• Author Author: colporteur
  • Date Created: Date Created
  • Views 160 views
  • Likes 7 likes
  • Comments 6 comments
Related
Recommended

POST 9 Partnering SDR & GNU Radio ...Red Pitaya STEMlab 125-14 PRO Gen 2 Starter Kit

colporteur

This my final post before writing the review. I’ve struggled up to this point to deliver on the objectives in my application. Well I am happy to report I was successful with wrangling the SDR part of the Red Pitaya into what I promised, a radio receiver

When I first read the call notice for RoadTesters on this product, I recalled assembling a AM/FM radio receiver using RTL-SDR and GNU Radio. My goal was to attempt that same exercise using the Red Pitaya.

Before starting the exercise I needed a practice run. Some project that used Red Pitaya SDR capabilities that worked. This article from controlpath was an example I could use. An input a signal to Red Pitaya DSP engine and play it back on GNU Radio.

I don’t have a workshop with test equipment. My signal generator was 5kHz generated from a sox application running on a raspberry Pi, out the audio jack to IN1 input of the Red Pitaya. The Red Pitaya network connection would carry a signal to GNU Radio running on my Ubuntu computer.

What it would proved was that I could input a signal to Red Pitaya. I could output a signal to the network from the Red Pitaya. I could capture the network output on GNU Radio running on my computer.

image

It worked! Those two words took close to six hours of effort to produce. I found nosie artifacts in output signals I couldn’t explain. In the end, I focused on the 5kHz signal. I varied the Pi output signal and GNU Radio sink output graph varied.

The Red Pitaya source block for GNU Radio came from Red Pitaya Notes . I could not nail down who exactly Pavel Demin was but it has a strong association with Red Pitaya development. Without that code there is no connection to Red Pitaya. The Rosetta stone was the links in this screen shot from the SDR transceiver startup screen.

image

The caveat was that my use case used different hardware or software versions. Warning in documents made it clear the code may not work. IP address the Red Pitaya source block correctly and GNU Radio makes the connection.

Getting a successful 5kHz test under my belt prompted me to shoot for the application objective. Build a radio receiver.

If I could replace the 5kHz signal with one from the AM radio band, could I build an AM radio receiver. The Red Pitaya documentation indicates analog input frequency range approximately 0–60 MHz. FM Radio wouldn’t make the frequency cut.

I confess this was a long shot. Pavel Demin’s Red Pitaya notes allude to the possibility. SDR Transceiver architecture, GNU Radio source/sink integration, and existing AM transceiver flowgraph (trx_am.grc). GNU Radio reference documentation for signal-processing blocks and their configuration offered some insight. My AI research assistant did the reading.

image

Before even starting I was skeptical. I didn’t have a proper AM receiver antenna. Antenna impedance didn’t match the Red Pitaya input. There is a balan available for purchase to handle the mismatch. I was already out money for the case, I was not going to purchase more. The last obstacle could I find a station that I could receive.

With a set of rabbit ear antenna connected to the input of the Red Pitaya I used the Spectrum Analyzer to scan the 500kHz to 1750kHz band looking for a signal. Once I found a signal, (the waterfall graph was most useful), I started assembling the parts.

Six hour later I was listening to a talk radio show at 990kHz. The voice audio was understandable. I cut my losses and didn’t spend time trying to clear up the static. The proof of concept was what I wanted.

I'm done kicking the tires and Red Pitaya STEMlab 125-14 PRO Gen 2 Starter Kit. My next post will be the review.

  • in reply to DAB

    I really don't like generating a less than positive review. The stuff I pointed out is fixable but, that requires effort on vendors part.

  • in reply to colporteur

    Consider your work a public service that might prevent a member for spending a lot of money only to end up with a useless product. 

  • in reply to DAB

    I've enjoy participating in the RoadTest program. I believe the programs adds value to the E14 community. I take pride in the responsibility and set high expectations for myself. I like problem solving but the dogged persistence can be a blessing or many times a curse. I'm penning this reply at three in the morning.

    My application suggested four blog posts each with an objective and the fifth blog post the final review. Nine posts is the current count, with ten being the summary report. I question why five posts has turned into ten posts. This RoadTest was a problem that I didn't think I would solve.

