@pitot1988 My dsp prof is literally teaching from a 177-page set of notes scribbled on his ipad from his golden-year of the pandemic. He just scrolls through it with a zoom of 3 lines and ask "are you following?" every 5 minutes, followed by an awkward pause. Sometimes he says for us to just guess at his questions and that it's okay if we don't know, but if we do and don't answer correctly, he will just repeat exactly what no one understood the same way and move on, or he will give us a pop quiz on it. I once asked what section of the book correspond to his current lecture because his lecture notes were too messy for me to understand his proofs; he got upset and said he doesn't use the book and to just read his notes again. It's a 4-hour course at $600/hr/student. This isn't normal, is it?
Excellent lesson. Thank you. I learn 😊 ! In the 2.3 GHz version You could place the antennas on a turn surface; i.e. cake serving kitchen device and the receiving antenna is stationary. But only if Your wife allows (promise to use a protecting cover between antennas and the turn table.
After a ~year, 3.7% Likes of 9200 views. Or each Liker watched the video 27 times - since I don't know the number of dislikes, I have no covtrary evidense! (Other than the kind of videos that do "go viral", not so much 27x likelyhood) Since I'm neither physicist nor mathematician but aspire to implement phased receivers, I really appreciate your videos. Thank you.
EDIT: I should learn to watch the whole video before asking questions. ;-P Just found your channel, and this is super helpful to me as a beginner/amateur. Thank you for posting these. I'm curious, when you say that time delay can be helpful for wideband signals, is that in absolute terms or relative to carrier frequency? This will be useful for determining which direction I go down first. For context… I'd like to apply beam forming to experiment in the 70cm band. Frequencies of interest would be between 430 and 440 MHz, and signals would typically be
Ha, yes! You found the equation in the video relating instantaneous bandwidth and carrier freq. So 10kHz at 430 MHz will be no problem -- that is very narrow band. But it sounds you've got a very cool project in mind! Please do a video or blog on it and let me know the link/channel!
Very cool video! I wonder instead of time delay filter, you could take the FFT, shift the phase of each bin with a linear taper (centered about baseband), then IFFT? That time delay filter you used had sinc function taps, which would be flat window in the frequency domain which does not alter the phase, so maybe the FFT method would combine the time and phase shift. I suppose the filter is probably faster than doing an FFT, at least until the filter gets really long.
Jon, loving this pluto series! I was wondering, you've talked about the problems with spacing antennas WIDER than wavelength/2 but are there any cons to spacing antennas less than wavelength/2?
Yes, that's a great question! I can't think of any major downsides. Except for the most major downside of all: cost! A smaller antenna spacing means you'll need more elements to cover a desired aperture size. And so your array will have more cost, weight, power, complexity, etc. Maybe some others can chime in on any downsides? Or significant advantages to smaller spacing?
@@jonkraft Doh. I remember now that aperture is the physical space you’re trying to cover. I’ll look into tradeoffs of aperture size. Eg, how does a 4 antenna array covering 2 wavelengths compare to a 4 ant array covering 1 wavelengths. Thanks again!
@@stingy49 Yes, and you can actually try it out with this demo just by changing the spacing of the Rx antennas. You'll see grating lobes start to come in, but you'll also see that the main lobe gets tighter. It's pretty instructive to see it in action. You can also check out the grating lobe video here: wiki.analog.com/resources/eval/user-guides/circuits-from-the-lab/cn0566#labs_and_lectures
Thank you for explaining much of the math, I still don't understand 100% but your video really helped. I've had fantasies of building something similar to a parametric array, as in the "Ultrasonic Directive Speaker" But using small software-defined microwave transmitters. One goal, just for example , could be to send a regular FM transmission to a specific radio using frequencies not in the FM band.. How many transmitter antenna combinations do you think such a project would require? Do you think the software would be able to calculate the timing and phase differences quick enough for an array large enough for such a goal? I need to know, so I can figure how much of a budget I would need to even attempt such a project.
Thanks! You could check out the first video in this series where I give some equations for beamwidth--i.e. how well you'd be able to target one radio vs another. But a big issue with FM radio freq's could be the size of the array -- 100MHz means your spacing for each antenna in the array would be 1.5m.... But I'd encourage you to start small, build a simple example, and then play around with that for a bit.
Jon. The antennas operate at Carrier freq, so with this antenna setup, the Carrier phase-timing between antenna must be adjusted by a phase-altering component(preferably digital) placed between the mixers & Center-Fed by the Carrier signal from the Rx-LO.
Hello Jon, awesome video as always ! I am planing on purchasing an SDR for myself. Now I would like to understand if transmit beamforming would work in a similar way. For example I would like to have the standalone antenna fixed and steer my beam until the received power on the single antenna reaches a maximum. Could you confirm that for me ? Best, Max
Right, the array is reciprocal, so forming the Tx beam will be the same as the Rx beam. I like to use the Rx beam, because then we can see it. But for your use case, it'll be no problem to modify the code for transmit. Let me know how it works!
