Building and tuning a narrow band 2GHz cavity filter

[erikka]

During my 2 GHz SA experiments it became more obvious that having three IF's (2.5GHz, 110MHz and 10.7MHz) was leading to all kind of problems so I went looking for a very narrow band cavity filter. To be able to go from above 2GHz directly to 10MHz the 20MHz offset suppression  should be more then 90dB. I could not find a filter with these specs.
According to this calculator
https://www.changpuak.ch/electronics/interdigital_bandpass_filter_designer.php
it should be possible to build a 5 resonator interdigital filter at 2 GHz with acceptable loss and narrow enough so I decided to build it. As I have no access to a machine shop I followed the construction proposed here
http://lea.hamradio.si/~s53mv/cavity/cavity.html
and using the square tube and the input/output antenna's from the later link and I build the calculated filter from the first link.
Result is a 5 resonator interdigital filter in a square aluminium tube with solid copper resonators as can be seen in this picture

You can just see the input antenna and two copper resonators with their tuning screws
The only special tool I used was a 4mm wire tap for attaching the resonators and the tunng screws.
After building the real problem start. A filter of this order will not let anything through if not tuned correctly. I could not find how other people did this initial tuning so I created a field sensor as shown in this picture

It was small enough to be inserted from one side of the filter past all the untuned resonators till the resonator up for tuning so I could tune one resonator at a time.
The end result was a 2MHz 3db bandwith at 2.016Ghz and sufficient suppression at 20MHz offset to ommit my 110MHz IF.
This led to an updated SA with this block diagram

Short specs; 0-2GHz input, 0-30dB attenuator IP3 at +17dB with 0dB attenuator. Noise floor around -100dB with 300kHz bandwidth, 300/30kHz HW resolution filters and FFT resolution filters down to 1Hz. Almost no spurs.
You can see the full set of components in this picture

As the ADF4351 have two outputs I added a 3GHz bridge and a triple receiver so the HW doubles as a 35MHz till 3GHz VNA
The unmarked module at the left bottom is a third ADF4351 that can be used is mixed (0-2GHz) or direct (35MHz-2GHz) tracking generator
I hope this inspires more people to build their own GHz cavity filters

[sp9bsl]

After building the real problem start. A filter of this order will not let anything through if not tuned correctly. I could not find how other people did this initial tuning

Hi Erikka,
I used my version of NWT6000 to perform such measurement. Some pictures of my first experimental cavity filter:

[erikka]

Nicely done, where you able to measure the higher harmonics modes of the comb filter design you used? Would be difficult to measure at over 4GHz I guess
As I was uncertain about the impact I choose for a interdigital filter setup.

How did you do the initial tuning?

[sp9bsl]

NWT6000 goes up to 6GHz, so I can measure directly only 2nd harmonic response. If you need more then you need the up/down converter using proper microwave mixer and BPF/LPF, for example a YIG based. Sometimes possible to find used on Ebay. Not easy to set up as the whole microwave matter...  Matjaz (S53MV) uses just microstrip low pass filter after his cavity solution to supress unwanted responses.
Anyway, I'm grateful for his work and showing people that there is no magic in this area of RF and everyone can try

The initial tuning: just rotated the screws and by trial and error found their right positions. There is possibility to find resonance of each resonator with S11 bridge (using bridge like the one you have on the picture).

[erikka]

I did it a bit different.
Connected the VNA in S21 mode.
Connected port 1 to one side of the filter and the field sensor on the end of a long coax to port 2
Or you can use a RF power meter (connected to the filed sensor) and a signal generator (connected to the input of the filter)
Now you can (just) slide the field sensor underneath the resonators from the other end of the filter till you are close to the second resonator.
Then tune the 1st resonator.
Then slide close to the 3rd resonator and tune the second (and retune the first)
End so forth.
Went very quickly

Can you measure if there is indeed second order leakage?
I should be able to do the same.

[erikka]

To complete the building I measured the S21 of the cavity filter

First the pass band



As the home build GHZ VNA does not have any shielding the dynamic range is rather limited but the -3dB bandwith is about 4MHz and the loss is -3dB. From other measurements its clear the suppression of the image at 21.4MHz offset is good enough.

Certain cavity filters do have harmonic modes that will allow 2nd or higher order harmonics to pass.

I can only measure S21 till 4.3GHz so here is a wide sweep.



The peak at 4.3GHz is an artifact of my SW, its not there when you zoom in. So the choice of an interdigital filter instead of a comb filter enabled the suppression of harmonics modes was good

The S11 measurement shows there is still room for improvement



but for now I'm happy.

