Building a data acquisition system to capture raw IQ data

Question

I was planning to build a data acquisition system to capture raw IQ data for a hobby project and the short-term goal is to gain practical experience on RF front ends and ADC's.

My background is mostly Wireless DSP implementations in MATLAB/C (and some basic GNU Radio experience using USRP B200).
Did anyone try something like this ? Maybe using all individual components from mouser and solder it up and then process data on raspberry pi (maybe)? Does anyone have any recommendations/any other blogs that talk about such experiments ?

PS: As long as frequency range is concerned , looking for 5GHz or 6GHz band

Farheen

Answer 1

As a front-end radio designer with experience in the microwave frequency ranges, I concur with the advice to just buy a solution such as the ADALM-PLUTO SDR Learning Module. However that particular solution has a maximum frequency of 3.8 GHz and 20 MHz bandwidth, so if acquiring 5-6 GHz signals is important you would not be able to do it directly with that. If going even lower to 2.3 GHz can be acceptable, these much lower cost RF receivers may be attractive:

https://www.nooelec.com/store/sdr/sdr-receivers/nesdr.html

If you must get to 6 GHz, and can live with <10 MHz BW, this HackRF SDR is an option at $330:

https://www.nooelec.com/store/sdr/sdr-transceivers/hackrf/hackrf-bundle-100715.html

Anything else that provides both the high carrier frequency and wider bandwidths that I am aware of starts getting above the $1000 price point and higher.

To do this with discrete parts will be much more expensive in comparison and require specialized layout practices (especially at the 5-6 GHz frequency ranges, this is definitely not something you "can just solder up" with parts from Mouser; the shape of every trace and wire, and it's proximity to ground has a significant impact on operation and performance). To avoid doing the layout, you could buy connectorized eval boards and/or connectorized components, but those are even more expensive.

If lowering the carrier frequency (significantly) is an option, this can open up a lot more possibilities if the idea is to have hands-on experience with receiver design and ADC's at the component level. But if the goal is just data acquisition of RF signals for subsequent processing, then I recommend the other solutions mentioned.

If the goal is to learn RF design techniques, here is a course coming up in the Boston area on PCB layout techniques and considerations for RF design:

https://ieeeboston.org/event/practical-rf-pcb-design/?instance_id=3308

Not cheap, but Besser Associates offers a course on RF Design:

https://www.besserassociates.com/Courses/Course-Description/CTID/248?gclid=EAIaIQobChMIufucue6j-gIVuzizAB3TwwPTEAAYASADEgJJr_D_BwE

For a lower cost introduction, the ARRL Handbook (and the Ham Radio community) is a good next step in getting introduced to the larger nuggets involved:

http://www.arrl.org/news/the-2022-arrl-handbook-for-radio-communications-is-now-available

Answer 2

Can you use any of the NI DAQ modules? Or does it need to be designing/soldering an RF front end from scratch and then reading it on some embedded processor/doing some FPGA?

I think option (1) is pretty straightforward as you can get all this info from the NI website (i remember seeing some blog/white papers/similar questions on NI forums)

I'm not really sure of option (2) as that's pretty time-consuming and yes, if that can be done or any blogs available, its a very practical way of learning RFIC design

Ash

Answer 3

I'd suggest buying a Adalm Pluto board. It's very inexpensive.

Answer 4

As Dan mentioned , check ADAM Pluto This might be a good starting point . I didn't look at the lab tutorials 1 and 2 and electronic tutorials 1 and 2 , but looks promising. You can probably order the ADALP2000 kit and go through all labs .

https://www.analog.com/en/technical-articles/the-world-is-analog-how-do-i-obtain-the-necessary-knowledge.html

Ash