Broadcast FM duplicate image at +240mhz

Post information or questions regarding SDRplay products here
usagi
Posts: 6
Joined: Wed Jul 29, 2015 6:53 am

Broadcast FM duplicate image at +240mhz

Post by usagi » Fri Aug 07, 2015 9:10 am

With my SDRPlay I'm getting a 1:1 exact duplicate image of the FM broadcast band at exactly +240mhz offset.

E.g. KUBE FM 93.3mhz shows up again at 333.300mhz, KMPS 94.1mhz shows up again at 334.1mhz.

Is anyone else seeing this?

I don't get it with an RTL dongle with the exact same antenna.
Last edited by usagi on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

Tech_Support
Posts: 499
Joined: Mon Jun 01, 2015 7:00 pm

Re: Broadcast FM duplicate image at +240mhz

Post by Tech_Support » Fri Aug 07, 2015 11:48 am

Hi Usagi
This results from the architecture that the RSP uses to cover the range 250 MHz to 380 MHz. Across this frequency range, the signal is first down converted using a 120 MHz LO, so a signal at 333.3 MHz will be mixed down to an IF of 213.3 MHz. The synthesizer is than set to 213.3 MHz and the signal is mixed to I/Q baseband using a second quadrature conversion process. There is a high pass filter at the RF input, but the image frequency for this first mixing process will be at IF - LO = 213.3 MHz - 120 MHz = 93.3 MHz
We specifically picked a frequency plan to place these images at frequencies away from where we felt people would have signals of interest (away from the amateur bands). The RF filter does provide some rejection of the FM images, but if this band is important to us and you and you need more rejection, we recommend using either a FM trap at the antenna or a antenna more highly matched to the band of interest.
These kinds of artefacts are commonplace with dual conversion heterodyne receiver architectures and we plan to put out a white paper, explain this effect in more detail.
I should stress that outside of this and the HF band, you will not see such artefacts as the architecture used is pure homodyne (direct conversion) which does not suffer from these kind of image responses. So in the range 70 MHz - 250 MHz and 430 MHz - 2 GHz, these artefacts will not be present.

Sincerely

Tech_Support
Last edited by Tech_Support on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

usagi
Posts: 6
Joined: Wed Jul 29, 2015 6:53 am

Re: Broadcast FM duplicate image at +240mhz

Post by usagi » Fri Aug 07, 2015 11:58 am

should I also see mirror images around the tuned frequency? around 137mhz receiving NOAA 19 APT, I noticed mirror image below the center freq.
Last edited by usagi on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

Tech_Support
Posts: 499
Joined: Mon Jun 01, 2015 7:00 pm

Re: Broadcast FM duplicate image at +240mhz

Post by Tech_Support » Fri Aug 07, 2015 1:33 pm

Hi again Usagi,
A homodyne receiver will always have quadrature phase and amplitude imbalances. S/W packages such as HDSDR and SDR Sharp contain facilities to correct these imbalances and effectively suppress these images. This will not work for the images that you were seeing at 333 MHz.

Sincerely

Tech_Support
Last edited by Tech_Support on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

User avatar
mikeladd
Posts: 191
Joined: Wed Jul 08, 2015 3:18 am

Re: Broadcast FM duplicate image at +240mhz

Post by mikeladd » Sun Aug 09, 2015 3:08 am

Just my 2 cents. You might want to invest in a par HPF and or a FM Notch (fm trap) it will solve most if not all of you image/intermod issues. I use both and have not had any issues since...

You can get a radio shack fm trap on ebay for $3.00-$5.00 if one does not work then run 2 of them back to back. The insertion loss will not be much of and issues nor will the 75/50 ohm miss match from the trap to the play.

Mike
Last edited by mikeladd on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

kevtheskin
Posts: 7
Joined: Tue Mar 31, 2015 4:17 pm

Re: Broadcast FM duplicate image at +240mhz

Post by kevtheskin » Sat Aug 15, 2015 11:45 am

Hi there.
I am also getting bad imaging . Edinburgh Forth fm on 97.3 is imaged on 385.3125. For me personaly it seems like a waste of time opening up these frequencies it they are affected as bad as this with harmonics?