    I am disappointed that there are so few positives in my work. I estimate I have close to 60 hours invested and little to show for it. Nothing was easy. Post nine was two six hour sessions to get arrive at the application objective "Complex radio using GNU Radio on Ubuntu Linux"

    I don't consider myself a novice in technology and far from an expert. This RoadTest felt like I was doing development and that I was under developed for the task. You know what you wanted to accomplish but there is no feeling you are going to succeed. Just try. I have to say, I didn't like the feeling.

    I'm giving some time for the summary report to percolate before writing it. I'm looking to take what I have learned about the Red Pitaya and turn it into insight for someone to use.

  • Nice update.

    I admire your persistence, I would not have continued to flog this horse.

  • 5 kHz — Making Something Out of What Was Lying Around

    I needed an audio source, and I didn't have a signal generator. Time to improvise.

    I dug out a Raspberry Pi, installed SoX, and soldered a connector onto a cable. That gave me a 5 kHz audio output. Now, how to connect it to the Red Pitaya?

    A search through my collection of electronic junk turned up a coaxial cable with a BNC connector. I also uncovered a little test jig with clip leads that could connect to the Pi's audio output cable. Add a BNC-to-SMA adapter, and I had a connection to the Red Pitaya.

    Time for a coffee and bathroom break.

    Now for the other end of the circuit: GNU Radio.

    Installing GNU Radio presented its own challenges, and challenges consume time. Next came locating and installing the Red Pitaya source and sink blocks, followed by combing through documentation to figure out how to configure everything.

    I started this project around 4 PM, after work. By the time I climbed into the shower a little before 10 PM, I had a working system.

    AM Radio — Can I Hear Something?

    With the 5 kHz experiment behind me, I wanted to see whether I could receive an AM radio station.

    First, an antenna. I sampled several antennas I had kicking around, looking for something with a usable connector and at least a fighting chance of working at AM broadcast frequencies. I settled on a four-foot rabbit-ear antenna, the kind you see perched on top of televisions in old movies.

    I loaded the Red Pitaya spectrum analyzer and spent some time figuring out how to configure it. The waterfall display at the bottom was ideal for what I wanted to do. The frequency range controls, however, are something I still haven't mastered. Make a mistake and suddenly you're looking at the wrong frequency range. Searching for a signal in the kHz range while looking at the wrong part of the spectrum doesn't work very well. :)

    Once I narrowed the range, I worked through the antennas, looking for a blip suggesting I had a receive signal. I found three.

    An internet search turned up call signs for radio stations operating around those frequencies around my area. The signals were way down in the mud. Not surprising, considering the mismatch between the Red Pitaya's high-impedance, capacitive-coupled input and an antenna arrangement that wasn't exactly designed for the job.

    But the waterfall kept painting a red line at the same frequency. That suggested there might be enough signal to recover some audio. Back to GNU Radio.

    I started with the layout from the 5 kHz experiment, replacing the audio signal with the antenna broadcast. The Red Pitaya source block showed a signal, so at least I knew something was getting through.

    Now came the challenge of turning that signal into intelligible audio.

    I had to research and make some AI assisted guesses about the values and settings for the GNU Radio blocks. Knowing that the Red Pitaya was delivering a signal gave me enough confidence to keep working through the processing chain. 

    I hadn't touched GNU Radio in five years, and I wasn't particularly interested in relearning all the details. So I cheated. A confession: ChatGPT did much of the GNU Radio reading.

    Tell me that a block needs a particular input type, or that its output must match the next block, and I can follow along. Ask me to remember which settings make a particular combination of blocks work, and that's another matter entirely.

    The first working GNU Radio configuration produced audio, but it was extremely faint. Voices were there, buried under a substantial amount of static.

    Giving ChatGPT the problem of low audio it suggested adding a Multiply Const block to increase the signal level. That made the voices louder. I could actually understand the conversation.

    It wasn't pretty. There was plenty of background noise, and nobody would mistake it for a commercial AM receiver. But I was listening to a radio station through the Red Pitaya, with GNU Radio doing the demodulation. That was the result I was looking for.