Dissappointed in RTL-SDR FM Bandstop Filter

[MDScanFan]

I purchased an RTL-SDR FM filter to use as part of a 150-160 MHz setup between an antenna and preamp. I decided on this filter because it's inexpensive, the 88-108 rejection is very strong, and the reported loss around 160 MHZ is <0.7 dB. The excellent RR thread on the "FM Bandstop Filter Comparison" was used in my decision.

I tested the filter using the signal strength meter of my 8600 and confirmed the very strong rejection of 50 to 60 dB across the FMBCB. In the airband the loss is around 3 dB, which matches what I had seen documented. Testing on some strong NOAA signals showed very high loss at around 4 dB. This is much higher than reported.

I repeated these tests with a PAR FM filter and its response matched expectations from FM through VHF high. Loss was <1 dB on the NOAA signals. This tells me my test method is working correctly.

I need several FM filters to support different 150-160 MHz setups and I was hoping to use the low cost RTL-SDR filters for this purpose. I was hoping to not spend $300+ dollars to get a set of minicircuits or PAR filters. I am wondering if I just got a dud and should try another. Has anyone seen similar issues with high loss for these at VHF high? Are there any other <$50 filters that provide around 50 dB FM rejection with <1 dB 150-160 MHZ loss?

(I also use a HPN-30118 FMBCB filter. Loss at 150-160 is low but, like the PAR, it does not provide enough FM rejection (25 to 40 dB) for some of the strong nearby stations.)


...........RTL-SDR....PAR
88...........50...........44
100...........58...........30
108...........55...........30
128...........3...........<1
162...........4.5...........<1

 

[dlwtrunked]

I have all the filters you mention and have measured them. I only now use the Minicircuits one and it is worth the cost.

 

[merlin]

Really, the only way to evaluate filters is to sweep them with a spectrum analyzer W/tracking generator or a network analyzer.
Something like a Nano VNA is good for that.
Center frequency, pass bandwidth, and ripple play major rolls, as do insertion loss.
The number of poles largely determines the skirting at band edges.
Anything -35Db in the FM band is more than acceptable, the idea is reducing strong broadcast from overloading preamps or front end amps.
So you have bad NOAA signals, "Turn the preamp OFF" you probably don't need it and your front end system is being overloaded. You may even need the opposite of gain called 'attenuation'.
(I need that in many cases)
Referring to the RR article the diagram comparing 3 filters "They are different" the Streisberg being Chebychev, the others, low pass/high pass with bandstop about 88 and 108 Mhz. (hence the ripple)
All 3 still effective for their purpouse.
73s

 

[Ubbe]

I am wondering if I just got a dud and should try another.

I've seen reports about the RTL-SDR filter not working as expected, too much attenuation in all bands. The specs do not show that, and the filter comparison thread measured 0,7dB loss at 150MHz, so it's probably bad quality control and should be replaced doing a warranty claim.

/Ubbe

 

[dlwtrunked]

Really, the only way to evaluate filters is to sweep them with a spectrum analyzer W/tracking generator or a network analyzer.
Something like a Nano VNA is good for that.
...

I used a spectrum analyzer with a tracking generator when I did my tests (including combinations of filters). A -35 dB for the FM band would not do here with an FM broadcaster visible 0.8 miles away de-sensing receivers used for weak signals . Some of the filters do not cover well the low end near 88 MHz and others do attenuate too much above 118 MHz. The attached shows the results with me choosing two Mini-Circuits in series although one alone would have done. The others lost primarily due to either not covering the low end near 88 MHz or attenuating too much above 108 MHz.

(Some of us do need pre-amps as we press our setup to the limits to hear things we want to hear more interesting weaker things and use multiple filters to deal with that. That does take money, experimentation, and experience but results in things not otherwise heard. )

 

[MDScanFan]

Agreed. This one is going back and a new one is on its way. If the new one tests poorly as well then I am going to order some different. Thanks.

I've seen reports about the RTL-SDR filter not working as expected, too much attenuation in all bands. The specs do not show that, and the filter comparison thread measured 0,7dB loss at 150MHz, so it's probably bad quality control and should be replaced doing a warranty claim.

/Ubbe

 

[MDScanFan]

One of these days I see myself trying out the mini-circuits FM filter. For now I plan to stick with the <$20 filters.

