Interpreting Output Signal

[GokcenMerve]

Hi Margarita, I'm a young engineer and working on Gpr data.I've compile Python code on mbed as described in your documentation.I have no problem until interpreting output signal. I do not understand about interpreting (i.e myRecording = getaudiodata(recObj)).What is the GPR_FMCV? How can i see GPR_FMVC screeen? Are there any MATLAB Code about that? I'm waiting for your help. Thanks, Best Regards, Gökçen Merve HORUK

[Ritchizh]

Hi Gökçen Merve, I'm glad you found the Project interesting. Have you already built the electronic part of the radar, on what stage of radar implementation are you now? What kind of antennas do you use?

I participated only in this radar firmware and software development during a short-term scientific mission in 2017. It concerned only the transmitter part.

This project was launched by COST Action TU1208 - However, they had not published all the necessary files, so I shared what I had. The radar was implemented at the Sapienza University, Rome, including two log periodic antennas, and presented at Maker Fare Conference. Unfortunately, I don't know the details about the antennas and the MATLAB software used to visualize the received signal.

This repository containes firmware for the Mbed microcontroller - it makes the VCO generate the desired transmitted signal. And it containes software in Python - that creates a GUI to chose the transmitted signal parameters and send them to Mbed microcontroller.

As for the received signal - in the "Guide" a sophisticated MATLAB GUI is shown - named "GPR_FMCV" - since we don't have it, it should be replaced by some simple Python / Matlab code to read the data from audio channel.

Kind regards, Margarita Chizh

[GokcenMerve]

Hi Margarita, We use electromagnetic antenna in project. We haven't combined the electrical circuits yet. We printing pcb boards now.We are also trying to improve ourselves in software. Thanks for your helping. We may have questions later. Can I ask you again? Best Regards, Gökçen Merve HORUK

[Ritchizh]

Dear Gökçen Merve, Sure, I will try to answer your questions, feel free to ask. Good luck with your work! Kind regards, Margarita Chizh

[GokcenMerve]

Hello Margarita,

In our previous talks you asked me that 'What kind of antennas do you use?' When we read the TU1208 (Guide to Building a GPR Radar) , we noticed the circuit is connected to two green triangles. The picture I added also shows. How should we design and optimize the antenna?Which parameters should we consider?

Best Regards, Gökçen Merve HORUK

[Ritchizh]

Dear Gökçen Merve, the antennas are the most challenging part of this GPR design. As I wrote in the project's description there is no tutorial for antennas so far. The radar was implemented at the Sapienza University, Rome, including two log periodic antennas ('green triangles'), unfortunately, I don't know the details about them.

The choice of antennas depends on your transceiver frequency band and desired application - you may look up for the typical antennas used in your area of interest.

Originally in the MIT course, on which the COST Action radar project was based, they made circular waveguide antennas from coffee cans. Such antennas have too wide directivity pattern for GPR application - they were tested in air for movement detection and target imaging. For the first test you could probably also use such an antenna, (explore the MIT course for inspiration: https://ocw.mit.edu/resources/res-ll-003-build-a-small-radar-system-capable-of-sensing-range-doppler-and-synthetic-aperture-radar-imaging-january-iap-2011/projects/ ).

Then, for a well functioning GPR, horn antennas are needed. Their parameters can be standard, and they are well known - the problem is usually the fabrication, because a drawing is required and metalworking is usually expensive. Also some commercial antennas could be bought.

There is also an option of producing some unconventional antennas - like printed on a 3D printer from plastic, polished and covered with metal paint. There are researches that show that such antennas have characteristics comparable to ones grinded from metal. Such antennas are really cheap and easier to reproduce in an education project.

At this moment, several researchers from different countries are trying to build a GPR based on this work, so I hope someone will finally succeed in creating the antennas and share the instructions.

Kind regards, Margarita Chizh

[GokcenMerve]

Hi Margarita,

I'm very grateful for the information you shared. I'il work on that. Thank you so much.

Best Regards, Gökçen Merve HORUK

[Ritchizh]

Dear Gökçen Merve, You are welcome 🙂 I hope, it helps.

Kind regards, Margarita Chizh

[GokcenMerve]

Hello Margarita,

After my last message, we completed the circuit and made an antenna. Circuit elements and python codes are working, but in Matlab we have problems getting and processing the sound wave. We could not find the concepts of Doppler, Sweep time and SweepBandWidth in the TU1208 GPR document. Do you have any information about these parameters? Do you have a document about receiving a radar signal and finally how can I contact other people (Andrea Pietrelli and Vincenzo Ferrara) involved in the project?

Best Regards, Gökçen Merve HORUK

[Ritchizh]

Dear Gökçen Merve,

This is great news, could you please tell more about the antennas - which type did you select and how did you produce them?

You don't really need to use Matlab for signal processing - I would use Python for it - to read and save data from audio channel and then visualize it (image the recorded dependencies of received signal amlitude and frequency on time). At first you should prove you can see the response to an object moving near the antennas. Then you should make a calibration experiment with a metal target at a known distance - measuring the frequency modulation allows calculating the distance to the target (https://www.radartutorial.eu/02.basics/Frequency%20Modulated%20Continuous%20Wave%20Radar.en.html) - when you get this distance right, you can be sure the radar works. I think, this is the only way of developping radars - delving into the physics. Ofcourse, you can try to adopt the advanced Matlab software (https://www.mathworks.com/help/phased/ug/fmcw-range-estimation.html) - studing the provided documentation, but there is always a chance of missing some detail. In what context were these terms (Doppler, Sweep time and SweepBandWidth) mentioned? I assume, Sweep time is equal to one period of the sounding wave (time the radar needs to emit signals at all programmed frequencies); SweepBandWidth is the range from minimum to maximum emmitted frequencies. The specific values depend on your application - the size of the target, distance to it, whether it is moving, properties of the medium, etc. I don't have the document about receiving a radar signal - but I think that you have already completed the most difficult part of the project, and processing the received signal is rather straightforward once you apply the physics principles. Professor Vincenzo Ferrara is working on other projects now - I could not reach him by email to ask for other files. Andrea Pietrelli is my friend, he is working in other area of research and has only the same files as published in this repository. There is also a research team from University of Split, Croatia, supervised by Maja Skiljo, who had implemented the radar following the tutorial - but on my request she sent me all the same files as I shared here.

Kind regards, Margarita Chizh

[GokcenMerve]

Hello Margarita,

We used the Broadband Log-Periodic Antenna. Unfortunately, we do not have detailed information about antenna design. As a result of our research, we made the drawing using Proteus (Proteus is a capable circuit drawing, simulation, animation and PCB drawing program that can visually simulate electronic circuits.) And we printed the double-sided printed circuit board. In this way, when we tested the circuit with a spectrophotometer, we saw that the signal wave was emitted.

We shared your informing about Doppler, Sweep time and SweepBandWidth with our Supervisor. According to our Supervisor's answer, I will share with you the context of Doppler, Sweep time and SweepBandWidth parameters.

Best Regards,

[olivierace]

Hi Gökçen Merve HORUK,

you reproduced the system?

I would like to know if it is working well? How deep can you see objects and on what surface area?

I am very interested in producing a GPR that can see at great depths and over a large area of ​​land! If it's possible !

What do you think ?

Thank you Olivier