Cheers Kev
ps still the best sdr I own :) ;)
Last edited by kevtheskin on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

Tech_Support
Posts: 499
Joined: Mon Jun 01, 2015 7:00 pm

Re: Broadcast FM duplicate image at +240mhz

Post by Tech_Support » Sat Aug 15, 2015 12:57 pm

Dear Kev,
The RSP uses a block converter to cover two regions of spectrum: For HF (100 KHz – 60 MHz), it works as a block up-converter and for the lower end of the UHF band (250 MHz-430 MHz), it works as a block down converter. The chipset was never intended to used in this second band, but when we developed the RSP, we found that we could ‘re-use’ the HF block up-converter as a down-converter to provide coverage for the majority of this band. The original API had a fixed 1st LO of 120 MHz for the block converter, which meant that we could not guarantee full coverage from 250 MHz to 430 MHz. Depending upon the skew of the chipset, you might get full coverage, but you might not. All we could guarantee was 250 MHz to 380 MHz and at 430 MHz, we switched to a conventional single conversion to I/Q architecture that is used to cover all of the other bands.

With the new API, we came up with the idea of ‘tricking’ the chipset into making other frequencies available for the first LO. Specifically, we could select 144 MHz or 168 MHz. There are a lot implications to changing the frequency of the 1st LO (in terms of the calibration of various functions on chip), and so it took a long time working with Mirics to get this all working correctly. Whichever LO you use, there will always still be a coverage gap, but by dynamically changing the 1st IF frequency, it is possible to achieve contiguous coverage and close the gap that was present between 380 MHz and 430 MHz when using the 120 MHz 1st LO.

The problem with block converters is that they will always have some spurious responses. This occurs because the mixers are ‘switching mixers’ which means that the Local Oscillators are effectively square waves. They therefore contain odd order harmonics that descend with a 1/X roll off characteristic. Spurious responses occur when signals from other bands ‘leak’ to the input of the block converter and mix with harmonics of the 1st and LO to produce a signal at the first IF that is either equal to the wanted IF frequency or is equal to an odd order harmonic of the wanted 1st IF frequency.

The RF filters mitigate these spurious responses to some degree, but if there are very strong signals at VHF (as is often the case with FM signals), there is a good chance that some of these signals will get through the front end filter and create a spurious response.

There are two possible ways to deal with this. Either add extra filtering (as has already been suggested) and an FM trap is a good way to go, or change the first LO to move these spurious responses away from the band of interest. At the moment, it is not possible to change the 1st LO frequency in SDR Console, but we hope to work with Simon Brown to get this new feature implemented. If you are using either HDSDR or SDR Sharp, try toggling between a 144 MHz and 168 MHz 1st LO under the advanced tab of the ExtIO window and you should see these spurious responses move by 24 MHz.

You will only get these kinds of spurious response in these two bands as it is a consequence of the use of a block frequency converter. In all other bands, the single conversion architecture is not susceptible to these effects.

We plan to release a ‘white paper’ in the next few weeks describing the architecture of the RSP in more detail so that people can better understand how such responses can arise and help devise strategies to effectively mitigate them.

Sincerely

SDRplay Tech_Support
Last edited by Tech_Support on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

usagi
Posts: 6
Joined: Wed Jul 29, 2015 6:53 am

Re: Broadcast FM duplicate image at +240mhz

Post by usagi » Wed Aug 19, 2015 8:04 am

what specific fm trap is recommended?
Last edited by usagi on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

F1BJB
Posts: 118
Joined: Wed Jul 15, 2015 4:24 pm
Location: Beauvais France

Re: Broadcast FM duplicate image at +240mhz

Post by F1BJB » Wed Aug 19, 2015 9:21 pm

Hi
As filter I had good results using a wall mounted TV/FM/SAT socket.
They are lowcost and easy to find .
My main problem was second order intermodulation products from the FM band falling in the DAB band.
I used an RTL dongle at the time.
HTH
Last edited by F1BJB on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

SPole
Posts: 2
Joined: Fri Aug 14, 2015 11:37 pm

Re: Broadcast FM duplicate image at +240mhz

Post by SPole » Tue Aug 25, 2015 10:38 pm

mikexeno wrote:Just my 2 cents. You might want to invest in a par HPF and or a FM Notch (fm trap) it will solve most if not all of you image/intermod issues. I use both and have not had any issues since...

You can get a radio shack fm trap on ebay for $3.00-$5.00 if one does not work then run 2 of them back to back. The insertion loss will not be much of and issues nor will the 75/50 ohm miss match from the trap to the play.

Mike
I went to radioshack and picked up a cheap fm trap as you recommended and it does contain most of the problem but it still see ghosts. It's a cheap fix and it works pretty well. thanks.
Last edited by SPole on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

Post Reply