This reminds me of why I subscribed in the first place, you run down all the pertinent questions, do the math, test the idea, find the answers and prove it can be done! GLORIOUS!!!
Beam Forming FOR THE WIN! This is exactly how the starlink dish works - instead of moving the dish all day long, they steer the radio frequencies and account for doppler shift all at the same time. Pretty cool to see this at the repeater level.
There is much easier way how to cut cable to electrical length using S11 displayed as smith chart. The same example, frequency 146.7MHz and 60° where you end up with electrical length 0.341m. The property of λ/4 line is that converts short to open and vice versa, and there is frequency on which electrical length of 0.341m would be a λ/4, in fact, its 300/(0.341·4)=220.05MHz. So if you take some coaxial cable with one connector mounted, cut it to something little bit longer than what you expect and then connect to the first port of the VNA. If you sweep from (almost zero) to 220.05MHz, you should see something near to half of circle, because on the low frequency it's just open, so it touches the right horizontal axis of the smith chart. The opposite side should be somewhere near to short (which is left side of the horizontal axis) and then you can easily cut small pieces of that cable until you touch the "short" exactly. This is not a problem, because you have nothing on the opposite side of that cable. As the S11 touches the "short" at 220.05MHz you are sure you have cable of electrical length of 0.341m. In fact, you can use the same method to determine the electrical length of your cable to found the exact velocity factor, which is useful to know before you "pre-cut" your cable. The method you used is perfectly correct, but it requires you to have connectors on both sides of the cable, so you cannot easily fine-tune the cable to the required length. And I would like to mention that I am happy to see someone from ham radio community knowing that repeater requires antenna with maximum at negative elevation.
Great job. This is the method used in broadcast to adjust coverage area . And in US/Canada MW AM stations change pattern from dat/night do to propagation change. And gets fancy. The hours of daylight change throughout the year. So the change in pattern is also time dependent . Also many AM (MW) stations cut TX power at night unless clear channel. When you are all done simply try then go out in the field and test signal . Then make changes as needed. The Earth not perfect . We has a new warehouse go up near a AM (MW) station. We had to adjust the phaser a tad because of reflection off building. 73
That's interesting that they can actively change it! I am going to make a couple of different delay lines up to see which provides us with the best coverage. Do some testing with a spectrum analyzer at various points to measure signal strength.
Absolute fantastic video, just a tip for the VNA Entering the freq in the NanoVNA - save yourself a step, instead of min, max and centre freq, just enter centre freq 146.7M and span 100M Start and stop is really for having a start and stop freq, which is voided when you enter a centre freq that is not the centre fo the range, see marker hard to the right @12:53 now shows 196.7 thus no longer 200M.
Kudos for introducing this concept. You might want to turn on real ground under the array and look again because ground reflection will have a profound effect on the main lobe angle with a low antenna height.
The lobes start to “fan” out, not sure if that’s the right descriptive word, but you know what I mean. And with the added delay, the “main” lobe is pointed still down at the desired angle. The good thing is I’ll make a couple of delay lines to test with.
TV transmission antennas are typically hundreds of wavelengths above ground where ground reflections have little effect, but Hayden's antennas are low and ground reflections will tilt the radiation angle up profoundly.
Complicating things too is that on one side of the mountain there is a sharp drop off - like several hundred metres off the side of the mountain.. and it's solid rock.. so modelling ground is somewhat something of a hard thing too. But in practical terms, it will be interesting to see what happens - if the snow ever melts!
Very good. The barrel inaccuracy could have been avoided by using both blue cables with barrel in center to calibrate. Then the barrel would be accounted for during measurement. Not a big error though.
That was very interesting, will you be doing some recorded signal tests with the phase shift in and out. If possible a few fixed antenna locations showing the before and after results. Thanks for sharing and educating me on the NanoVNA.
Hi mate, Yes I'll be doing that. Using the spectrum analyzer to take measurements before/after in various spots. Also testing using different delay lines. I'd like to get some real time measurements as the delay lines are switched out - but I need a few stations coordinated/using spectrum analyzers for that. Maybe even a TinySA will do. S meters on radios aren't accurate and they vary from radio to radio.
Thanks for that Hayden, a very informative video. What does the radiation pattern look like with real ground switched on as ground reflections will alter it a bit as will anything in the Fresnel Zone . Granted, it is no simple matter to calculate these effects over terrain such as in your situation but it would be interesting non the less. You have got to love these Nano VNAs, they have brought so much that was out of reach of all but a few in recent years. Cheers
Thanks mate. Yes with real ground on you get a different pattern (more pronounced nulls, with lobes that have higher gain). Once you start feeding the bottom dipole out of phase (lagging) then it starts to flatten/pancake a bit and those nulls disappear and the lobes combine (at a lower gain). I should've done that calculation on the model during the video. Be good to see it work in practice.
