Cavity Filter

It has been a (long) time since I’ve written a post. I have to admit that just thinking about building the cavity filter was enough to stall my progress. It shouldn’t have. It’s not all that hard to build, though there are some significant challenges.  The construction instructions from Scotty are located here, and I’ve pasted a copy of his schematic diagram for reference:

Scotty’s 1013.7 MHz Cavity Filter

Cash’s kit doesn’t include any parts for the cavity filter, so you have to find them yourself.  To start, you need some RG-141 coax cable.  From the pictures Scotty has posted, it appears that he used actual “hard pipe” coax with a solid metal outer conductor. The cable I ordered on eBay is similar to the RG-405 used in the rest of the kit, but it’s larger and has a outer plastic jacket. The eBay supplier was “rfsupplier_com”, here is a search link that may or may not work.  Also, you need to find some 1″ copper pipe and som 1/4″ copper tubing.  Both are available from Home Depot, though you have to buy them in 10′ lengths.  This is enough pipe for about 9 cavity filters and costs almost $90.  Finally, you need to find some .0625 brass sheet.  This isn’t available at Home Depot.  I was able to find it in town at a hobby shop that sells R/C airplane, helicopter and car stuff along with model trains.  It was in the bulk materials section where different plastic extrusions and metals are.  A piece of material big enough for the filter cost about $8.  While I was there, I also bought some steel wire to align the cavities during soldering.

Tubing straightening device.

The first step to building the filter is setting up the materials.  The most tedious step is straightening the tubing and cutting the pipe.  To straighten the tubing, I used a piece of pressure-treated lumber with a hole drilled in it slightly larger than the tube.  I forced it through the hole and pulled it straight.  There was still a little bend still left even after this process, but when it’s cut into pieces it appears straight enough.

Cut Pipe Segments

Next, it’s necessary to cut the outer pipe segments.  I put a stripe of tape down the length of the 10′ pipe and marked the cut point.  It turns out that didn’t need the tape, and it’s a pain to remove so I don’t recommend it.

Polished pipes

Once the tape is removed, polish the pipe segments.  I used a much bigger drill bit than the holes to clean off the burrs.  I didn’t mention the process of drilling the pipes, and I don’t have any pictures of it.  They were all drilled at a friend’s house.  He used his drill press and a drill vice to make up a jig.  This made it easy to drill many precisely measured and well centered holes.

Centering guide installed

Scotty suggested using a wooden dowel or a really long drill bit to align the holes.  I used steel wire that was about the right size.  I’ve only ended up soldering the wire to the assembly once so far.  To remove it I heated the joint while pulling and it came loose.

Pipes clamped and fluxed, ready to solder.

I didn’t have any C-Clamps, or any other appropriate clamps, around that were the right size, so I used some vice-grips.  This way, I couldn’t solder all 4 at the same time.  I soldered the outside pairs, then I soldered the pairs together.  I pressed the pipes flat onto a flat surface, and tried to make sure that they were aligned longitudinally.

Soldering the pipes

Heat the pipes up using a torch until the flux is molten, and the pipes are nice and hot.  After a while, I’m not sure exactly how long, drop some solder onto the joint.  Once everything is at the melting temperature, move the torch back and forth along the joint to make sure that it’s evenly heated and soldered.

Soldering finished

After each side pair is soldered and cooled I soldered them to each other.  This photo is from the first time I tried soldering like this.  There was lots of splatter and carbon everywhere.  After a while I got the hang of it.  The biggest thing for me was that my torch got really rich combustion upside down and would have a long yellow flame, then got out.  I learned to reduce the valve setting when the torch was inverted for a good burn.  Also, some steel wool or Scotchbright pad polishes these back up nicely.

Tuning screw nuts soldered

On the top plate of the filter, there are a series of tuning screws.  There are a few ways to make the tuning screws work.  One is to drill and tap the brass plate itself, which has an elegant simplicity (but I don’t have any taps), another is to solder nuts onto the plate.  I chose the later.  I drilled the plate first, then threaded the screws into the nuts to align the nuts over the drill holes.  It was a little difficult to remove the screws, but once I did the thread easily in and out.

Pipes soldered and polished

Another step that is missing pictures is soldering on the top plate of the filter.  Scotty recommends soldering the bottom plate on first, then the top (For good reason!  See below).  I didn’t have any of the RG141 coax at the time, so I decided to solder the top on first.  It was hard to make sure it was centered.  What I ended up doing was using some wooden stock with a hole drilled into it for the tuning screws.  The pipes slipped over them, holding them centered.

