MSi.SDR sound crackle pulses and spurs

Talk about other SDR products like the FunCube, HackRF, BladeRF etc.
artm
Posts: 16
Joined: Mon Jun 24, 2019 3:51 am

Re: MSi.SDR sound crackle pulses and spurs

Post by artm » Mon Jul 01, 2019 3:00 am

Sure. There is markings. Voltage regulator is on top.
DSCN2112.JPG
DSCN2112.JPG (881.94 KiB) Viewed 604 times

radiolistener
Posts: 28
Joined: Thu Jan 25, 2018 10:48 pm

Re: MSi.SDR sound crackle pulses and spurs

Post by radiolistener » Mon Jul 01, 2019 3:34 am

Thanks. Also, can you please check, how many positions your MSi.SDR has for gain slider in SDRuno?
My one has just 4 positions with strange values.

Interesting, it looks that my PCB is more old (the year is not readable, but it looks like 2018.09 or something like that), it has more vias than your new PCB. May be these vias leads to some kind of instability? Strange, technicaly it should be better with more vias...

So, this is not flash memory, this is controller... But pin layout looks very similar to flash memory with two SPI pins wired to external connector. May be this is flash memory emulator implemented within STC 15W100 microcontroller? :)
According to datasheet this MCU has 0.5K flash and 128 bytes memory. Working voltage is 2.5...5.5V

For what reason they are scratched off components marking? :evil:


L00B - 3.3 STD LP2980IM5X: Ultra Low-Dropout Regulator 3.3V

YJ28 - SGM2019-2.8YN5/TR: Low Power, Low Dropout, 250mA, RF - Linear Regulator - 2.8V

A7W - BAV99

S79 - looks like RF switch AS179-92, working frequency 20 MHz - 4 GHz.

artm
Posts: 16
Joined: Mon Jun 24, 2019 3:51 am

Re: MSi.SDR sound crackle pulses and spurs

Post by artm » Mon Jul 01, 2019 5:33 am

I have continuous gain slider, it's older EXTio version. The newest has these 4.

Now i found these crackles.. When frequency was exactly f.o. 6Mhz, was not all zeros before... :lol:
Does something with that oscillation too.
I have even more tests. On (hdsdr) zero-if almost all sample combinations was some oscillation, like picture below 8Mhz/8Mhz seen spur bump oscillation. At 10Mhz it seems fly over screen really fast. Low-if mode, no oscillation but any change for gain moves it somewhere and then drifting.

With SDR# i cannot reproduce these at all. There was max 8Mhz so at 10Mhz noise level jumping can't tested. All steady and spur bump almost steady. No oscillations!
Gqrx linux oscillation was similar as hdsdr but bit less fluctuation.
msi_8Mhz_8Mhz.JPG
msi_8Mhz_8Mhz.JPG (127.24 KiB) Viewed 583 times

artm
Posts: 16
Joined: Mon Jun 24, 2019 3:51 am

Re: MSi.SDR sound crackle pulses and spurs

Post by artm » Mon Jul 01, 2019 1:00 pm

I think i found the cause for that interference.. or not..
When i listening msi's input with rtl v3 and looking it's vco leak. Found better in higher freq like 600Mhz.
In picture, signals (yellow) are msi's own spur left and vco right. When tuning (red arrow direction) closer to spur frequency, mixing product occur. When it's zero, signal pumping. This was strongest on 600Mhz, 360Mhz and xtal 24Mhz and harmonics.
If tuning Lo and tune to 600.001, then strong 1khz interference was audible.
Using low-IF mode when tuning vco, same interference is there also but out of IF passband so not audible.

Crackle sound, maybe some spurs are internal USB/digital signal leak? Then they mix/modulate with vco signal.
Why 10Mhz mode is worst, I don't know.
Capture.JPG
Capture.JPG (91.75 KiB) Viewed 551 times

radiolistener
Posts: 28
Joined: Thu Jan 25, 2018 10:48 pm

Re: MSi.SDR sound crackle pulses and spurs

Post by radiolistener » Mon Jul 01, 2019 4:15 pm

artm wrote:
Mon Jul 01, 2019 1:00 pm
When tuning (red arrow direction) closer to spur frequency, mixing product occur. When it's zero, signal pumping.
yes, I also found that when you change LO, these spurs are moving with different speed. And when they are merged together in a single spur, sound crackle and pulses appears. It happens when LO is tuned at 1'000'000, 2'000'000, 3'000'000, 4'000'000, 5'000'000, 6'000'000 Hz, etc... It also present near that frequencies for example 6'000'001, 6'000'002, etc, but pulsing period is different.