Actually I used this method to double confirm (not in the video). I was confused at first as to why I was getting 120 degrees - and then realized it was due to the reflection - i.e. round-trip delay.
You didn't show any real test results, I am a bit disappointed with that. Why didn't you show us a reception test at a mobile station and at the base. I would have used a completely different antenna setup. I would have put the antennas nearer the pole to give a cardioid pattern and put the splitter near the aerials. You can then easily adjust the cable for different phasing - the cable to just one aerial. I'm not convinced that these simulators work as it hasn't in your case because it doesn't take a lot of the surrounding or radiation angles of antennas in to account. Using a 3 element beam at a downward angle would be far better than the two dipoles and give more gain and better coverage. Failing that, just a 2 element would widen the pattern. I would have used a J Pole as that will cover the entire area and still have a fairly low angle of radiation. Why make it more complicated and have to use far more power up two bits of coax only to lose power splitting it between two inefficient aerials?
1) I’ll let you come and clear the snow 2) what happens with a 3 element beam? You lose a lot of your signal in any direction other than the way it the beam is pointing. Not all that desirable for repeaters. 3) I suggest you watch the video again. You may then notice/hear the reasonings behind running two antennas that were given. If it were as simple as sticking up a J-Pole, that would’ve been done. In the commercial world, you use high quality antennas that will last. Folded dipoles such as this are built to last in the environment that is presented to them.
This reminds me of why I subscribed in the first place, you run down all the pertinent questions, do the math, test the idea, find the answers and prove it can be done! GLORIOUS!!!
Thanks! The icing on the cake will be the real world tests!
Beam Forming FOR THE WIN! This is exactly how the starlink dish works - instead of moving the dish all day long, they steer the radio frequencies and account for doppler shift all at the same time. Pretty cool to see this at the repeater level.
Starlink is next level with it that's for sure! Thanks TO - science!
Great demonstration! This is exactly what I look for in a video. I use MMANA and have been practicing with my NANOVNA alot lately.
Awesome! MMANA is good fun to play with. I've used it on Yagi antennas too, optimizing them etc
There is much easier way how to cut cable to electrical length using S11 displayed as smith chart. The same example, frequency 146.7MHz and 60° where you end up with electrical length 0.341m. The property of λ/4 line is that converts short to open and vice versa, and there is frequency on which electrical length of 0.341m would be a λ/4, in fact, its 300/(0.341·4)=220.05MHz. So if you take some coaxial cable with one connector mounted, cut it to something little bit longer than what you expect and then connect to the first port of the VNA. If you sweep from (almost zero) to 220.05MHz, you should see something near to half of circle, because on the low frequency it's just open, so it touches the right horizontal axis of the smith chart. The opposite side should be somewhere near to short (which is left side of the horizontal axis) and then you can easily cut small pieces of that cable until you touch the "short" exactly. This is not a problem, because you have nothing on the opposite side of that cable. As the S11 touches the "short" at 220.05MHz you are sure you have cable of electrical length of 0.341m. In fact, you can use the same method to determine the electrical length of your cable to found the exact velocity factor, which is useful to know before you "pre-cut" your cable.
The method you used is perfectly correct, but it requires you to have connectors on both sides of the cable, so you cannot easily fine-tune the cable to the required length.
And I would like to mention that I am happy to see someone from ham radio community knowing that repeater requires antenna with maximum at negative elevation.
Great job. This is the method used in broadcast to adjust coverage area . And in US/Canada MW AM stations change pattern from dat/night do to propagation change. And gets fancy. The hours of daylight change throughout the year. So the change in pattern is also time dependent . Also many AM (MW) stations cut TX power at night unless clear channel. When you are all done simply try then go out in the field and test signal . Then make changes as needed. The Earth not perfect . We has a new warehouse go up near a AM (MW) station. We had to adjust the phaser a tad because of reflection off building. 73
That's interesting that they can actively change it!
I am going to make a couple of different delay lines up to see which provides us with the best coverage. Do some testing with a spectrum analyzer at various points to measure signal strength.
This is a brilliant demonstration of "beam tilt" and phasing requirements. Bravo Hayden! 73 OM
Thanks!
Absolute fantastic video, just a tip for the VNA
Entering the freq in the NanoVNA - save yourself a step, instead of min, max and centre freq, just enter centre freq 146.7M and span 100M
Start and stop is really for having a start and stop freq, which is voided when you enter a centre freq that is not the centre fo the range, see marker hard to the right @12:53 now shows 196.7 thus no longer 200M.
This is awesome stuff. Including the maths that I'm following along with. THIS is the reason I got into amateur radio.
Kudos for introducing this concept. You might want to turn on real ground under the array and look again because ground reflection will have a profound effect on the main lobe angle with a low antenna height.
The lobes start to “fan” out, not sure if that’s the right descriptive word, but you know what I mean.
And with the added delay, the “main” lobe is pointed still down at the desired angle. The good thing is I’ll make a couple of delay lines to test with.