View into the filter

In this view, you can see the tuning screw holes and a screw.  It’s a little gross in there, probably because of the wood spacer I used, and I hope it doesn’t matter too much.

Bottom plate drilled and soldered

The bottom plate has many drill points.  It’s somewhat difficult to get all of them done precisely.  I didn’t have a drill bit that was the “right” size for the hairpin couplers, so I just choose the smallest one available.  I think this is a good thing because it gives some extra clearance for small drilling errors.

Stubs soldered onto bottom plate

To solder the stubs onto the bottom plate, I made another wooden jig with a hole for each stub cut to exactly the right depth.  This made it rather easy to assemble.  Once the tubing was put in the holes, the bottom plate was placed over them and they were soldered on.  The wood definitely scorched, but because it was mostly starved of oxygen, it didn’t burn.

Bent hairpin coupler

Finally, the hairpin couplers need to be installed.  I made the hairpins first, then installed them into the cavity tubes.  I’m not sure if there is any other way to do this, it’s all I could think of.  To make them easier to install, I bent the longer leg a little bit so I could feed it through the inter-cavity aperture.  The shorter side has a tendency to catch the cavity wall, so it may help to pull it back a little.

Hairpin installation

I used a pair of needle-nose pliers to insert the hairpin into the cavity, thread the opening, and pull it into place.  this was a somewhat challenging process.  The hairpins get mangled a little in this process, so I flattened them along the cavity wall to straighten them out.  Once the long side is straight, I put the insulators on that side.

Hairpin coupler installed

This is a close-up of the hairpin installation.  You can see the insulator between cavities and the 2 insulators along the wire within the cavity.

Bottom plate installation

At long last, you can install the bottom plate.  I have to stop here, because I don’t have any connectors for RG141.  I’m going to order some next month, so I’ll be able to install those hairpin couplers.  Once I do that I’ll solder it up for good.  So exciting!!!  I’ll post here again once I do that, and I’ll hopefully be able to include some plots detailing their filtering performance.

Update 8/29/2010:

It is very important that you do not attempt to solder the bottom plate to the filter the way I did it.  When I was heating it to solder the hairpin coupler wires, the center tubes slipped into the cavities, ruining the filter.  I had to clamp the top plate into the vise and desolder it.  This allowed me to recover the lost center tubes.

Top plate desoldered

After recovering them, I had to re-solder them to the bottom plate by clamping the bottom plate into the vise with the filter upside-down.  I vice-gripped a nail to the work bench vertically so that the point of the nail kept the tubing (mostly) centered and (mostly) the correct length.

Bottom plate soldering

This is the end result of the bottom plate’s soldering.  You can see the wires on each side of the cavity walls and the center tube soldered in place.

Inner tubes replaced

With the center tubes back in, I straightened them with a long allen wrench.  I also re-polished everything, because I’m O.C.D. like that.

Cavity filter internal view

I took the opportunity to see what the hairpin couplers looked like while installed in the filter.  You can see the insulator through the cavity aperture and the 2 spacing insulators against the cavity wall.

Finished filter

Finally, here is a picture of the finished filter.  I haven’t installed the connectors on the coax yet (because I don’t have them yet), but other than that, it’s done.  There is a lot of nasty soldering here, but I was learning as I went.  I’m sure that next time it’ll be much nicer.

I still can’t wait to get some connectors and try it out!  I’ll keep this page updated.

  1. #1 by Gareth on May 25, 2011 - 3:14 am

    Hi William,
    I’m just begining to build scotty’s Spectrscope, and was going to tackle the cavity filter first, your page has provided a great deal of insight into its construction, do you think solid brass/copper inserts would be a better option rather than the “bendable” pipe?

  2. #2 by hpux735 on May 25, 2011 - 5:45 am

    Sure, some people have done that. It seems like a good idea. Someone even tapped the holes and threaded the rods so they didn’t need any additional tuning screws. I imagine it will be more expensive, though.

  3. #3 by Malcolm on August 3, 2012 - 10:29 am

    Hi William,

    I’ve read your articles with great interest and would like to know where you can buy a kit of parts for Scotty’s spectrum analyser or at least the printed circuit boards?



  4. #4 by hpux735 on August 6, 2012 - 1:07 pm

    The best course of action is to ask on the Scotty’s Spectrum Analyzer yahoo group:

(will not be published)

Please complete this capcha. I get almost 1000 spam comments a day! * Time limit is exhausted. Please reload CAPTCHA.