For those in Canada, are there any suppliers/stores that has a saw filter? I know that there’s a place in Mississauga but was wondering if anybody knew of another place.
Thanks,
Mike
I’ve been buying EPCOS 1090 MHz SAW filters on eBay from a guy in Europe – delivery takes about 10 days to the U.S.
Remember that the front-end of most of these SDRs consists of a very wideband variable gain low noise amplifier (LNA) followed by a wideband detector which is used to set the LNA gain. This means that the LNA gain is going to be set by the strongest signal or signals fed to the LNA – and that ain’t going to be 1090 MHz! This makes it important to filter out the strong signals we don’t want, which means mostly broadcast radio and especially television in the 80 - 850 MHz portion of the spectrum. (Yes, there are other sources as well – I’ve got some paging transmitters nearby, and cellular might be a concern.)
The insertion loss for the EPCOS SAW filter is on the order of 4dB. Before you go to a SAW, try putting a high-pass filter on the input such as the Mini-Circuits SHP-1000 (or NHP-1000), also available on eBay. The insertion loss for these filters is negligible (under 0.5dB at 1090 MHz), and they’ll cut out almost all the broadcast TV and FM. They won’t do anything for cellular signals around 1900 MHz, but response of many SDRs is falling off at that point. The SHP-1000 also gives you the benefit of providing some front-end protection from static buildup and such.
The configuration I use is the antenna (a discone on the top of our 2-story house) feeding a SHP-1000 high pass filter followed by a low-noise amplifier driving a EPCOS 1090 MHz SAW, then the SDR. Putting the SAW after a 20dB low noise amplifier reduces that 4dB insertion loss.
The EPCOS parts are a 3.5 x 3.5mm surface mount part, but hand prototyping is possible. I’m looking at doing a pc board layout for the low noise amp, SAW, and surface mount high pass and low pass filters (Mini Circuits HFCN and LFCN parts).
If there’s interest, I can write up what I’ve done – I’ve got spectrum analyzer data from the discone with and without the high-pass filter; the difference is substantial! Just using the high-pass, the majority of the energy outside the band of interest (1090MHz) is gone; I still see some residual energy in the 850 MHz range from digital TV, and energy in the 1800-1900 MHz range from cellular. I tried quarter-wave stubs to further filter these, but didn’t notice any marked improvement in number of aircraft - number of positions observed.
– Bob K6RTM
^ Thanks for info and the explanation. Doing a buy it now on ebay for the filter before I try anything else.
–J
The VBFZ-1065 is a good bandpass filter.
(1) It has a higher insertion loss than the SHP-1000 or NHP-1000. (1.7 dB for the VB, 0.5 dB for the HP) – not a really big deal
(2) It has lower attenuation at 800 MHz than the HP-1000 (7 dB for the VB, 16 dB for the HP) – this is more important
(3) The SHP-1000 and NHP-1000 show up on eBay quite often. Don’t recall seeing the VBFZ-1065.
Unless you have significant energy in the 1100 - 2000 MHz range, I’d go with the HP-1000 high pass filter. If you have significant energy in that range, then the VBFZ-1065 is a better choice.
– Bob K6RTM
Thanks, I’ve ready about the VBFZ-1065, it’s the shipping from the US that’s a killer (brokerage fees) as minicircuits doesn’t ship using USPS. Guess I’ll wait to hear back from the vendor out in Mississauga to get back to me regarding the K&L Microwave 4B341-1090 filter. I have a LNA4ALL filter here that I’m planning on using but not until there’s a filter.
How can you tell? Do you need a spectrum analyzer? If so, what would you recommend?
While a spectrum analyzer can help, in the end, the measurement that counts is what do dump1090 and PiAware show – did the change I made result in increased birds seen and positions reported?
I have a spectrum analyzer (an older one, but it works and it’s paid for). You need something that matches the range of your SDR; for the RTL SDRs, something like mine which tops out at 2.9 GHz works fine. I’ll get a web page on this stuff started and post the spectrum analyzer before and after shots from my discone antenna and the high pass filter.
Better equipped ham radio operators in your area will probably have a spectrum analyzer; they are an extremely versatile tool. The usual suspects are Anritsu, HP/Agilent, Rhode Schwarz, and at the low end, Rigol. You can pick up a used Anritsu 2711 portable for $1800 or less.
Unless you’re in an area with cell towers staring you in the face, I’d start out with a high-pass filter. The SHP-1000 and NHP-1000 are usually available on eBay for around $20 (half of their price new). (a quick look shows 3 of the NHP ones for $20 each and 2 of the SHP for $32.) New from Mini Circuits they’re around $40.
In my area, I have significant energy in the 750 - 800 MHz region from broadcast TV. When you compare the frequency response of the SHP-1000 with the VBFZ-1065, the SHP-1000 gives you better than 10 dB more attenuation in this range, and that’s significant.
