Filter in a die-cast box

Sunday comes bright and early and I decide to put the FM broadcast band notch filter into a die-cast aluminum box.

First, I found in the garage the G112T box I had purchased a long time ago. Measuring somewhat carefully, but not enough, I estimate the spots on the two ends of the box that I’ll have to put 1/2″ holes to install these little TNC-f to coax pigtails.

The TNC-f chassis mount bodies are just about as big as the sidewalls of the box, so aiming carefully is important. However, I miscalculate low on the first hole, and have to oval it just a hair to fit the body of the connector. The second hole I overcompensate, and it’s a little high. But, within 30 minutes I have the case drilled and ready for the filter/coax assembly.

img_4462_sm img_4461_sm

The little TNC coax pigtails came from a buddy of mine – he had a billion of them from some former project.

Before I take the filter off the roof, I measure it one more time while it’s foil-wrapped. Didn’t have a photo of the install from last night, so here it is today.


Kinda ugly, no?

Using the Icom setup, it appears a little worse than yesterday night. Could be so for a number of reasons, including propagation, xmtr power output changes between night and day, or the foil moved a bit during the night. Not sure.


Taking the filter down off the roof, I cut off the SMA-f board mount connectors, clean off the excess solder, and prepare the TNC jumpers to solder directly to the board in place of the former connectors.



The reason I prepared the coax open end as shown in the above picture is apparent in the next photo when I solder the TNC jumpers to the filter board.




Now to install the filter and coax assembly into the enclosure and prepare the sealing gasket.






The box is almost ready to close up. Don’t want to lose the little screws that are intended to hold down a board inside the case so I install them now.


Next, put on the lid and tighten down the screws. Use a compression clamp to squeeze the case closed instead of making the screws do the work. Aluminum strips out very easily.



All done. Now, since the rooftop cable from the antenna is an N-m connector, and the feed down to the shack is currently a UHF-f adapter, I add those to the case.




Voila! All done. Now, to the roof!


Here’s a shot of the discone up on the roof. The thing attached to the chimney is a 2.4 GHz Ubiquti 2.4 GHz Bullet M2 set up as an access point, so I can get Wi-Fi pretty much anywhere around the neighborhood.




It does appear that the box is certainly no worse than the foil and at least for a few stations it’s actually improved on the rejection. That pesky 107.9 MHz station is down 10 dB with the box.

Since that wasn’t as hard as I thought it’d be, I need to order a few more of those nice little G112T boxes, or something similar!

While I’m at it, what TV channels are detectable here?

A couple posts ago I ran a scan of radio spectrum observed from a discone atop the house. There were a bunch of TV stations observed.

This FCC site comes in handy to check these.

Channel Callsign Start Freq
Stop Freq
XMTR location Distance (km)
8 KAET 180 186 South Mtn, PHX 31
10 KSAZ 192 198 South Mtn, PHX 31
12 KPNX 204 210 South Mtn, PHX 31
15 KNXV 476 482 South Mtn, PHX 31
17 KPHO 488 494 South Mtn, PHX 31
20 KPAZ 506 512 South Mtn, PHX 31
22 KNAZ 518 524 Mormon Mountain 158
24 KTVK 530 536 South Mtn, PHX 31
26 KUTP 542 548 South Mtn, PHX 31
30 KUAT 566 572 Mt Bigelow 174
31 KSAZ 572 578 South Mtn, PHX 31
33 KTVW 584 590 South Mtn, PHX 31
35 KFPH 596 602 South Mtn, PHX 31
36 KAZT 602 608 South Mtn, PHX 31
38 K38IZ-D 614 620 South Mtn, PHX 31
39 KTAZ 620 626 South Mtn, PHX 31
40 KEJR-LD 626 632 South Mtn, PHX 31
41 KPDF-CA 632 638 South Mtn, PHX 31
42 KVPA-LD 638 644 South Mtn, PHX 31
44 K44CN-D 650 656 Mingus Mtn 122
45 K45MW-D 656 662 Sentinel, AZ 143
46 KDPH-LD 662 668 South Mtn, PHX 31
47 KDFQ-LP 668 674 Mingus Mtn 122
48 K48NH-D 674 680 Globe 104
49 KASW 680 686 South Mtn, PHX 31
50 KFPB-LD 686 692 South Mtn, PHX 31
51 KPPX-TV 692 698 South Mtn, PHX 31

