I found that 2 things are quite difficult to get right - the exact adjustments of each resonator - to get the desired response shape; and the other is the coupling - you can get a better or worse result (higher or lower pass band attenuation) by changing the exact amount of coupling into the first and last resonator. So with the devices I've built, the coupling is adjusted by changing the exact tap point.
@@FesZElectronics Tuning resonators accurately is best done in time domain. Set the analyser by CF and Span. CF goes to the centre frequency of the filter. The span should be about 3 X the filter bandwidth. Now look at S11 in time domain and you will see three dips, one for each resonator. Tune them by making them as deep as possible. The resonators are now perfectly in tune, and all that remains is the coupling.
Nice explanation, with very approachable builds, thanks! Now I'm going to have to play with these even though I don't have any real application at the moment. (When do we get a diplexer video? 😅)
Very nice work, I'm sure my construction would look more like VK3YE's video on cavity filters, but using the blank PCB's seems like a good approach for this type of thing.
I used generic PLA, nothing special; but I guess there are better materials, anyway, you should be able to reduce losses by having a very low infill ratio - use the bare minimum plastic to support the structure.
I'v found that silver plating the copper helps reduce losses. Also in your helical filters, the dielectric losses in your coil forms could be significant.
I did think of trying to silver-plate one of these devices, but looking into the chemicals needed, I decided it was cheaper and safer not to try... maybe if you know a small workshop or company that can do it for you (I guess a jeweler) it might be relatively affordable. Did you ever silver-plate any electronic device?
Thank you. I was wondering about the thermal expansion/contraction effect from home-brew PCB material filter cans. They wouldn't really follow the expansion/contraction coefficient for pure copper. I suppose I will need to do thermal frequency shift experiments with the materials of interest. But even used commercially made cans are so expensive it might well be worth it for projects less important than a ham club repeater up on a mountain. Did you print your coil forms with some sort of polystyrene to keep the dielectric losses down, or was that not a concern?
It is quite common to make PCB based filters - most transmission line filters are built on a PCB substrate when going for high frequencies; here you can work with FR4 when losses are not important or at relatively low frequnecy, or with some more exotic materials - like Rogers PCB when losses need to be minimized. Regarding the coil forms, I just used PLA, I was not trying to take too much care with losses though...
My HAMster wheel brain is trying to make me figure out how this would scale at 160 meters. #10 can? Steel 5 gallon paint bucket? 40 lb. propane canister? Of course, then i would need a legal-limit amp... (The dogs says I'm crazy, and to go to bed.)🙂
A quarter wavelength at 160m is 40m.. I think a tank container is appropriate. It would be completely impractical, but what a sight to see it would be though.
![[SFM] Risky Recital](http://i.ytimg.com/vi/RzNkY1_Nk3o/mqdefault.jpg)
I work in electronics for 15 years know. Your videos are a true gifts. Many thanks
Thank you very much for sharing and explaining your courageous RF experiments results. 👍👍👍
Nice work. They all need adjustment to the coupling to flatten the tops, but they work, and they work well.
I found that 2 things are quite difficult to get right - the exact adjustments of each resonator - to get the desired response shape; and the other is the coupling - you can get a better or worse result (higher or lower pass band attenuation) by changing the exact amount of coupling into the first and last resonator. So with the devices I've built, the coupling is adjusted by changing the exact tap point.
@@FesZElectronics Exactly what I would do
@@FesZElectronics Tuning resonators accurately is best done in time domain. Set the analyser by CF and Span. CF goes to the centre frequency of the filter. The span should be about 3 X the filter bandwidth. Now look at S11 in time domain and you will see three dips, one for each resonator. Tune them by making them as deep as possible. The resonators are now perfectly in tune, and all that remains is the coupling.
@donepearce "look at S11 in time domain": Do you mean a Fourier transform of S11?
@@tomyorados974 Time domain is a selectable option on this analyser.
The best video on practical cavity filters on RUclips. Excellent work! ❤
Great video, as always! 🙂
Thank you! I'm happy you enjoyed it!
Good work. It would be interesting if you tunned the cavity screw live and show the change or sensitivity.
Nice explanation, with very approachable builds, thanks! Now I'm going to have to play with these even though I don't have any real application at the moment. (When do we get a diplexer video? 😅)
Very nice work, I'm sure my construction would look more like VK3YE's video on cavity filters, but using the blank PCB's seems like a good approach for this type of thing.
5:00 not only a cylinder will be better but also using a pipe will save you a lot of time in construction. Great Job anyways
Again, it's a thought-provoking video. Respect!
Do you know the dielectric losses of the material of the helical coil supports?
I used generic PLA, nothing special; but I guess there are better materials, anyway, you should be able to reduce losses by having a very low infill ratio - use the bare minimum plastic to support the structure.
I'v found that silver plating the copper helps reduce losses. Also in your helical filters, the dielectric losses in your coil forms could be significant.
I did think of trying to silver-plate one of these devices, but looking into the chemicals needed, I decided it was cheaper and safer not to try... maybe if you know a small workshop or company that can do it for you (I guess a jeweler) it might be relatively affordable. Did you ever silver-plate any electronic device?
Muy buen video. Gracias por la calidad del contenido. Saludos.
Nice video!👍
Thank you.
I was wondering about the thermal expansion/contraction effect from home-brew PCB material filter cans. They wouldn't really follow the expansion/contraction coefficient for pure copper. I suppose I will need to do thermal frequency shift experiments with the materials of interest. But even used commercially made cans are so expensive it might well be worth it for projects less important than a ham club repeater up on a mountain.
Did you print your coil forms with some sort of polystyrene to keep the dielectric losses down, or was that not a concern?
It is quite common to make PCB based filters - most transmission line filters are built on a PCB substrate when going for high frequencies; here you can work with FR4 when losses are not important or at relatively low frequnecy, or with some more exotic materials - like Rogers PCB when losses need to be minimized. Regarding the coil forms, I just used PLA, I was not trying to take too much care with losses though...
👍👍
Use of copper sheets and scissor is much easier to build. Copper sheet is available in any hardware store.
9:03 - ...and using... What? :)))
How much dielectric loss does the 3D printed coil form add to an otherwise identical 2M helical filter?
Aren't these the principles Tesla was trying to use to transmit power wirelessly with the wardenclyff cavity resonator?
Your callsign please?
My HAMster wheel brain is trying to make me figure out how this would scale at 160 meters. #10 can? Steel 5 gallon paint bucket? 40 lb. propane canister? Of course, then i would need a legal-limit amp... (The dogs says I'm crazy, and to go to bed.)🙂
A quarter wavelength at 160m is 40m.. I think a tank container is appropriate. It would be completely impractical, but what a sight to see it would be though.
Id rather have a gravity filter. My house is full of it, and nothing i've tried works.