Direct sampling interference vs upconverter

Need help installing or figuring out something about your SDR? Ask here.
Post Reply
artm
Posts: 17
Joined: Mon Jun 24, 2019 3:51 am

Direct sampling interference vs upconverter

Post by artm » Sat Jul 06, 2019 5:19 am

Wondering what cause this. Rtl-sdr V3 On direct sampling mode and feeding it 6Mhz xtal oscillator there is lot spikes around and tuning causes these spikes move wildly. Signals are well below overload and decreasing level (w/ext. attenuator) may cause spikes grow a bit as seen on picture. 14Mhz LP filter won's change much, leves same.
Same happens wit other signal sources.
6Mhz_direct2.JPG
6Mhz_direct2.JPG (41.27 KiB) Viewed 991 times
When using (125Mhz) up-converter, no problem. Noise around carrier comes from converters oscillator.
6Mhz_mixer.JPG
6Mhz_mixer.JPG (46.87 KiB) Viewed 991 times

radiolistener
Posts: 42
Joined: Thu Jan 25, 2018 10:48 pm

Re: Direct sampling interference vs upconverter

Post by radiolistener » Sat Jul 06, 2019 8:31 am

I don't have such spurs around 6 MHz on directsampling mode with open input.

If it appears when you connect high level 6 MHz source, then this is intermodulation distortions.
This is known issue of any receiver. Smaller level of these spurs means better IMD performance of the receiver.

If it is missing with upconverter it means better IMD performance.
When you're using upconverter, it works through 820T2 tuner.
When you're using direct sampling it works directly.
Direct sampling mode uses different RF path from antenna.
So, may be there is some component on that path, which leads to non-linear distortions with high level signal on the input. For example diode or amplifier may cause such issue.
Attachments
rtlsdrv3-direct-sampling-6MHz.png
RTLSDRv3 direct sampling mode 6 MHz
rtlsdrv3-direct-sampling-6MHz.png (373.38 KiB) Viewed 973 times

artm
Posts: 17
Joined: Mon Jun 24, 2019 3:51 am

Re: Direct sampling interference vs upconverter

Post by artm » Sat Jul 06, 2019 11:07 am

radiolistener wrote:
Sat Jul 06, 2019 8:31 am
I don't have such spurs around 6 MHz on directsampling mode with open input.

If it appears when you connect high level 6 MHz source, then this is intermodulation distortions.
This is known issue of any receiver. Smaller level of these spurs means better IMD performance of the receiver.
I have clean screen without signal, (but i have abt 40db DC-spike). About -90dBm is enough to start it. They seems to be mixing products.

In picture signal just below overloading. More level more interference but no signal level increase.
DC-spike rigth at 6.28Mhz, DC-removal will remove it but no affect interference.
6Mhz near overl.JPG
6Mhz near overl.JPG (64.91 KiB) Viewed 955 times

artm
Posts: 17
Joined: Mon Jun 24, 2019 3:51 am

Re: Direct sampling interference vs upconverter

Post by artm » Sat Jul 06, 2019 11:58 am

Addition info, when LO is same as oscillator, no wide spurs (upper in picture). Within 1Hz and noise floor pumping.
2kHz offset and spurs are wider (bottom in picture). More offset more wider spurs.

If using other freq, same happens any frequency if sdr LO is same as input signal.. :shock:
test1.JPG
test1.JPG (45.53 KiB) Viewed 938 times

radiolistener
Posts: 42
Joined: Thu Jan 25, 2018 10:48 pm

Re: Direct sampling interference vs upconverter

Post by radiolistener » Sat Jul 06, 2019 4:27 pm

artm wrote:
Sat Jul 06, 2019 11:07 am
In picture signal just below overloading. More level more interference but no signal level increase.
These spurs are receiver IMD distortions. RTLSDR has IMD about 50-60 dB, so these spurs will appears 50-60 dB below strong signal.
artm wrote:
Sat Jul 06, 2019 11:07 am
DC-spike rigth at 6.28Mhz, DC-removal will remove it but no affect interference.
6Mhz near over
I also have DC-spike 30 dB (measured at 500 Hz bandwidth), on my picture I enabled IIR filter with bandwidth 1 Hz in order to remove it.

Username
Posts: 433
Joined: Sun Oct 09, 2016 7:27 am

Re: Direct sampling interference vs upconverter

Post by Username » Sat Jul 06, 2019 10:09 pm

Hi
I use an Upcongverter with LNA from a friend. The quality is good and the results to.
I will post a Picture tomorrow. I would say a LNA is reqiered.

artm
Posts: 17
Joined: Mon Jun 24, 2019 3:51 am

Re: Direct sampling interference vs upconverter

Post by artm » Sun Jul 07, 2019 4:14 am

radiolistener wrote:
Sat Jul 06, 2019 4:27 pm

These spurs are receiver IMD distortions. RTLSDR has IMD about 50-60 dB, so these spurs will appears 50-60 dB below strong signal.
Well, I found it can tolerate only less than 40dB in direct sampling mode. Signal level abt. -85dBm at bottom.
DC spike is quite high.
imd_direct.JPG
imd_direct.JPG (25.23 KiB) Viewed 844 times
In up-converter mode, abt 70dB IMD. Signals from 6Mhz xtal osc and A&A DDS combined, level abt. -65dBm.
rtl_mixer_imd.jpg
rtl_mixer_imd.jpg (26.45 KiB) Viewed 844 times

artm
Posts: 17
Joined: Mon Jun 24, 2019 3:51 am

Re: Direct sampling interference vs upconverter

Post by artm » Mon Jul 08, 2019 5:58 am

I thought is it actually IMD? It behaves strange when its so wide (whole band), occurs so low level, way below blocking level. Spike at other side of carrier is much lower. This not seen on normal IF-mode. Some way it is mixing product with "dc-spike"...
I watch videos on youtube and hard to say, they look better.

radiolistener
Posts: 42
Joined: Thu Jan 25, 2018 10:48 pm

Re: Direct sampling interference vs upconverter

Post by radiolistener » Tue Jul 30, 2019 10:56 am

I also have something at LO=6001.05 kHz, but there are just two spurs, it is different from your case. I'm using 2.4 MHz sample rate.

What sample rate you're using? What is your dongle?
Attachments
6001kHz.png
spurs at 6001 kHz
6001kHz.png (604.83 KiB) Viewed 250 times

artm
Posts: 17
Joined: Mon Jun 24, 2019 3:51 am

Re: Direct sampling interference vs upconverter

Post by artm » Tue Jul 30, 2019 1:10 pm

I use always 2.4Mhz, dongle is V3. I bought another dongle and it behave just same. Both from rtl-sdr-blog ebay store and both date stamp 2018-9-13 on pcb.

I have only dc-spike (6.001.05Mhz)and -35dB both side. Same at 12 and 24Mhz, but more spikes around.

Post Reply