product, but I do know that the effective precision and sensitivity can be increased by oversampling and decimating. That is, it takes 32 samples for I and 32 for Q, does digital filtering, resulting is more than 14b of precision. For some uses, an 18b 3MHz signal is more useful than a 14b 60 MHz signal, especially since it is hard to get that much data off of a USB device.
Can you explain your reasoning ? – Especially with regard to Johnson–Nyquist noise and dynamic range.
Looking at the block diagram the design looks good – http://trcvr.ru/wp-content/uploads/2017/07/ColibriNANO-Blockdiagramm.jpg
It is designed for HF (DC to 55MHz), and it is not like there are many signals wider than a few kHz apart from some really high power over the horizon radar systems.
And how does it compare with this, which has been around for ages already? As I understand, it’s bandwidth is all the HF spectrum.
http://uvb-76.net/p/sdr-mk15-andrus.html
Also the receiver used in one of the WebSDR nodes (but no schematics and design documentation):
http://websdr.ewi.utwente.nl:8901
It’s for LF though and no higher than 55MHz. So you only need so much.
Still, $300? uhm.. is there really that much interesting stuff on shortwave?
But if shortwave is your thing then I guess it’s OK
I agree, JUNK, especially with the cost of the unit! KV4WM
Wow! $300, that’s right up there with some of the other nice SDRs.
Ahem, I’ll just leave this here…
https://www.crowdsupply.com/lime-micro/limesdr
And when it comes to HF performance the above device will run circles around it. Amazing performance in HF bands is not an area where a LimeSDR is optimal.
Lime is destined for VHF, UHF and microwave – where all the modern wireless communication happens, that’s what it’s RF ASIC was originally designed for…
I used to be skeptical too, but the hardware RF pipeline reasoning becomes clear when you dig into the lectures.
Recall the lime is also TWO full duplex systems with multiple RF input switching to select different HF/VHF/UHF line inputs.
I was used to looking for LNA and filters for SDR (specifically notching out FM 88MHz-108MHz to reduce the noise floor), but when I got my lime it already had respectable amps built-in to the front end.
The app notes can be thick to read, but it was designed by people that wanted everything but the power amp and output filter on a single chip. Now if someone makes a dual wide-band digitally tunable 1W bandpass filter, than the device will be 2/3 complete digital transceiver/repeater.
More Data does not always mean better SNR.
I have 3 various chipped RTL-SDR dongles. All work great; scanning, trunking… They’re all on a shelf now that I have a LimeSDR. Killer setup, tons of bandwith.
since when is 14 adc bits equal to 110 dB dynamic range. … its ~6db per bit folks.
There is this DSP thing called oversampling where, provided there is enough noise!, you can increase the dynamic range by sampling at a higher sample rate. For a dynamic range of 110 dB, that would imply 18 bits, since the ADC above only has 14-bits (~86.dB of dynamic range ), that would imply that they are gaining 4 additional bits (Or an additional 24dB of dynamic range) by oversampling – 122.88 MSPS real @ 14-bits is 61.44MSPS IQ @ 14-bits and if you decimate that by 256 for 4 additional bits of dynamic range that is 240 KSPS IQ @ 18-bits.
To decimate means to remove 1-in-10, as in decimal. Or as used by the Romans, line up everybody and count off. Every tenth one is killed. It is true that, like news readers, DSP jargon uses a different meaning, I wish there were something that makes more sense. Removing 1-in-10 will almost never happen in DSP. Hexamate? Octomate? It doesn’t really work since in DSP you are removing 7-of-8 or 3-of-4, etc.
“Upsampling” and “Downsampling” are also used, but have been overloaded in lay speak (to include the antialiasing filtering) such that in practice they’re more imprecise.
“Decimate” is a sufficiently rarefied term that it doesn’t have that problem.
There is good explanation of ADC dynamic range here: http://www.qsl.net/z33t/dynamic_range_eng.html
The 14 bit ADC on 122.88 MSPS is equal to about 86 dB dynamic range, but with decimation process, when you decreasing the spectrum bandwidth to 48 kHz you gained to 18 bit with 110 dB dynamic range!
>when you decreasing the spectrum bandwidth to 48 kHz
which you can get with $50 consumer sound card(UMC202HD/Asus Xonar DSX) SDR these days …
http://www.telepostinc.com/soundcards.html
Its junk. Good SDRs give you all of the bandwidth over USB 3.0/pcie, and let YOU decide what to do with it (oversample or listen to whole >100MHz band).
I guess your main problem is that the stick is made by russians.
NAH. Reads like a sponsored post/advert.
Sorry you thought so. I don’t think it reads that way.
We absolutely, categorically don’t do sponsored posts or sneaky adverts. I know the rest of the Internet is going to hell with twitter bots and fake news and clickbait and etc. But this is Hackaday.
Meh!
DC to daylight receivers were a thing advertised in the back of electronics magazines that I read over 20 years ago.
Waterfalls are nice but 3MHz isn’t all that wide and that’s a lot of money.
I’ll stop being bored when a ‘good’ SDR doesn’t cost $300!
Then don’t buy it. I guess you don’t care about HF performance, only being able to receive some IoT crap or something.
Can some one tell me why the SDRs in general do not go down to the 1-10 Khz and lower, thinking of applications with a metal detector
On the face of it, it doesn’t compare value wise against the SDRPlay. 1KHz to 2GHz, version 1 is now US$99 and version 2 US$170
http://www.sdrplay.com/docs/RSP2_Datasheet.pdf
It is a nice piece of technology shrinked in to the size of a USB-Stick. I would like to see a software working on android.