Announcement

Collapse
No announcement yet.

Implementation of Transmitter and Receiver for GPR

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Penetration will not change for better with high gain antenna because in near field it performs worse than the low gain antenna. All the gain of any antenna is attributed solely to the far field sum of all the signal paths from the antenna geometry ... which in case of high gain antenna you don't have when you turn your antenna facing ground.

    The most important feature for an antenna you can look forward is broadband coverage because your distance resolution depends upon it. With 1GHz bandwidth you have ~30cm resolution

    Comment


    • #17
      Yes, that's the system I want to design but still dubious about it. I don't know where to start. ;(

      Comment


      • #18
        Davor, you so want me to change the antenna. :P

        Anybody read the Seeker's low cost GPR?

        Comment


        • #19
          You probably have seen this page. To me this was the closest to being a diy gpr. They went quiet with no further details. You can see photos of the antenna and circuit boards.
          I think it ran in the 300 to 600mhz range?
          http://www.pulsediscriminator.net/

          Comment


          • #20
            Yeah, saw it. I am late, they took it commercial quite some time ago so no chance of help from them.

            Comment


            • #21
              Originally posted by Altra View Post
              Does any one know how to compress a pdf file? Trying to upload a 2.4mb which exceeds 1mb limit
              Compressing a PDF file will not help much with the size.
              However, it is often useful to compress the PDF in multiple rar files to get past the size limit. WinRAR can be used on Windows.

              Comment


              • #22
                This book might be worth reading ->
                http://www.amazon.co.uk/Ground-Penet...2303033&sr=1-1

                and these publications are downloaded for free ->
                http://www.ees.nmt.edu/outside/Geop/...vey_design.pdf
                www.eltesta.com/resources/application-notes/PDF/AN-1.pdf
                http://groundprobe.site.aplus.net/ur.../bellett01.pdf
                http://rrsg.uct.ac.za/theses/phd_the...man_thesis.pdf

                I suspect these will either provide some inspiration, or scare you off the whole idea of GPR.

                Comment


                • #23
                  This is interesting one, I almost missed this post. Challenge is interesting too. If I understand this properly, idea is to emulate VNA functionality in some hardware and use it as GPR.
                  Good idea (considering decent VNA is more expensive than decent GPR).

                  Not so impossible for DIY, but some experience with RF stuff is needed. (Long time ago, I was slightly more enthusiastic about…) Some parameters must be determined first, and then more than one solution can be found, about in this order:

                  Aside freq. range required, dynamic range and sensitivity (noise floor) is important, then phase noise requirements for LO, then IF processing.

                  In order to emulate VNA operation in S11mode, most critical, but surprisingly DIYable part is decent directional bridge (well, considering VNA is available to adjust it). Not even need to be too wideband. From this point, all performances will be limited by directivity achieved. If you make it better than 50dB directional over 2.4GHz, 1GHz span, you deserve a kingdom.

                  With given dynamic range, receiver part can be simple, actually some Analog Device log-amp chip with amplitude and phase capability (AD8302) will do the job with enough sensitivity at given freq. To fix sensitivity issue, some cheap amp like MMG3004\5\6 can boost TX power enough, or some wideband amp can be added at AD chip input.

                  Then, phase noise is not an issue anymore, due to limited dynamic range, low resolution will permit cheap and dirty design using some popular prescaler\PLL chip, like in SAT tuners, TSA5055 or similar, and discrete VCO (or some cheap Chinese brick). For S11 operation, and AD log-amp for phase extraction at operating freq. only single source is needed, no IF and another LO.

                  This will not make state of the art performance, but can be almost credit card sized and not too expensive, on board controlled with some low end MCU as interface.

                  (Or at least, will make S11 capable VNA, GPR use not guaranteed)

                  Only thing unclear is, why someone needs to build SFCW GPR in order to get degree in Philosophy??? (Article from previous post)

                  Comment


                  • #24
                    Thank you Qiaozhi and Tepco for replying and keeping this post alive. It means a lot to me.

                    @ Qiaozhi

                    I have that book in hard copy. Its useful for those who are interested in getting a good idea about GPR. What I require is a bit far-fetched than its scope.

                    At the moment, I am scared of my choice of this project. :o

                    @ Tepco

                    Thank you for NOT missing this post and for your detailed reply.

                    As I stated, I am a bachelors student. Don't know much about Microwave/RF Engineering as we only studied a single subject of it.

                    About the 'kingdom deserving part', a lot of people said that to me that IF you do this, Gold Medal is all yours. No one will even think of snatching it from you.