I have all the filters you mention and have measured them. I only now use the Minicircuits one and it is worth the cost.

 

[vagrant]

@MDScanFan - For under $20 there are also the Nooelec FM filters ( Flamingo ). They have two versions with one being a little deeper, but more attenuation at the skirt where one may not want it if monitoring the air band. They can be purchased on Amazon as well. I have tried the RTL-SDR FM filter and had to repair it due to the poor soldering and or my normal wear and tear. It sits in a box now unused. I use the Nooelec AM & FM notch filters with an SDR now. In particular I like that the Flamingo AM filter is a notch and not an HPF ( High Pass Filter ).

https://www.nooelec.com/store/sdr/sdr-addons/rf-blocks.html

We get what we pay for with these filters. I have a PAR FM filter on the way for testing as well. It is not sealed, so I can adjust as needed and see about keeping things friendly for the air band.

 

[MDScanFan]

The price is right on those nooelec filters. Do you know the estimated insertion loss from 150-160 for the deeper attenuation version? It is hard to tell from the scales of the figures they show

I also looked into the GPIO labs FM notch filter (also listed on the thread noted above). Rejection looks good and VHF high IL looks good as well. I may try one of those if my second RTL-SDR purchase is a fail.

@MDScanFan - For under $20 there are also the Nooelec FM filters ( Flamingo ). They have two versions with one being a little deeper, but more attenuation at the skirt where one may not want it if monitoring the air band. They can be purchased on Amazon as well. I have tried the RTL-SDR FM filter and had to repair it due to the poor soldering and or my normal wear and tear. It sits in a box now unused. I use the Nooelec AM & FM notch filters with an SDR now. In particular I like that the Flamingo AM filter is a notch and not an HPF ( High Pass Filter ).

https://www.nooelec.com/store/sdr/sdr-addons/rf-blocks.html

We get what we pay for with these filters. I have a PAR FM filter on the way for testing as well. It is not sealed, so I can adjust as needed and see about keeping things friendly for the air band.

 

[merlin]

Most of my applications call for band pass filters. I usually make them up myself or get them from K&L microwave.
FM traps, I tend to buy the cheap stuff off feebay. Many of those I have to tweak by swapping out a chip cap or two.
Quite easy to do. I'll try to post an image of these simple filters, usually 3 pole, as are the Noo Electric.
I am 200 yards from a major repeater/cell site so I had to make an 11 pole filter to reduce the 800 Mhz signals down to still monitor 700/800 Mhz LMR. A railroad repeater half mile away I also have to trap or it wreaks havoc with all my VHF stuff. In the works now is 2, 6 pole filters and LNA for 52 Mhz.

 

[pro92b]

NanoVNA Saver - Part #1

This page would act as a basic user guide, to get this app up and running.

zs1sci.com

The link above shows measurements on the RTL-SDR FM filter, about half-way down the page. His measurements confirm the graph published by the filter designer. I also measured one of my RTL-SDR filters and got the same results (file attached).

If you don't monitor the civil air band the RTL-SDR filter is nearly perfect. The Mini-Circuits filter favors passing the civil air band with less attenuation at the cost of reduced attenuation at the top of the FM broadcast band. Mini-Circuits lists their filter as 88-105 MHz, an indication that the last 3 MHz of the FM band is not as fully attenuated.

 

[vagrant]

@MDScanFan - I have the original Flamingo ( Nooelec ) version with less FM attenuation. I show 34 dB max in the FM BCB although the Mfg. claims >40 dB.

150 MHz = 0.4 dB
155 MHz = 0.38 dB
160 MHz = 0.32 dB

and for those that may be wondering...
118 MHz = 2.5 dB
125 MHz = 1.3 dB
130 MHz = 0.9 dB

-----
I swept my RTL-SDR FM notch and it is pretty good if one does not care about VHF air. I have around 55 dB of a notch for most of the FM BCB as you noted. For the cost it notches rather well. I also took a look at 150 - 160 MHz and it is around 0.5 dB.

 

[MDScanFan]

OP back with an update. I received a second RTL-SDR filter and this one matches expectations. The rejection in the FM band is around 50 dB and the insertion loss at 160 MHz is around 0.6 dB. The first part I received must have had a QC issue with its high insertion loss at VHF high.

For the price this filter should work perfectly for my needs.