So, the time delay seems to be a better solution as it provides phase shifts for all frequency component contained in the signal? So I don't fully get why people use phase-shifting instead of true time delays then. Am I missing something?
There are some cases, like with polarized signals, where you might prefer phase shifting. But yes, in general time delays are what you're after. However.... it's a LOT easier to do phase shifting than time delays. In an all digital array, it doesn't really matter too much (though it is a more computations to do a time shift). But most arrays are not all digital. And phase shifting is almost always easier to implement in a hybrid subarray or an all analog phased array.
BladeRF is a really great SDR, but I don't know how to set it up. If you can ask them for help on how to use with Python/Matlab, then all the other programming should be similar.
I once had to (curiosity) reverse engineer a TVRO 2 axis parabolic antenna. It had an arrangement of dispoles around the LNB waveguide. By switching these dispoles on and off (i suppose grounding them) the effect was a useful distortion of the receive beam. Enough to sample if the RSS increased or decreased. What I never really understood, as the KU-band signal i linearly polarised could work in both E and M fields?
Hello, awesome video ! Can you help me pls, i have problem with adi librarry, when i run code i have error, i cant install adi, thank you for help! if '/' in name or '\\' in name: TypeError: argument of type 'NoneType' is not iterable
In order to help, I would need a lot more info and screenshots of what you are doing. Can you post that information and all the details here: ez.analog.com/adieducation/university-program/
If only college professors can explain things that clearly and concisely, more students will actually learn more.
@pitot1988
My dsp prof is literally teaching from a 177-page set of notes scribbled on his ipad from his golden-year of the pandemic. He just scrolls through it with a zoom of 3 lines and ask "are you following?" every 5 minutes, followed by an awkward pause. Sometimes he says for us to just guess at his questions and that it's okay if we don't know, but if we do and don't answer correctly, he will just repeat exactly what no one understood the same way and move on, or he will give us a pop quiz on it. I once asked what section of the book correspond to his current lecture because his lecture notes were too messy for me to understand his proofs; he got upset and said he doesn't use the book and to just read his notes again. It's a 4-hour course at $600/hr/student.
This isn't normal, is it?
I praise your dedication. I don't understand a lot of things. But you are doing an incredible work.
Thanks, appreciate that!
Excellent lesson. Thank you. I learn 😊 ! In the 2.3 GHz version You could place the antennas on a turn surface; i.e. cake serving kitchen device and the receiving antenna is stationary. But only if Your wife allows (promise to use a protecting cover between antennas and the turn table.
After a ~year, 3.7% Likes of 9200 views. Or each Liker watched the video 27 times - since I don't know the number of dislikes, I have no covtrary evidense! (Other than the kind of videos that do "go viral", not so much 27x likelyhood) Since I'm neither physicist nor mathematician but aspire to implement phased receivers, I really appreciate your videos. Thank you.
can't wait to try this out!
Thanks Kurt! Yes, please play around with this and let me know any feedback/corrections/ideas, etc. Thanks!
Very very interesting and good explained! keep this way!
Gracias muy interesante!!! Saludos desde Argentina!
¡Gracias Sebastián!
@@jonkraft wait, Jon speaks Spanish as well? No lo puedo creer!
@@INSOFTUSA ¡Ha, no! ¡Pero Jon sí sabe cómo usar el traductor de Google!
Brilliant and clever! Than you a lot, for educating community.
Excellent Video. Thank you!!!!
EDIT: I should learn to watch the whole video before asking questions. ;-P
Just found your channel, and this is super helpful to me as a beginner/amateur. Thank you for posting these.
I'm curious, when you say that time delay can be helpful for wideband signals, is that in absolute terms or relative to carrier frequency? This will be useful for determining which direction I go down first.
For context…
I'd like to apply beam forming to experiment in the 70cm band. Frequencies of interest would be between 430 and 440 MHz, and signals would typically be
Ha, yes! You found the equation in the video relating instantaneous bandwidth and carrier freq. So 10kHz at 430 MHz will be no problem -- that is very narrow band. But it sounds you've got a very cool project in mind! Please do a video or blog on it and let me know the link/channel!
Very cool video! I wonder instead of time delay filter, you could take the FFT, shift the phase of each bin with a linear taper (centered about baseband), then IFFT?
That time delay filter you used had sinc function taps, which would be flat window in the frequency domain which does not alter the phase, so maybe the FFT method would combine the time and phase shift. I suppose the filter is probably faster than doing an FFT, at least until the filter gets really long.
Jon, loving this pluto series! I was wondering, you've talked about the problems with spacing antennas WIDER than wavelength/2 but are there any cons to spacing antennas less than wavelength/2?
Yes, that's a great question! I can't think of any major downsides. Except for the most major downside of all: cost! A smaller antenna spacing means you'll need more elements to cover a desired aperture size. And so your array will have more cost, weight, power, complexity, etc. Maybe some others can chime in on any downsides? Or significant advantages to smaller spacing?
