Evaluating RF filters?

Information and discussions on Spectrum Analysis
Mike2459
Posts: 318
Joined: Mon Jul 24, 2017 8:24 pm

Re: Evaluating RF filters?

Post by Mike2459 » Sun Jul 14, 2019 5:17 pm

Here is a scan of my homebuilt 3 pole (not 2 pole) elliptic filter. Had to set the RSP2 to maximum RF gain and disable IF agc and manually set it to a low level to keep receiver noise from masking the low level response while trying to avoid ADC OL. Used a program called RF Filter Designer to determine component values. All of the capacitors are SMD's, the coils are wound on small toroids. Set up my FG504 to start the sweep at 100KHZ and end around 4 MHZ.
HIZ 3p LP filter.jpg
HIZ 3p LP filter.jpg (69.29 KiB) Viewed 2560 times

Reason: No reason

Mike2459
Posts: 318
Joined: Mon Jul 24, 2017 8:24 pm

Re: Evaluating RF filters?

Post by Mike2459 » Sun Jul 14, 2019 6:10 pm

Here is a scan of the filter in use. My local RF environment is no way near as challenging as that of LA!
The real world.JPG
The real world.JPG (95.25 KiB) Viewed 2551 times

Reason: No reason

sdrom33
Posts: 128
Joined: Sun Dec 09, 2018 4:38 pm

Re: Evaluating RF filters?

Post by sdrom33 » Sun Jul 14, 2019 9:25 pm

Mike, it seems to me that your filter is low pass filter, e.g. suppresses MW & higher band to let you get the LW band. I am under the impression, .maybe I am wrong, that Franko wants a high pass filter, to suppress MW in order to listen to HF from 1.9 MHz & higher, as shown by glovisol.

Reason: No reason

franko
Posts: 55
Joined: Mon Aug 28, 2017 9:33 pm

Re: Evaluating RF filters?

Post by franko » Mon Jul 15, 2019 12:07 am

glovisol wrote:Franko, this is what you need to get rid of MW interference, it is effective and easy & inexpensive to build, just look at the thread.

https://www.sdrplay.com/community/viewt ... 973#p12847
Very interesting thread, thanks for the link. It sounds like it was a fun project.

My challenge is that I'd like to build a band-reject filter that will pass frequencies from 0 to about 520 kHz, and also pass frequencies above about 1700 kHz, but strongly attenuate the band from ~520 to 1700 kHz. The reason for this is that I have a single antenna connected to a high-quality four-way splitter, which feeds four SDRs monitoring bands from 2200m to 6m.

One idea would be to build a low-pass filter and also a high-pass flter, and place them in parallel so that they function together. But I've been warned that combining multiple filters runs the risk of having them interact in undesirable ways.

The other alternative would be to use a separate RF filter for each SDR, placed between it and the splitter. This would be okay, except that one of my SDRs is configured to monitor both VLF/LF and HF frequencies.
Mike2459 wrote:Here is a scan of my homebuilt 3 pole (not 2 pole) elliptic filter. Had to set the RSP2 to maximum RF gain and disable IF agc and manually set it to a low level to keep receiver noise from masking the low level response while trying to avoid ADC OL. Used a program called RF Filter Designer to determine component values. All of the capacitors are SMD's, the coils are wound on small toroids. Set up my FG504 to start the sweep at 100KHZ and end around 4 MHZ.
Looks like very nice results.

I tried a web search for the RF Filter Designer software but didn't find it right away. Do you know if it's currently available? In addition to high- or low-pass filters, does it also deal with band-pass or band-reject designs?

Also, I checked the manual for my Siglent SDG2042X waveform generator, and it will do sweeps. How do you configure SDRUno so that you see the signal level for all the frequencies across the visible bandwidth after the generator completes its sweep? Is there a selection in SDRUno to freeze the highest signal level at each frequency? Or is the sweep so fast that it ends up being essentially instantaneous?

Reason: No reason

Mike2459
Posts: 318
Joined: Mon Jul 24, 2017 8:24 pm

Re: Evaluating RF filters?

Post by Mike2459 » Mon Jul 15, 2019 12:24 am

It's RF Filter Design and not Designer. Sorry about that.
As to the sweep. My unit has an adjustable sweep from .1ms to >10Sec. For this I used .1ms but it's not critical for a spectrum display as long as it is not too slow. Just adjust it for the best display. Too view it on an O'scope the timing becomes more critical.

Reason: No reason

Mike2459
Posts: 318
Joined: Mon Jul 24, 2017 8:24 pm

Re: Evaluating RF filters?

Post by Mike2459 » Mon Jul 15, 2019 1:11 am

Here is the link to a description of RF Filter Design. At the bottom of that page is a link to the download section of the web site.

http://www.iowahills.com/9RFFiltersPage.html

Mike

Reason: No reason

Mike2459
Posts: 318
Joined: Mon Jul 24, 2017 8:24 pm

Re: Evaluating RF filters?

Post by Mike2459 » Mon Jul 15, 2019 2:18 am

I talked to one guy who designs and sells RF filters, and he said he looked into this but gave up because the filter design was going to be so complex.
After plugging some numbers into RF Filter Designer for a notch filter for the MW Band, I can see what he was referring to. Not a lot of wiggle room.

Mike

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: Evaluating RF filters?

Post by glovisol » Mon Jul 15, 2019 6:40 am

Hi Franko,

Having long ago prepared Excel spreadsheets for quick calculation of any type filters, once you plug in the available filter prototype tabulated data, I could quickly check your specification and the result is shown below. This filter is complex, but with easily realizable components, provided you stay on a Zo = 500 Ohm. You can then translate Zin & Zout to any impedance you need with binocular core ferrite transformers: this is a big advantage if you use long wire antennas and/or the HI Z input of the RSP. If you need less than 40/45 dB attenuation in the stopband, I can compute a simpler filter, perhaps a Cauer N= 3 prototype.

I am wary of automatic calculation software, because then the user tends to calculate filters without really understanding what he is doing.

Finally, if I may venture advice, for filter development and measurement the easiest and most practical setup is a cheap noise generator, such as the BG7TBL and the new Spectrum Analyser software, applicable to RSP processors. You can find many examples in my filter threads. For a lively discussion on Noise Generators, see:

https://www.sdrplay.com/community/viewt ... +generator

Hope this helps.

glovisol

MW band elimination notch filter.jpg
MW band elimination notch filter.jpg (62.54 KiB) Viewed 2476 times

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: Evaluating RF filters?

Post by glovisol » Mon Jul 15, 2019 3:13 pm

ERRATA CORRIGE: in the schematic I omitted to put the value of the capacitor in series with the 336 uH inductor: value is 215 pf. Sorry for that!

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: Evaluating RF filters?

Post by glovisol » Mon Jul 15, 2019 4:09 pm

This filter is simpler & easier to make, it would be good as a first experimental unit. It is a Chebyshev N=5, r=0.5 dB, Zin=Zout=500 Ohm as the previous Cauer filter.
Chebyshev MW band elimination notch filter.jpg
Chebyshev MW band elimination notch filter.jpg (57.26 KiB) Viewed 2398 times

Reason: No reason

Post Reply