Notch Filters and Effects

[KevinC]

Interesting thread. I would only comment that perhaps 95% of our filters go to the aero, commercial FM broadcast, 2 way etc. These are not the scanner filters per se- but some are similar in topology.
We have filters at Cape Canaveral, in front of GPS systems, flying in space, on avalanche sites in Asia, on probably the majority of commercial ships on the Mississippi, hundreds of muni airports and FM translator sites- on and on.

Regards,

Dale W4OP
for PAR Electronics, Inc.

I'm glad you chimed in Dale! I've purchased numerous filters from you and ALWAYS recommend your company. Your customer service can't be beat and your products are top-notch in performance and construction.

Keep up the good work.

 

[mancow]

Here's a real work example and maybe you guys can help me out.

Uniden BCD996T
DB224 4 bay VHf antenna
Antenna height ~ 140 ft atop a water tower
Antenna fed with LMR600 sized cable
Unknown power level 158.100 MHz pager 1 mile away
PAR 158.100 filter installed at the scanner

The system is inundated with NOAA intermod mainly across the 163-174 range. 162.475 approx 50 miles away, 162.425 approx 40 miles away, 162.525 approx 30 miles away, 162.550 approx 60 miles away.

The PAR filter has taken care of most of the offending 158.100 pager trash. However, the FED band remains plagued bythe NOAA signals mixing with 153-160 MHz LMR traffic.

Should I use a stub at 162.525? Will the height and gain of the system compensate for the loss?
Should I use a NOAA PAR filter instead?

 

[prcguy]

Do you have a spectrum analyzer that can look at the VHF band to see what is contributing to the intermod?
prcguy

 

[KevinC]

Do you have a spectrum analyzer that can look at the VHF band to see what is contributing to the intermod?
prcguy

X2...otherwise you're just shotgunning it.

 

[W4OP]

NOAA WX Notch Filter

Here's a real work example and maybe you guys can help me out.

Uniden BCD996T
DB224 4 bay VHf antenna
Antenna height ~ 140 ft atop a water tower
Antenna fed with LMR600 sized cable
Unknown power level 158.100 MHz pager 1 mile away
PAR 158.100 filter installed at the scanner

The system is inundated with NOAA intermod mainly across the 163-174 range. 162.475 approx 50 miles away, 162.425 approx 40 miles away, 162.525 approx 30 miles away, 162.550 approx 60 miles away.

The PAR filter has taken care of most of the offending 158.100 pager trash. However, the FED band remains plagued bythe NOAA signals mixing with 153-160 MHz LMR traffic.

Should I use a stub at 162.525? Will the height and gain of the system compensate for the loss?
Should I use a NOAA PAR filter instead?

I have not looked at stubs in many years. Maybe someone else can chime in. From memory, typical notch depth of a single stub is in the region of -30dB. If one uses hardline, a greater notch can be obtained. The bigger issue is that being a simple 2 pole filter, the -3dB bandwidth is very broad. In fact I just put a stub on our VNA. The center of the notch was at 100MHz and 32dB deep. The -3dB points were +/-30MHz. At +/- 10MHz the loss was a bit over -10dB. So, the -3dB BW is +/-30% of center.

Remember also that the stub will also have notches at 3/5/7 etc times the fundamental notch frequency and VSWR upset quite a ways out from the notch.

By contrast, the PAR NOAA filter has a -3dB BW of +/- 1.5% of center or +/- 2.43 MHz.
So in the end it probably comes down to how close in to the WX notch do you want to listen.
I hope this helps and that I got my numbers correct.

Dale

 

[zz0468]

Dale, your numbers sound pretty reasonable to me, and fit with the plots I posted to this thread a couple of years ago.

In the case that mancow is dealing with, I would do nothing until I put a spectrum analyzer on the end of the feedline and see what's really going on. I've seen plenty of cases where some out of band signal, like an FM broadcast transmitter, causes enough overload to a receiver to make the front end go non-linear, and then the mixes that show up are in-band and have nothing to do with the FM transmitter that's causing the overload in the first place.

I have a similar situation right here at home that's caused by nearby cell sites, and the intermod manifests itself in the 2m and 440 ham bands. I get multiple 440 mixes across the band, and when I notch out 860-900 MHz, the problem goes away. I never would have guessed that without a spectrum analyzer.

 

[W4OP]

Your advice is spot on. It is very difficult to troubleshoot w/o being at the site. I 100% agree on FM broadcast- they run high ERP and are on 24 hours a day. That energy in a front end just takes away from precious headroom.
We developed a 4 cavity notch filter for the cell spectrum you mentioned- initially for the Houston area but finding other markets as well. It is not listed on our site, but I will be happy to send plots and info for anyone that is interested.

