2M Bandpass Filter From RG142 Coax

The statistics for this blog show that a 2m bandpass filter is a popular project. So at a recent hamfest I was surprised no-one was really interested in some ex-commercial circulators I had on my bench. Two male N connectors on quality RG142 coax, a chassis mount N connector and a 50 ohm termination. What's not to like?

One of the activities taking place at the NCRG Hamfest was a measurement of handheld spurious emissions. Which triggered a thought - could a 2m transmitting filter be made from the parts taken from one of these circulators? It would clean up the output and a bandpass response can help reduce the interference on reception all modern radios seem to suffer from.

This post just concerns a filter made with RG142. Later attempts with other coax will be blogged separately.

Circuit

 

Ingredients

• two 19cm lengths of RG213 coax
• two 25pf transmitting variable capacitors (salvaged from a VHF transmitter)
• two 3.9pF high quality capacitors (I had some recovered ATC chip capacitors)
• some enameled wire for a gimmick capacitor (or a 0.7pF chip capacitor)
  
• an enclosure

Alignment

I used my nanoVNA and Dishal's method but whatever method works for you.

Results

The first attempt inside the recycled case of the circulator. I wasn't happy with the insertion loss of over 4dB but the filter tuned up easily and had a 5MHz bandwidth. Still, a 4dB insertion loss for such a narrow 2m filter compares well with a filter made with conventional coils.

 

 

 

 

 

 

 The second attempt started with a piece of scrap PCB as shown below.

Once I fitted some stubs, caps and connectors it looked like this:

It still had a loss of 4dB but the bandwidth was around 5MHz.

 

m

Some further iterations and use of the highest quality parts I had got the loss down to just under 3dB. 

With the bandwidth extended to around 9MHz the loss fell to 2.3dB. 

 

So after several attempts the short answer is not really. The biggest issue was insertion loss. While I could get an insertion loss of 3dB or lower I felt that was too high. Some back of the envelope calculations suggest this is still less loss than a conventional direct coupled bandpass filter using inductors. However, I did achieve the desired bandwidth of 5MHz and significant harmonic suppression.

The coax needs to be larger in diameter. I found a text book which while poorly written made the point that unloaded Q of coaxial stubs is a function of the coax diameter. I assume that is the reason why a cavity filter is large in diameter!

 However, you can build a better bandpass filter but not with RG142. More on that later.

 

 

 

Footnote: 

You might benefit from one of these when stripping the RG142 

I got mine for next to nothing on Temu.  Click on the picture to go there and search for "Wire Stripper Knife" with no quotes. It has been suggested I earn something if you go there and if that's the case I will discontinue with ads on my blog.