SUPER turtorial. - Thanks! - Cheers!
I see beam tilt used on TV stations
when I have looked into them. 😊
TV transmission antennas are typically hundreds of wavelengths above ground where ground reflections have little effect, but Hayden's antennas are low and ground reflections will tilt the radiation angle up profoundly.
@@michaelross-aj6fgvk3rz Thanks 😊
Complicating things too is that on one side of the mountain there is a sharp drop off - like several hundred metres off the side of the mountain.. and it's solid rock.. so modelling ground is somewhat something of a hard thing too. But in practical terms, it will be interesting to see what happens - if the snow ever melts!
Nice work Hayden!
Thank Ape!
Very good. The barrel inaccuracy could have been avoided by using both blue cables with barrel in center to calibrate. Then the barrel would be accounted for during measurement. Not a big error though.
Yes very true!
Great video and really informative.
Very good explanation. Looking forward to when it's installed, as I have a spot where, assuming my understanding is correct, this will be fixed
That was very interesting, will you be doing some recorded signal tests with the phase shift in and out. If possible a few fixed antenna locations showing the before and after results. Thanks for sharing and educating me on the NanoVNA.
Hi mate,
Yes I'll be doing that. Using the spectrum analyzer to take measurements before/after in various spots. Also testing using different delay lines.
I'd like to get some real time measurements as the delay lines are switched out - but I need a few stations coordinated/using spectrum analyzers for that. Maybe even a TinySA will do. S meters on radios aren't accurate and they vary from radio to radio.
Great video! Thanks!
Glad you liked it! Thanks!
Awesome info Hayden. Thanks.
Thanks for that Hayden, a very informative video. What does the radiation pattern look like with real ground switched on as ground reflections will alter it a bit as will anything in the Fresnel Zone . Granted, it is no simple matter to calculate these effects over terrain such as in your situation but it would be interesting non the less. You have got to love these Nano VNAs, they have brought so much that was out of reach of all but a few in recent years. Cheers
Thanks mate.
Yes with real ground on you get a different pattern (more pronounced nulls, with lobes that have higher gain). Once you start feeding the bottom dipole out of phase (lagging) then it starts to flatten/pancake a bit and those nulls disappear and the lobes combine (at a lower gain). I should've done that calculation on the model during the video. Be good to see it work in practice.
Great content thanks
Interesting thank you 🙏
Nicely done! 73
Hayden, instead of flipping the model in MMANA, why wouldn't you set your antenna elevation to be the same or similar to real world?
What is the new down tilt with the minor change in coax length from 60 degree to 62.5 degree?
According to modelling it's an additional -0.4 degrees difference
Nice, very interesting and clever hack of the software. :)
Will you do a real world measurement before and after down in the valley ?
Yes. Plan to use a spectrum analyzer to measure the before and after signal levels in several spots
Way too much math. That was my antenna. The people down below would be out of luck
NanoVNA can measure round-trip delay of the delay line, with a short or open at the far end of S11 phase measurement.
Actually I used this method to double confirm (not in the video). I was confused at first as to why I was getting 120 degrees - and then realized it was due to the reflection - i.e. round-trip delay.
sehr interessant, welchen vertikalen abstand haben ihre 2 antennen. viele grüße aus deutschland.
3/4 of a wavelength. About 1.5m
@@HamRadioDX Danke für ihre info, 73
Do you mean mana-gal?
Yes
You didn't show any real test results, I am a bit disappointed with that. Why didn't you show us a reception test at a mobile station and at the base.
I would have used a completely different antenna setup. I would have put the antennas nearer the pole to give a cardioid pattern and put the splitter near the aerials. You can then easily adjust the cable for different phasing - the cable to just one aerial. I'm not convinced that these simulators work as it hasn't in your case because it doesn't take a lot of the surrounding or radiation angles of antennas in to account.
Using a 3 element beam at a downward angle would be far better than the two dipoles and give more gain and better coverage.
Failing that, just a 2 element would widen the pattern.
I would have used a J Pole as that will cover the entire area and still have a fairly low angle of radiation.
Why make it more complicated and have to use far more power up two bits of coax only to lose power splitting it between two inefficient aerials?
1) I’ll let you come and clear the snow
2) what happens with a 3 element beam? You lose a lot of your signal in any direction other than the way it the beam is pointing. Not all that desirable for repeaters.
3) I suggest you watch the video again. You may then notice/hear the reasonings behind running two antennas that were given. If it were as simple as sticking up a J-Pole, that would’ve been done.
In the commercial world, you use high quality antennas that will last. Folded dipoles such as this are built to last in the environment that is presented to them.
My wife when younger had a NANA. 😊
You need some down tilt. Oops spoke too soon. I should have waited to comment. Good on you. You got it handled. 73, N1FQF
😂
Cant do show and tell this wednesday maybe zombie wednesday?