[sp9bsl]

Erikka,
I've built my analyzer at home and of course have the shielding in the right places. Beside the metal shielding itself there is a must for ESD foam as the microwave absorber at these frequencies (S53MV mention that). It is unbelievable but it supress the I/O leak by over 10dB! I've build this thing as a tool for building tracking generator for my SA (HP 8591E) and I found it very helpful - would say essential.

My analyzer inside:


Back to the topic, the wideband response (the sub 2.4G spikes are present because there is lack of any filter preceeding the first mixer):


The noise floor of my analyzer with terminated input (the steps on the level around 3, 3.5 and 4GHz are caused by change in output amplitude of LO to compensate mixer loss):


S21:


S11:
and there is a problem - I saw the things I didn't want to see My filter has five peaks, but overall S11 is at the level of -8dB - IMHO too poor. This is probably due to I/O coupling - probes too short.
I was wrong saying that you can tune the filter by looking for resonances of resonators, I saw this elsewhere but this was for one resonator cavity.... This would be true if we could have separate cavities connected in series (Scotty Sprowls analyzer has this arrangement of cavity filter). In the case of our filters - five resonators in one cavity, we have endless interaction between resonators tuning... At least I see that. I think (may be wrong of course) your way is not good too. If you wanted to build a filter based on Matjaz Vidmar concept, you must have trapezoidal shape response (flat top) - not peak as you have... He explained this in "Practical cavity filters" article on his website.

Anyway, a lot of experiments and measurements ahead of us

[erikka]

The antenna designed by S53MV are too close to the first resonators. I had to bend them outwards to get to the right impedance.
You suggest all resonators can see all other resonators but this is not true otherwise an in between out of tune resonator could not have such a big impact. And S53MV mentions the fields decay exponentially so it's really coupling between successive resonators.
Tuning one at a time worked very well if the field sensor is at least as far away as the next resonator.
Given the very narrow bandwidth I am not sure a flat top filter is possible but when measured in situ it shows sufficient flatness to place the 300kHz BW of the widest resolution filter on a fairly flat part.
I found that by placing a attenuator in the path of the reference signal of the VNA I can increase the S21 dynamic range to over 60dB. Next step is shielding and foam.
Thanks for the input!

[erikka]

After some very careful tuning the flat top did appear.

At least flat within 0.5dB
Center frequency at 2020.7MHz, S21 -5.1dB
-3 dB width 3.3MHz
-40dB width 14MHz

[coppercone2]

the real filters (I sawed a big ass one apart out of curiosities sake) have little disks attached to rods in them. at least the bandpass ones

[erikka]

I think you are referring to filters at much lower frequencies that use disks at the end of the resonator to create a tunable capacitor.
The capacitance required to tune the 2GHz resonators is so small there is no need for a disk. Just the proximity of a screw is enough capacitance.

[coppercone2]

no, its a 2GHz filter

this one is high order though, I don't know if that changes things

[KE5FX]

Looks good!  Note that for this frequency range, pipe-cap filters are a whole lot easier to build and align, though.  Possibly worth trying next time.

Didier used a nut at the end of the tuning screw, but that's not usually necessary at the higher frequencies.  Overall this is a good technique for ad-hoc filter construction in the 2-10 GHz range.

[coppercone2]

when does surface finish on components start  becoming an issue ?

Say you have the time to put the bells and whistles on it, like if you have threaded rods that you know approximately where they will end up, you turn most of the thread away besides the ones needed for the tuning range. I.e. you determine the depth that you need on a VNA measuring it with a caliper afterwards, you figure out where it is, and you leave yourself a bit of room, then machine most of the thread away besides what you think you need (with some kind of engineering tolerance to account for the difference the lack of threads makes). Then maybe you silver plate the shaft.

What kind of a difference does this stuff make?

[erikka]

No clue. Even without silver coating the insertion loss of a 3MHz wide 5 resonator filter at 2GHz is only 5 dB
I used stock copper rod without any finishing improvement.

The components I used where
- 30cm aluminium tube
- 20cm copper rod to cut the resonators
- 2 SMA female connectors
- some 2mm copper wire for the antenna
- 4mm screws for rod mounting and tuning

The only mechanical tools I used where:
- metal saw
- regular soldering iron (for mounting the antenna on the SMA connectors)
- something to grind the length of the copper rods to correct length
- 4 mm screw tap to enable mounting the rods and for the tuning screws in the aluminium tube