2m Series Band Pass Filter

Well it took a few days longer thanks to some awful weather putting me off the trek to the workshop. But I finally finished the 3 pole series bandpass filter I alluded to in my last post. Here are some pictures of the filter:

What you might immediately notice is there are no trim caps. Since my last post I had removed another turn of wire and I had re-measured the inductors at 119nH. So I used this value in the design and since I'm confident in my technique I didn't feel the need to include trim caps in the first instance.

So how did it go? The response and losses were almost exactly as modelled. The model is shown below:

My centre frequency measured 141.7MHz, between the 3db points. And a peak at 140.5MHz. A few spot checks further down the response curve on higher and lower frequencies showed my filter had a very slight shift towards the left, or to the lower frequencies, compared to the predicted response. And I couldn't noticed any ripple in the response suggesting it wasn't worth trying to trim up the filter. With my loss measured at 1dB I adopted that old adage "If it ain't broke, don't fix it". I'm calling the filter finished.

You might notice I only fitted one RF connector. That's because I plan to mount this back onto the board the coils were removed from. I have a project in mind now that the filter worked so well.

Regards
Richard VK6TT

Footnote: This is the second most viewed page of my blog. While it shows what can be done with high Q inductors the results are not as good as a filter made with quality RG213 cable. While the RG213 filter is physically larger it would be a much better option for a homebrew project should space allow. The filter above had a bandwidth of 25MHz to keep losses around 1db. A 2 pole filter built with RG213 could have a bandwidth of 3MHZ for the same amount of loss.

June and the Joy of the Junk Box

I'm having a break from my WSPR project while I wait for more parts to arrive. But I couldn't go cold turkey so I rummaged through the junk box and one of the many boards I pulled out is shown below. And it got me thinking. What could I use this for? What parts could I re-use? Is it really worth keeping?

The board has some nice helical coils used for filtering so it was probably the UHF variety of Motorola Syntrx radio. These helical coils are presumably high Q and perhaps could be re-used. I measured the dimensions and if the free end was unwound and used as a could they would appear to around 130nH. Some quick number crunching suggests that might be useful for front end filtering at 6m and 2m.

I pulled one out, unwrapped half an turn from the floating end and inserted the resulting coil into my inductance measuring fixture. The measured inductance turned out to be 141nH with a Q of 160. Popped that all into some filter software which showed that at 2m a 3 pole filter would have a loss of about 1.2dB. That's about 0.4dB worse than a 2 pole filter but perhaps the slightly higher attenuation of pagers would be worth the extra loss and complexity.

2m Filter, 2 poles (dashed) versus 3 poles (solid)

The issues at 2m are the inductance makes for very small values of capacitance. Not insurmountable though. Having removed all the helical coils from the board the space available does have me intrigued to give it a go.

Then why I was playing with the Iowa Hills software I noticed a red message: "First Tank Z is Large". I re-ran the design in Elsie and then realised that in practice the message meant the coils were too large.Then I discovered the series option in Iowa Hills. This is not something I had ever seen before in Elsie or AADE. It gave me a design, after tweaking, like the following (shown in comparison to the previous 3 pole topology):

I worked out how I could manually tweak the capacitor values to what I wanted and found that as long as the parts were close to design I wouldn't need trimmers. Since I've never built a filter like this I decided to see how well it worked in practice. Over the next week or so I will build this and report back on the results.

I have a project in mind for the filter and board. Anyone else have thoughts on what it could be used for?

Regards
Richard