1. SDR-Radio | Main
  2. Topics

Bandwidth-/ADC-Settings with RX-888 MKII

Dirk
 

Perhaps just doing sth. wrong in SDRC ...
I tried to show the whole FM broadcast band on the screen.
I thought that would be quite simple with its 21 MHz bandwidth.
I used 64 Msps and a bandwidth of 32 MHz and got the attached screen.
(Also tried 80 Msps and 48 MHz with same result ...)
The marker is set to a station at 104.7 MHz and it is clearly audible.
My question:
1. Why can I see only about 10 MHz of the whole FM spectrum (87.5-108.5 MHz)?
2. Why is there a sinus-like peak above this spectrum?
3. Why can I hear a station on 104.7 MHz (not in the shown 10 MHz spectrum)?

Perhaps give me a push into the right direction...?
Tony_AD0VC
 


I believe the RX888 is limited to 8 Mhz BW on VHF. You will only get the 32Mhz on HF.

Tony

From: main@SDR-Radio.groups.io <main@SDR-Radio.groups.io> on behalf of Dirk <dottensm@...>
Sent: Sunday, October 9, 2022 6:24 AM
To: main@SDR-Radio.groups.io <main@SDR-Radio.groups.io>
Subject: [SDR-Radio] Bandwidth-/ADC-Settings with RX-888 MKII
 
Perhaps just doing sth. wrong in SDRC ...
I tried to show the whole FM broadcast band on the screen.
I thought that would be quite simple with its 21 MHz bandwidth.
I used 64 Msps and a bandwidth of 32 MHz and got the attached screen.
(Also tried 80 Msps and 48 MHz with same result ...)
The marker is set to a station at 104.7 MHz and it is clearly audible.
My question:
1. Why can I see only about 10 MHz of the whole FM spectrum (87.5-108.5 MHz)?
2. Why is there a sinus-like peak above this spectrum?
3. Why can I hear a station on 104.7 MHz (not in the shown 10 MHz spectrum)?

Perhaps give me a push into the right direction...?
I
Simon Brown
 

Correct.

From: main@SDR-Radio.groups.io <main@SDR-Radio.groups.io> on behalf of Tony_AD0VC via groups.io <canthony15@...>
Sent: 09 October 2022 13:41
To: main@SDR-Radio.groups.io <main@SDR-Radio.groups.io>
Subject: Re: [SDR-Radio] Bandwidth-/ADC-Settings with RX-888 MKII
 

I believe the RX888 is limited to 8 Mhz BW on VHF. You will only get the 32Mhz on HF.

Tony

From: main@SDR-Radio.groups.io <main@SDR-Radio.groups.io> on behalf of Dirk <dottensm@...>
Sent: Sunday, October 9, 2022 6:24 AM
To: main@SDR-Radio.groups.io <main@SDR-Radio.groups.io>
Subject: [SDR-Radio] Bandwidth-/ADC-Settings with RX-888 MKII
 
Perhaps just doing sth. wrong in SDRC ...
I tried to show the whole FM broadcast band on the screen.
I thought that would be quite simple with its 21 MHz bandwidth.
I used 64 Msps and a bandwidth of 32 MHz and got the attached screen.
(Also tried 80 Msps and 48 MHz with same result ...)
The marker is set to a station at 104.7 MHz and it is clearly audible.
My question:
1. Why can I see only about 10 MHz of the whole FM spectrum (87.5-108.5 MHz)?
2. Why is there a sinus-like peak above this spectrum?
3. Why can I hear a station on 104.7 MHz (not in the shown 10 MHz spectrum)?

Perhaps give me a push into the right direction...?
I

--
- + - + -
Please use https://forum.sdr-radio.com:4499/ when posting questions or problems.
Dirk
 
Edited

Thanks.
In principle I know that, but the hardening of arteries  ...
And to 2. and 3.?
How can I receive out of the 8 MHz shown in the spectrum and what is the sinus-shaped curve?
Simon Brown
 

What you see is the response from the tuner chip, that's it.

From: main@SDR-Radio.groups.io <main@SDR-Radio.groups.io> on behalf of Dirk via groups.io <dottensm@...>
Sent: 09 October 2022 14:35
To: main@SDR-Radio.groups.io <main@SDR-Radio.groups.io>
Subject: Re: [SDR-Radio] Bandwidth-/ADC-Settings with RX-888 MKII
 

[Edited Message Follows]

Thanks.
In principle I know that, but the hardening of arteries  ...
And to 2. and 3.?
How can I receive out of the 8 MHz shown in the spectrum and what is the sinus-shaped curve?

--
- + - + -
Please use https://forum.sdr-radio.com:4499/ when posting questions or problems.
Max
 

On Sun, Oct 9, 2022 at 01:24 PM, Dirk wrote:
1. Why can I see only about 10 MHz of the whole FM spectrum (87.5-108.5 MHz)?
2. Why is there a sinus-like peak above this spectrum?
3. Why can I hear a station on 104.7 MHz (not in the shown 10 MHz spectrum)?
See here for block diagram of BBRF103 which is the original SDR designed by Oscar Steila, IK1XPV on which the RX666 and RX888 are based:

https://www.rtl-sdr.com/the-breadboard-rf103-a-homemade-16-bit-0-1800-mhz-software-defined-radio/comment-page-1/

You can see it's only direct sampling on the HF bands up to about 30 MHz. For frequencies above that then there is another path via the RTL Dongle tuner chip fed from second SMA socket which forms the limitation of the design for VHF/UHF and up, as this chip (on BBRF103 it's the Rafael Micro R820T2) only allows for 10MHz bandwidth. Actually I think it's usually 8 MHz on most dongles but somehow on the BBRF103/RX666/RX888 it's stretched to 10MHz. This explains why only 10 MHz bandwidth above 30 MHz. Actually I think with the RX888 the direct sampling bandwidth is stretched to 60MHz for the higher available sampling rate. Others know better than I. I don't have an RX888 (yet!).

