A friend has asked me for help re-tuning a 4-section helical bandpass filter for the 2-meter ham band. It has no adjustments, so there won't be too many chances to recover from botched cutting.
As I recall, there's considerable interaction between the sections. And since he's been bending on the resonators, the bandpass is quite out of whack. He wants to move it up 2 MHz. and restore the shape.
A couple of decades ago, I had a procedure that involved measuring return loss. You disable (short out) all but the first resonator and set the dip of the curve here. Then enable the second section and put that dip there. So on till you get to the end.
AS I recall, it got you very close in one pass. The second dip moved the first one, but the end result bandpass came out close to what you want.
Problem is that I've misplaced the details of where "here" and "there" are. I've also misplaced much of my memory.
Anybody remember the details of that technique? Or any technique that might be practical on a filter with no tuning adjustments and little chance to recover from mistakes?
Is this out of a commercial radio? If so, the maker and model please. Perhaps a photo. I can't imagine anyone shipping something like that without tuning adjustments.
That's an understatement. It's nothing but interactions. Change one thing and all the adjacent resonator tuning changes.
2MHz / 146MHz = 1.4% That should be easy without unwinding turns. Any clue on the expected bandwidth?
Well, I'm not sure this is the correct procedure, but when I was making 902-928Mhz solid cavity duplexers, I put a reflection coefficient bridge on the input (no network analyzer available) and swept for lowest return loss and best looking sweep. As I vaguely recall, it was 5 sections, with a total of 9 adjustments for tuning and coupling:
I might have shorted out downstream sections to reduce the display clutter, but there were no provisions for that inside the cavity. So, you start at the input, tune until the curve looks best. Move on to the next stage, and tune again. Because the insertion loss of the first stage is presumably very low when it's properly tuned, the 2nd stage will have a big effect on the pattern. However, if the insertion loss through the first stage is high for some reason, tuning the 2nd stage will have minimal effect on the displayed pattern. After you're done making the return loss look pretty, check the overall response. If the return loss looks good, the bandpass response will also look good. However, don't bother trying to tune it for best bandpass response. It just doesn't work, which is probably what your friend is struggling with.
If you have access to the helical resonators, you can see which stage is responsible for which dip by waving a ferrite or brass rod near the coil.
Yep. That's it. Everything affects everything else, but when you're done tuning for best looking return loss, the bandpass response will magically be perfect.
Try googling for repeater cavity duplexer tuning procedures. It's very similar, except that you're tuning a notch instead of a peak, with the added bonus of overly critical, white knuckle, adjustments.
This Agilent app note looks interesting, but I don't have time to read the whole thing:
A digital camera is your friend. Take lots of photos so you at least have a chance to put Humpty Dumpty back together again.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Design instructions but no tuning info per se. You could contact them and see if they have more. Then there is a book by Zverev. The best "McGyver style" alignment instruction I ever saw was in an ARRL handbook but I can't recall what year. You might want to ask in a ham radio group if someone still has it.
formatting link
Not sure how far the Zverev book goes into tuning, I don't have it here.
It does and thanks for jogging my failing memory. I it here. "Handbook of Filter Synthesis" by Anatol I. Zverev. (1967)
9.4 (P513 to 521) is on "Alignment of Helical Resonators" with references to Dishal's method. It's too much to type here, but I can scan the pages and post them if needed.
Google finds quite a few hits for Dishal's Method.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Dishal's method. For high-Q filters, mistuned series sections look like opens, and mistuned parallel sections look like shorts.
If you know what the resonant frequencies of the sections should be, you can do that with a return loss bridge from each end.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510 USA
+1 845 480 2058
hobbs at electrooptical dot net
http://electrooptical.net
I dunno if it counts as an official method, but I've used a spectrum analyzer+tracking generator. Each resonator makes a "bump" in the passband; you can locate any given bump by grossly mistuning it (with a hand-held cap, for example), and watching the bump jump back and forth. Tune, repeat.
Give up now, while you're both still sane. I don't think you can do much without some way to make measurements. It might be easier to just ship it to someone with the proper equipment and get it done right the first time.
It looks like you're trying to move a 144.0 to 146.0 MHz filter up to
146.0 to 148.0 MHz, or at least that's what you'll get if you don't change the coupling port size.
The guts: Hmmm... you're right. No second chance to get it right.
Looks like all you have to do is chop a little off the end of the helical resonator. How much? Dunno. Cut a tiny amount and sweep the filter to see what happened. If it looks good, cut some more.
Instead of butchering the helical resonators, why not just shove an aluminum or brass rod down the center of each coil? I've done that many times moving 135MHz military band stuff into the ham band. You'll loose some Q but I don't think it's going to take much brass to move the frequency 1.4%.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.