                    But the problem remains the same, I am looking at a sea of material on GPR and don't know where to begin. I studied some papers, they were helpful but it looks like i'll be designing the circuitry myself as there are no schematics available anywhere.

                    I'll look deep into your idea to. A lot to learn.

                    I have a few questions. What if I totally get out of the idea of UWB SFCW GPR? Design FMCW GPR? It seemed a bit approachable to me as it lies in MHz range, not GHz thus decreasing cost of manufacture. But I'll need to change antenna for it. Not an issue if I make the transceiver part.

                    Any thoughts about this?

                    || Hoping you would keep energizing this post.

                    Regards

                    Comment


                    • #25
                      I think that GPR is one of the dreams of all serius detectorist and i hope some day we can build one unit.




                      Originally posted by Tepco View Post
                      This is interesting one, I almost missed this post. Challenge is interesting too. If I understand this properly, idea is to emulate VNA functionality in some hardware and use it as GPR.
                      Good idea (considering decent VNA is more expensive than decent GPR).

                      Not so impossible for DIY, but some experience with RF stuff is needed. (Long time ago, I was slightly more enthusiastic about…) Some parameters must be determined first, and then more than one solution can be found, about in this order:

                      Aside freq. range required, dynamic range and sensitivity (noise floor) is important, then phase noise requirements for LO, then IF processing.

                      In order to emulate VNA operation in S11mode, most critical, but surprisingly DIYable part is decent directional bridge (well, considering VNA is available to adjust it). Not even need to be too wideband. From this point, all performances will be limited by directivity achieved. If you make it better than 50dB directional over 2.4GHz, 1GHz span, you deserve a kingdom.

                      With given dynamic range, receiver part can be simple, actually some Analog Device log-amp chip with amplitude and phase capability (AD8302) will do the job with enough sensitivity at given freq. To fix sensitivity issue, some cheap amp like MMG3004\5\6 can boost TX power enough, or some wideband amp can be added at AD chip input.

                      Then, phase noise is not an issue anymore, due to limited dynamic range, low resolution will permit cheap and dirty design using some popular prescaler\PLL chip, like in SAT tuners, TSA5055 or similar, and discrete VCO (or some cheap Chinese brick). For S11 operation, and AD log-amp for phase extraction at operating freq. only single source is needed, no IF and another LO.

                      This will not make state of the art performance, but can be almost credit card sized and not too expensive, on board controlled with some low end MCU as interface.

                      (Or at least, will make S11 capable VNA, GPR use not guaranteed)

                      Only thing unclear is, why someone needs to build SFCW GPR in order to get degree in Philosophy??? (Article from previous post)

                      Comment


                      • #26
                        Originally posted by ali_cmi View Post


                        But the problem remains the same, I am looking at a sea of material on GPR and don't know where to begin. I studied some papers, they were helpful but it looks like i'll be designing the circuitry myself as there are no schematics available anywhere.
                        Most useful info will be to describe in short what effort is made and what results are achieved so far using VNA in S11 mode, setup or some published papers about.

                        Time ago, I conceived to try straight high power monopulse UWB radar, but abandoned that idea, just not my league, due to missing T&M equipment required, not exactly DIY hobby project. So I’m bit off GPR stuff now.

                        From description you give, reflection (amplitude and phase) from single port S11 and single antenna is used for signal extraction, and then processed somehow (don’t ask me..)
                        Some basic parameters can be deciphered from setup\instrument type description. Maybe some very cheap ready-made hardware, like “antenna analyzers”, in essence S11 measuring wideband sets, can be used?

                        With more info, I may point you to some particular design\reading\hardware etc.

                        (Song remains the same, only instead of balancing coils at kHz, this is about balancing bridges at GHz…)

                        Comment


                        • #27
                          Originally posted by Tepco View Post
                          Maybe some very cheap ready-made hardware, like “antenna analyzers”, in essence S11 measuring wideband sets, can be used?
                          Aren't they all using some darn slow sweep?

                          Comment


                          • #28
                            I found this page. It looks like the same principle as Ali's school setup.

                            http://www.ngi.no/en/Contentboxes-an...trating-Radar/

                            Comment


                            • #29
                              Originally posted by Davor View Post
                              Aren't they all using some darn slow sweep?
                              Yes, and even worst, only few (more expensive) ones can provide phase measurement, usually just magnitude.

                              Comment


                              • #30
                                Originally posted by Altra View Post
                                I found this page. It looks like the same principle as Ali's school setup.

                                http://www.ngi.no/en/Contentboxes-an...trating-Radar/
                                Yes, it is almost the same.

                                Comment

                                Working...
                                X