@@jonkraft ahhh so I would need more elements to achieve the same half power beamwidth? Got it. I’ll look into that more. Thank you!
@@jonkraft Doh. I remember now that aperture is the physical space you’re trying to cover. I’ll look into tradeoffs of aperture size. Eg, how does a 4 antenna array covering 2 wavelengths compare to a 4 ant array covering 1 wavelengths. Thanks again!
@@stingy49 Yes, and you can actually try it out with this demo just by changing the spacing of the Rx antennas. You'll see grating lobes start to come in, but you'll also see that the main lobe gets tighter. It's pretty instructive to see it in action. You can also check out the grating lobe video here: wiki.analog.com/resources/eval/user-guides/circuits-from-the-lab/cn0566#labs_and_lectures
Thank you for explaining much of the math, I still don't understand 100% but your video really helped.
I've had fantasies of building something similar to a parametric array, as in the "Ultrasonic Directive Speaker"
But using small software-defined microwave transmitters. One goal, just for example , could be to send a regular FM transmission to a specific radio using frequencies not in the FM band..
How many transmitter antenna combinations do you think such a project would require? Do you think the software would be able to calculate the timing and phase differences quick enough for an array large enough for such a goal? I need to know, so I can figure how much of a budget I would need to even attempt such a project.
Thanks! You could check out the first video in this series where I give some equations for beamwidth--i.e. how well you'd be able to target one radio vs another. But a big issue with FM radio freq's could be the size of the array -- 100MHz means your spacing for each antenna in the array would be 1.5m.... But I'd encourage you to start small, build a simple example, and then play around with that for a bit.
Jon.
The antennas operate at Carrier freq, so with this antenna setup, the Carrier phase-timing between antenna must be adjusted by a phase-altering component(preferably digital) placed between the mixers & Center-Fed by the Carrier signal from the Rx-LO.
Super
I see that pluto doesnt have the module with the Rx1 and Rx2 anymore, do you know where to pick that one up? Your link is to revC
I'm not sure what you mean. The currently shipping version is Rev D, so that will work. Anything other than Rev A or B.
Spitting image of the Microwave Landing System we developed late 1980s! (MicroNav International. I did the MilSpec tech_docs.)
^5
We used phasing to shift a beam back and forth / up and down, which the plane would use to get precise location in space.
Hello Jon, awesome video as always ! I am planing on purchasing an SDR for myself.
Now I would like to understand if transmit beamforming would work in a similar way. For example I would like to have the standalone antenna fixed and steer my beam until the received power on the single antenna reaches a maximum. Could you confirm that for me ? Best, Max
Right, the array is reciprocal, so forming the Tx beam will be the same as the Rx beam. I like to use the Rx beam, because then we can see it. But for your use case, it'll be no problem to modify the code for transmit. Let me know how it works!
@@jonkraft Thanks for your quick response ! It is good to know ! I am planning on doing NF beam forming with it ! Thanks again
So, the time delay seems to be a better solution as it provides phase shifts for all frequency component contained in the signal? So I don't fully get why people use phase-shifting instead of true time delays then. Am I missing something?
There are some cases, like with polarized signals, where you might prefer phase shifting. But yes, in general time delays are what you're after. However.... it's a LOT easier to do phase shifting than time delays. In an all digital array, it doesn't really matter too much (though it is a more computations to do a time shift). But most arrays are not all digital. And phase shifting is almost always easier to implement in a hybrid subarray or an all analog phased array.
very happy to find your channel, very good video, thank u for your contribution
Thank you!
Hey John, loving your work ive always been afraid to dabble with rf. You might have me convinced
@JohnKraft ....can this setup work for bladeRF micro xa4. I dont have a pluto sdr
BladeRF is a really great SDR, but I don't know how to set it up. If you can ask them for help on how to use with Python/Matlab, then all the other programming should be similar.
Спасибо, очень интересно.
Finally great spotless clear to understand.
I once had to (curiosity) reverse engineer a TVRO 2 axis parabolic antenna. It had an arrangement of dispoles around the LNB waveguide. By switching these dispoles on and off (i suppose grounding them) the effect was a useful distortion of the receive beam. Enough to sample if the RSS increased or decreased. What I never really understood, as the KU-band signal i linearly polarised could work in both E and M fields?
Hello, awesome video ! Can you help me pls, i have problem with adi librarry, when i run code i have error, i cant install adi, thank you for help!
if '/' in name or '\\' in name:
TypeError: argument of type 'NoneType' is not iterable
In order to help, I would need a lot more info and screenshots of what you are doing. Can you post that information and all the details here: ez.analog.com/adieducation/university-program/
@@jonkraft Thank you for your reply, where i can send information to you? Maybe e-mail? or WhatsApp ? Thank you!
@@moskovenko Just post your question in that link above. Someone more qualified than I will be able to help you.