The other area where a spectrum analyzer comes in handy is for making traps. I see some energy in the 1900 MHz area from cellular. An open quarter-wave stub of coax a quarter wavelength long looks like a short at that frequency. (the fine print – that’s a quarter wavelength taking the velocity factor of the coaxial cable into consideration.) Quarter wavelength for 1900 MHz is a few inches long. With the spectrum analyzer I can start with a length of cable and clip it shorter a bit at a time, watching the spectrum analyzer until I get the response I want. When I did this for 1900 MHz, the spectrum analyzer showed mw I’d gotten rid of it. But FlightAware didn’t see an appreciable increase in the number of birds spotted or the number of positions reported. Not worth it!
But if you had a high pass filter with a lower cutoff, say 700 MHz, that filter and one or more stubs might do the trick.
–bob k6rtm
I purchased a couple of these K&L Microwave bandpass filters off eBay:
ebay.com/itm/151059589242
They were $50 when I bought them - looks like he only has a few left and the price went up to $75. Something like this is pretty effective at attenuating strong signals (away from the ADS-B frequency) that would otherwise overload the SDR dongle. Putting the SDR and Raspberry Pi in metal enclosures and adding this filter solved RF interference problems I was causing from ham transmissions.
Bob W1QA
That was the one I was trying to get from a place just west of Toronto that has these. I asked for a quote for one, I a response for a ridiculous $1050.00!
The either somebody punched in 100 quantities or they got the decimal in the wrong place… asked them to check the price again.
Bob–
I saw the K&L filters; they looked interesting, but without data on insertion loss or skirt steepness.
Once I get past the shakedown stage, I do simple “enclosures” using PCB material or adhesive-backed copper. One of my favorites is 1-inch copper tape with a conductive adhesive. That’s what’s on my LNA and the SAW. I should do the SDR dongle as well.
I’ve got a very nice cast enclosure with a SMA bulkhead connector on one end, Ethernet and barrel power connectors on the other for the ADS-B rig once I get it stabilized (you know how that goes).
Many moons ago I had a Heathkit weather station in the shack. If I operated on certain HF bands, it thought I was on Jupiter, the barometric pressure shooting up and the temperature plummeting…
73 bob k6rtm
Hello,
I’m new to RF but i have an RTL dongle on an RPI.
the SAW bandpass filters you were talking about seem surface mount? Or do they plug into anything that gives you the n-type plug?
ebay.com/itm/B1602-SAW-filte … 19c16842ba
I mean, I could make a board for it but… Probably would be more expensive than buying the NHP filter or somethign that comes with an n-type.
BTW Still using the stock antennae… 
Yes, the SAW part is surface mount – here’s one http://bob.k6rtm.net/adsb-saw.jpg
Mounted on a board with SMA connectors; the input is on the left, output on the right. The wiring is done with 30 gauge, and yes, the “board” is smaller than a fingernail.
Because of the relatively high insertion loss of the SAW (4dB), I wouldn’t start with that unless you have severe problems, which are unlikely if you’re still on the original antenna. Start with a simple high-pass filter, like the Mini Circuits SHP-1000 or NHP-1000; they have very low insertion loss (around 0.5 dB) and will clean up broadcast TV, FM, and the like. What they won’t get rid of is nearby cellular. These filters appear a lot on eBay. The SHP-800 will work if you don’t have a lot of DTV above 700 MHz – in our area, the DTC spectrum is full, with large signals in the 750 - 900 MHz portion of the spectrum.
(I’m going to do another post on filters, with spectrum analyzer shots of what these things can do – look for it in an hour or so).
–bob k6rtm
The VBFZ-1065-S+ from minicircuits arrived, hooked it up with the LNA4ALL amp and adjusted the gain to what looks is ideal based on the number of mode-s per sec according to PlanePlotter.
Looking at SBSPlotter, I was able to see an improvement right away, probably about 30NM at least. Will have to let it run and compare it to old plots with just the antennas.
A lot of cell towers in the U.S are using a couple different freqs. Between AT&T, Verizon, and Sprint use anywhere from 700 to 850 MHz and 1900 to 2500 MHz. 99% of each cell tower you see each carrier has 3 sectors, facing 3 ways. Just a thought I pass a long.
Ebay package came and filter installed, mini-circuits 15542.
Hard to tell how much of an improvement instantly.
IMHO, the new antenna helped more than anything. I DIY built one using the info here and the SO239 and some 12guage wire. That was like a 10x more planes.
Thanks again for the advice here!.
Is the flightaware 1090MHz filter standard SMA connectors or reverse polarity SMA (RP-SMA)? I am trying to figure out what connector pigtails to buy. – Thanks
Hi Tricycle, standard SMA
Bob k6rtm,
I’d forgotten about this post. What type of soldering iron/equipment do you use to solder these “bugs”.