Dale

 

[mancow]

Thanks guys for the replies.

I don't think it's broadcast intermod simply because it's a smaller town generally 50+ miles from any towers of appreciable strength. There is one 90 something Meg translator (think that's what they call it) booster type site on the same tower where the 158.100 pager is about a mile away. I never hear any hint of broadcast interference though. Everything mixing with NOAA seems to be local 154/155 range LMR and some business stuff I haven't yet identified.

The only real spec-an I have is an HP8924C. It's enormous but can be hauled out there. Other than that I just have the SDR dongles oh and an AOR SR200A with a rough spec-an feature.

I would really like to be able to focus in on the 165-172 Fed range if possible.

Also, I have some miniature repeater duplexers. Could they be of any use?

 

[zz0468]

...Everything mixing with NOAA seems to be local 154/155 range LMR and some business stuff I haven't yet identified.

Just remember, the energy in the receiver that's causing the overload does not HAVE to be anything that's audible to your receiver, either directly or as a result of the mixes you're hearing.

The only real spec-an I have is an HP8924C. It's enormous but can be hauled out there.

Pay note to the "two person lift" sticker on the side! It's an excellent instrument, but heavy. But it will do the trick for you. I'd pack the thing up and take it to the site. It's critical to understanding what's going on, so you can apply the proper fix. A NOAA WX notch filter won't do a lick of good if the problem is being caused by a nearby cell site. The 8924C will tell you. Be sure to look from 30-1000 MHz. And don't be surprised if the problem ends up being 1.9 GHz PCS. I've seen that, as well.

So long as that front end preamplifier transistor in the scanner is overloaded, the mixes you hear will merely be dependent on the math working out top create something that appears in the IF.

Other than that I just have the SDR dongles oh and an AOR SR200A with a rough spec-an feature.

Useful for seeing what's out there to listen to, not so much when it comes to finding interference. Partially because they're both going to be quite prone to getting overloaded themselves.

I would really like to be able to focus in on the 165-172 Fed range if possible.

You COULD start with a band pass filter to cover that range and see if that solves the problem, but it might not. It wouldn't be a bad thing to go from the antenna, to the band pass filter, to a notch on whatever specific frequency is causing the overload (if needed), to a good low noise preamp, and then to the receiver. You could end up with an uncrunchable front end that hears really really good.

Also, I have some miniature repeater duplexers. Could they be of any use?

Probably notch type. If the frequency that needs to be notched out is withing the range of the duplexer, you could tune one side to the frequency you want notched, and put a dummy load on the unused port. There's nothing in the rule book that says you must use both sides of a duplexer. Might save you a few bucks. And it might be narrow enough that it doesn't take out the entire band.

 

[KevinC]

We developed a 4 cavity notch filter for the cell spectrum you mentioned- initially for the Houston area but finding other markets as well. It is not listed on our site, but I will be happy to send plots and info for anyone that is interested.

Dale

And it works quite well!

 

[mancow]

Thanks zz0468. I will try to haul it over there tomorrow and see what's going on. I will note the largest signals across the range then post an update to see if you guys can maybe help come up with an idea of what to target and how.

By the way, the Vhf antenna works surprisingly well across the other bands. I suppose the altitude is the factor but it makes for a decent overall scanning system even up to 800 MHz.

 

[zz0468]

Thanks zz0468. I will try to haul it over there tomorrow and see what's going on. I will note the largest signals across the range then post an update to see if you guys can maybe help come up with an idea of what to target and how.

Set the spectrum analyzer on peak hold for a while and see what comes in. Then, maybe you can take a picture of the screen and post it along with your notes.

By the way, the Vhf antenna works surprisingly well across the other bands. I suppose the altitude is the factor but it makes for a decent overall scanning system even up to 800 MHz.

I'm sure the antenna elevation is a factor. But outside the antenna's designed frequency range, it's essentially a random chunk of metal. Anyone who's ever replaced the rabbit ears on a TV with a coat hanger knows that a random chunk of metal can usually collect enough signal to be useful.

 

[woodpecker]

I've built many passive filters from LC types to helical notches, the only way to really remove a single in band frequency close to where you want to listen is using loose coupled helical resonators or cavity resonators.

As mentioned to build a sharp notch you need a very high Q, coax stubs don't make sharp notch filters, you can improve them using RG214 and a small series capacitor.