Some of these are a little hard to believe, like the low-power transmitters at Sentinel and Globe. Mingus, Bigelow and Mormon are somewhat more likely…

Using a FM notch filter

Earlier I mentioned that I’d gotten an FM notch filter via eBay  from iseeabluewhale. That filter is now up on the roof in a metal box, wrapped with aluminum foil. The discone has a 6′ pigtail of LMR-240 coax, that then goes into the FM notch filter, the output of the notch filter goes into the LMR-240 that runs down into the shack. It’s not a super installation, but let’s see what it looks like from the receiver’s point of view.


Looks bad, but still better than the original (red, below) but significantly worse than the terminated version (green, below).


So, the filter does seem to do a fairly good job of cleaning things up.

Next, I removed the cast enclosure from the mix and just wrapped aluminum foil around the filter assembly and over the coaxial cables connected to the filter.


Even better than before. Filter blow-by is a real issue here. We are talking a filter that advertises 85 dB rejection, after all.


As well, I realize that the 25 cm RG-178 coaxial jumpers that I’m using to get from the LMR-240 to the filter are partially exposed to the world, so I’ll go do something about that right now.


More shielding is better, so it appears! I’ve wrapped the Al foil from LMR-240 connector to the other, covering all the RG-178. Perhaps I’ll fix that next. I believe I’ve a small die-cast box in the garage that I could use.

FM broadcast stations

The last post involved sniffing the spectrum around the house from 1 to 1300 MHz. It was obvious from the scan that the strongest signals were all in the FM broadcast band. Time to dig in a bit more and identify these culprits.

A very handy tool is, which is chock-full of useful information on commercial radio stations. If you’ve never used it, I encourage you to check it out.


Sweeping the FM broadcast band again, the strongest carrier is at 98.7 MHz. This is KMVP atop South Mountain, with an effective radiated power of 97 kW.  Another very strong one for me is 88.3 KNAI, located over on Shaw Butte, only 5 miles west.

Let’s now remove the antenna and replace it with a dc – 18 GHz termination. Ideally, there should be nothing but a flat line at the receiver’s noise floor. Sadly, this is not the case.



There are two pieces of LMR-240 that make up the coax run from the antenna to the shack. The first, from the shack up to the roof, is about 30′. The second, up on the roof from the cable outlet to the antenna, is about 60′. The following plot shows the reception with the second piece disconnected from the first and the termination attached to the roof end of the 30′ piece.


It appears that there’s not even 60 dB of isolation between termination on and antenna. Rats – there should be around 90 because it’s pretty good coax, and the connectors seem to be installed well. Let’s look more closely to see whether these are external or internal.


It appears that all the signals are from over-the-air transmissions, except the one around 96 MHz, doesn’t appear to correspond to any OTA signals.

Station KNAI (88.3 MHz) broadcasts from Shaw Butte, only 8 km or so west, while all the others have transmitters atop South Mountain, over the Phoenix mountains from this location.  The big guns are 100 kW ERP.

In any event, it appears that my setup is somewhat leaky, with only about 60 dB through-cable loss.

Local radio spectrum scan

Interested in installing a low-noise wideband amp on the rooftop discone to allow me to aggregate together a bunch of receivers and reduce the amount of clutter in the shack.

I have an old Diamond discone on the roof, only 5 – 6 m above ground, and about 30 m of feedline to bring it down into the shack where all the scanners and other receivers are.

I’d purchased on eBay a super-hot LNA from iseeabluewhale with NF of 0.5 dB at 400 MHz, 20 dB of gain, and a useable bandwidth of 10 – 2000 MHz.  Most importantly, it’s got an output IP3 of +42 dBm at 1 GHz, so it should be able to withstand nearly any level of signal that I can throw at it and stay linear.