I seem to remember that Justin Peng, originator of RX666/888 mentioned plans to produce a down-converter to make receiver true direct sampling at VHF/UHF but I don't think this has materialised yet, or maybe never will?

As Simon says, roll off at the ends of the sample bandwidth window is entirely normal with the RTL-SDR chip. Also I see this in the SDR-Play receivers. I don't see it on my Hermes Lite 2 direct sampling receiver. It's an almost perfect steep cutoff at either end of the sapling window. I would wager it's maybe also sharp cutoff when using the HF direct sampling part of the RX888?

How on earth you are hearing the station on 104.7 outside the sampling window I have no idea. Odd! However, the S-Meter is reading -140dBm which is extremely weak to put it mildly. Well I'd expect that to be about noise level with no antenna plugged in! Maybe it is possible to just about hear some stuff outside of the set bandwidth? At that level I am amazed it's either audible or that you can get RDS readout.

Max
jdow
 

The 104.7 MHz issue is almost certainly a simple aliasing issue. Filters are not perfect. All they do is reduce or accentuate some frequencies over other frequencies. The R820T chip family has an interesting tunable filter design. However, it is not perfect. The FM band generally contains very strong signals that can punch through the clever but still limited filters in the R820T family of tuners. Band reject filters are sold to deal with this in particularly bad cases.

AirSpy and AirSpy R2 use the same tuners and suffer the same issues. They work up to a 10 MHz sample rate. The reasonably anti-aliased bandwidth is about 8 MHz, the same as with the RX888 family.

The lack of flatness on the VHF passband is simply an artifact of the tuner design.

{^_^}

On 20221009 12:56:31, Max wrote:

On Sun, Oct 9, 2022 at 01:24 PM, Dirk wrote:
1. Why can I see only about 10 MHz of the whole FM spectrum (87.5-108.5 MHz)?
2. Why is there a sinus-like peak above this spectrum?
3. Why can I hear a station on 104.7 MHz (not in the shown 10 MHz spectrum)?
See here for block diagram of BBRF103 which is the original SDR designed by Oscar Steila, IK1XPV on which the RX666 and RX888 are based:

https://www.rtl-sdr.com/the-breadboard-rf103-a-homemade-16-bit-0-1800-mhz-software-defined-radio/comment-page-1/

You can see it's only direct sampling on the HF bands up to about 30 MHz. For frequencies above that then there is another path via the RTL Dongle tuner chip fed from second SMA socket which forms the limitation of the design for VHF/UHF and up, as this chip (on BBRF103 it's the Rafael Micro R820T2) only allows for 10MHz bandwidth. Actually I think it's usually 8 MHz on most dongles but somehow on the BBRF103/RX666/RX888 it's stretched to 10MHz. This explains why only 10 MHz bandwidth above 30 MHz. Actually I think with the RX888 the direct sampling bandwidth is stretched to 60MHz for the higher available sampling rate. Others know better than I. I don't have an RX888 (yet!).

I seem to remember that Justin Peng, originator of RX666/888 mentioned plans to produce a down-converter to make receiver true direct sampling at VHF/UHF but I don't think this has materialised yet, or maybe never will?

As Simon says, roll off at the ends of the sample bandwidth window is entirely normal with the RTL-SDR chip. Also I see this in the SDR-Play receivers. I don't see it on my Hermes Lite 2 direct sampling receiver. It's an almost perfect steep cutoff at either end of the sapling window. I would wager it's maybe also sharp cutoff when using the HF direct sampling part of the RX888?

How on earth you are hearing the station on 104.7 outside the sampling window I have no idea. Odd! However, the S-Meter is reading -140dBm which is extremely weak to put it mildly. Well I'd expect that to be about noise level with no antenna plugged in! Maybe it is possible to just about hear some stuff outside of the set bandwidth? At that level I am amazed it's either audible or that you can get RDS readout.

Max

Dirk
 
Edited

On Mon, Oct 10, 2022 at 04:29 AM, jdow wrote:

The 104.7 MHz issue is almost certainly a simple aliasing issue. Filters are not perfect. All they do is reduce or accentuate some frequencies over other frequencies. The R820T chip family has an interesting tunable filter design. However, it is not perfect. The FM band generally contains very strong signals that can punch through the clever but still limited filters in the R820T family of tuners. Band reject filters are sold to deal with this in particularly bad cases.

AirSpy and AirSpy R2 use the same tuners and suffer the same issues. They work up to a 10 MHz sample rate. The reasonably anti-aliased bandwidth is about 8 MHz, the same as with the RX888 family.

The lack of flatness on the VHF passband is simply an artifact of the tuner design.

Yes, I think you're right.

What I found out:
There is no 104.7 MHz carrier with the culture program "WDR 3" (as shown in my picture). But there are 2 very strong carriers with this program at 92.7 MHz and at 97.0 MHz (this one is strongest). How can these carriers be mirrored from within the selected band into the passband?
So all the same,- should be an aliasing-effect...

1 - 8 of 8