helical filter design

im run the helical cavity bandpass filter for very narrow bandwidth filter 0.5MHz, and center frequency is 30 or 90 MHz, I have done lots research on this topic, but few materials can be found, and this technology is being used for very long time, but this type of filter does not have any synthesis software compare to LC filters. anyone can help me? or anyone had done helical filter before?
THANKS IN ADVANCE.

You may find a solution in this book:

thanks vswr

i found your reply in other topic, which introduced two software to synthesis helical filter
http://tonnesoftware.com/helical.html
http://www.comnav-eng.com/dnload.asp

first one is quite good, but the minimum bandwidth cannot be less than 1 MHz, and i use the parameters for this structure, run in HFSS, the result was not good, but the resonant frequency located correctly.

the second one do the transform from helical to LC circuit, which state the helix structure acts inductor shunt with capacitor.

The book you recommend, i cannot find in the local bookstore, its really old one, if you have electrical version, please let me know.

Anyone know how to short the simulation time in HFSS, or how to do helical surf approximation in mesh?

Thanks in advance

Helical filter may not give you the best Q at the frequency you specified. Do you have to use helical filter? Can you consider crystals filter though crystal filter is not for high power applications.

thanks Phytech, cos this type filter needs 100w power connect after power amplifier. so the only solution is used helical cavity structure.

I see! In that case, crystal filters cannot be considered. Generally speaking, you can build any filter including helical filters by the coupling coefficient method. After you obtained the k values either from any commercial software or from g-value. You may find the right dimension and coupling apertures etc by simulations (HFSS for example) or by experimental measurements. In this way you can build any filter you want.

Thanks phytech

I have done the model simulation in hfss, and try to optimize the parameters for this helical structure, but it took long time to run, any suggestion to figure out the mesh setting for helix? and another question, i wanna add capacitor at the open end of the second, third and fourth helical in HFSS, how do i do? 1. draw a sheet vertial to the face of end of open-end helix and set the capacitance? but i can not define the location of the sheet. 2.set these three face as ports and export into ansoft designer, but computer just let me import only two wave ports into designer. any suggestions?

Thanks in advance

I have never simulated helical with HFSS. Actually I would prefer to make it experimentally because even you get an desired dimension from simulation, when you wind the helical coils, it differs from what you assumed in simulation, unlike combline resonators that are well defined. So it would be better to do it experimentally. Also introduce coupling screws to control the coupling strength. As for the capacitor loading in HFSS, I tried before but could not figgure it out. If anyone can share how, I would appreciate it as well.

hi, 3HzT,
I have the same design problem as you. Do you have some progress in the design?

Thanks

Hi All,

I am new here, just registered today.

I actually was helped in this site when I searched info for combline filter design. Someone mentioned the paper by Peter Martin, which incorporates the most important papers. So I am impressed by this site, and I'd like to join.

I have simulated 2-pole helical filter in HFSS for 108-174 MHz and 30-88 MHz, with bandwidth about 2 MHz. The projects were for tunable filters. So trimmer capacitors were added to the open end formed by a metal piece connected to the helix, and PIN diode switching circuit was used to switch capacitor in and out to tune the frequencies. In my cases, the helix length was just about 60-70% of that of a fixed filter for the higher frequency end. It was with relative smaller amount of meshes due to less helix turns.

In HFSS, only fixed passband filters were simulated. This may give good reference to find the proper tapping and aperture for inter-cavity coupling. I found both magnetic coupling and electric coupling works well for my cases. Notice the return loss. I had to change the impedance of the tapping wire by adding coil to reach optimal reflection.

Another book "handbook of filter synthesis" by Zverev contains one chapter (9) about helical cavity and filter design. The constraints of helical cavity design are there, to keep the Q as high as possible.

Freeman

hi, freeman,
how much capacitance at most can be connected to the top end of the helix? Can you simulate the connected capacitor in HFSS?

Thanks
y99lx

Hi y99lx,

The filter cavities were designed for its high frequency end (174 MHz or 88 MHz). In my cases, what I concern is how the performace changes for a tunable filter when increasing capacitance loading. There is a metal plate (cavity hot) connected to helix, which is for mounting trimmer capacitors. I had more than 10 of such capacitors on the plate cotrolled by PIN diode switches. In HFSS, these real capacitor components cannot be modeled. But I intrude another metal piece from the top (cavity ground) so that this piece forms a plate capacitance with the first one. Thru adjusting the gap distance, I got variety of capacitance value.

For 108-174 MHz, I got up to 25 pF from the plate capacitor, and for 30-88 MHz up to 110 pF, which were needed for shfting the filters from the high frequency end (174 MHz and 88 MHz) to the low end (108 MHz and 30 MHz).

What I desired thru doing this was to check the change in filter response and find out (confirm) proper coupling approaches. I think there is no need to simulate four-cavity filter if you are working on such. A two-cavity filter simulation supplies all the major physics characteristics I need, either for a four-pole or a two-pole filter. This may cut down the complexity of the simulation models and save some time.

Hopefully this may help.

Freeman

Thank you, freeman.
I can simulate the cap-loading by assigning a lumped RLC boundary. The problem is that the totally loaded cap is too much to implement in real world. Also, the loading is very sensitive when the loaded cap is small. Did you consider how to reduce such kind of sensitivity of small loaded cap?

Thanks,
y99lx

Hi y99lx,

I haven't used lumped RLC boundary so I don't know the performance and limits. The values of capacitance I got were from plate capacitors formed by drawing (putting) two metal pieces (hot and ground).

Freeman

hi, freeman, thank you any way.

I actually have learned a lot from this topic..

Hi rffilterdesigner, I am fine and buzy on looking for new employment. Where did I meet you?

Freeman

have you ever worked in RF Products Inc, (Camden)?

Many people said it is a junk company. The managner Carmine is stupid. A guy I know just resigned from it.

Hi, who resigned? Just tell me the initial of his name and your initial.