First thing is to see how noisy the environment is. I set up the Icom PCR-1000 radio to scan frequencies from 1 MHz to 1300 MHz and see what kind of average noise levels I was dealing with out there. The discone is about 1 – 2 dBi or so, the cable loss varies over the frequency span. The Icom was configured to step in 25 kHz increments with a 50 kHz bandwidth.

The antenna is really only competent from about 100 or 120 MHz up, but it has a bottom-loaded coil atop the disk to supposedly get down to about 50 MHz or so.


1 – 100 MHz: No surprise here, the strong signals are all FM broadcast band. Some are exceedingly strong, although I’m pretty sure the closest one is at least 8 km distant. No digital TV signals between 54 and 88 MHz.


100 – 200 MHz: The rest of the FM broadcast band, some peaks in the aviation, 2 m ham, and VHF 150 – 170 MHz bands. The big rectangular humps at the right are digital TV channels 8 and 10.


200 – 300 MHz: Here, there’s just TV channel 12, then it’s pretty quiet beyond, with the noise floor about as low as the Icom can report.


300 – 400 MHz: The innards of the Icom must switch receiver paths or filter setups at 350 MHz, since the floor was rising smoothly and then made the 5 dB step down. Otherwise, it’s pretty quiet still.


400 – 500 MHz: Still quiet, random narrow-band carriers here and there until the 450-470 MHz two-way band, plenty of carriers there, and then TV channels 15 and 17.


500 – 600 MHz: A bunch of TV channels, from left, 20, 22, 24, 26, 30, 31, 33, and the lower edge of 35.


600 – 700 MHz: All DTV still, 36, 38, 39, 40, 41, 42, 46, 47, 49 and 50, and maybe 44, 45, 47, and 48. Otherwise, quiet noise floor.


700 – 800 MHz: Mostly quiet, notice the noise floor’s come up from the last plot, and there’s LTE cellular for Verizon and AT&T.


800 – 900 MHz: In a normal receiver, there’d be all sorts of signals between 824 – 849 and 869 – 894 MHz, but because this receiver is not allowed by law to tune those frequencies (stupid law; nothing to hear but digital noise) there’s nothing there. However, my $10 USB SDR dongle can tune this spectrum, so maybe I’ll look there to see how strong the signals might be.


900 – 1000 MHz: The 902 – 928 MHz part 15/18 band has plenty of small carriers, lots of activity, but it’s all pretty weak. There’s more carriers up in the business bands all the way up to about 960 MHz, then gets quiet again.


1000 – 1100 MHz: Between cable loss and the absorption by foliage, the signals here are all generally pretty weak. Don’t know what that plateau is at about 1015 MHz, or the hump around 1025 MHz. The one at 1015 MHz looks suspiciously like a digital TV carrier… The ADS-B stuff at 1090 MHz barely shows, though there’s hundreds of transmissions per second.


1100 – 1200 MHz: Really not much here that my receiver can detect.


1200 – 1300 MHz: Same here.

So, it appears that the bulk of the stronger signals in range are all FM broadcast or digital TV stuff. FM broadcast really beats everything else. Let’s take a closer look at that next.


American AAL1920 ADS-B Capture

Living next to a large hub airport, and having a good view ADS-B-wise of the airport and a to 200+ nm around it, I get to see flights come into PHX, unload and load, then take off. Case in point: AAL1920.

The route it’s currently on is LAX-PHX-MIA. My ADS-B receiver first heard its beacons when it climbed over the Santa Ana mountains in Orange county. It’s now approaching Deming NM and it should drop off my display very soon.


Total time recorded is almost 2-1/2 hours, and nearly 9000 position reports.

When zoomed in on Sky Harbor, the path is


What I find interesting is that as the plane was on final landing to the east (right), it did a little bump in the path just a few miles before the runway. Why? Avoiding a PHX PD copter? Dunno.


Once it landed (rwy 8R), it taxied to the even-number side of the A15-A30 part of Terminal 4. Kinda looks like gate 16 or 18.

After loading up its passengers, it taxied out and took off to the east on runway 8L.