I have no idea how it is possible, that some spurs can lead to random pulses across entire bandwidth.
artm wrote:
Mon Jul 01, 2019 1:00 pm
Crackle sound, maybe some spurs are internal USB/digital signal leak? Then they mix/modulate with vco signal.
I tried to connect external ultra low noise lab power supply with ferrite bead filters. Tried to use usb cable with RFI filter and tried to add ferrite bead on it. Tried to add capacitors on power lines at pcb. I tried to connect it to different computers and notebook. But it doesn't affect this issue at all. These pulses are still there with the same level.

The worst thing is that these pulses appears across full bandwidth, so there is no way to get rid of it by tuning. Even when you change LO these pulses didn't disappears, their frequency just going to more high frequency (up to 100 kHz and more), so it is difficult to notice it on some LO frequencies, but they are still there and you can find it on AF spectrum waterfall with 192 kHz bandwidth. So, this defect can affect digital signal reception even when you don't hear it.

radiolistener
Posts: 28
Joined: Thu Jan 25, 2018 10:48 pm

Re: MSi.SDR sound crackle pulses and spurs

Post by radiolistener » Mon Jul 01, 2019 4:33 pm

artm wrote:
Mon Jul 01, 2019 1:00 pm
If tuning Lo and tune to 600.001, then strong 1khz interference was audible.
in my case there is no 1 kHz interference when LO=600'001'000.
Tried LO=600'000'001, and there is also no 1 kHz interference, just 1-2 Hz pulses, similar to these as on LO=6'000'000.

Tried to connect RTLSDRv3 and MSiSDR connectors with coax cable and I also see fixed spur at 600 MHz and VCO spur which is moving when I change LO. Hm... What is the source of that 600 MHz spur? Any idea how to fix it?

Also, found strange noise which appears from MSi.SDR when it's LO is smaller than 60 MHz.
Attachments
LEAK_6.png
LEAK_6.png (337.24 KiB) Viewed 514 times
LEAK_144.png
LEAK_144.png (313.89 KiB) Viewed 514 times

artm
Posts: 16
Joined: Mon Jun 24, 2019 3:51 am

Re: MSi.SDR sound crackle pulses and spurs

Post by artm » Tue Jul 02, 2019 3:38 am

We need to know better these chip's histoy, if they are defectives maybe faulty mixers inside.. etc.
I noticed too usb cable and ferrites wont help, no difference when pcb is in case or not.
radiolistener wrote:
Mon Jul 01, 2019 4:33 pm

in my case there is no 1 kHz interference when LO=600'001'000.
Tried LO=600'000'001, and there is also no 1 kHz interference, just 1-2 Hz pulses, similar to these as on LO=6'000'000.
I used both Lo and Tune as 600.001 zero-if (any sample rate) and AM.
On some freq. there is something digital whining etc.

Here pictures:
zero-if sample rate 1.563, 1khz interference show and many harmonics on audio.
interf.JPG
interf.JPG (102.96 KiB) Viewed 494 times
zero-if sample rate 8Mhz/10Mhz when noise pumping and sound cracling. Sound like something overloading or became unstable.
interf.1.JPG
interf.1.JPG (111.54 KiB) Viewed 494 times
Using Low-IF mode seems to be far more better, less spurs and cleaner signal. Zero-if is more useless.

radiolistener
Posts: 28
Joined: Thu Jan 25, 2018 10:48 pm

Re: MSi.SDR sound crackle pulses and spurs

Post by radiolistener » Tue Jul 02, 2019 4:28 am

artm wrote:
Tue Jul 02, 2019 3:38 am
Using Low-IF mode seems to be far more better, less spurs and cleaner signal. Zero-if is more useless.
no, I don't agree. With low-IF there are a lot of images (mirrors from borders of the Nyquist zones). And most of spurs are present in both modes, you can test it on wideband spur at 30 MHz. According to my tests, zero-IF almost always works better and gives you more control. Probably low-IF mode was added to hide some LO behavior under the hood for those who don't familiar with LO tuning. If you're familiar with SDR and LO tuning, low-IF mode will be completely useless.

It seems like root of cause of these instabilities and pulses is a defect in the chips. So this may be the reason, why these cheap devices appears on the market. Probably someone trying to sell factory rejected chips in such way.

10 MHz defect is not a big deal, because there is anyway 8 MHz IF filter. But this defect with pulses is really annoying issue.

I found a workaround, you can set LO to xxxx103000 and change just xxxx MHz values. With such LO frequencies this pulsing defect is minimized and almost not noticeable. There are a lot of shifting spurs and they can interfere with such LO values at some frequencies. But it's better than nothing :)

In overal my MSi.SDR works worse (due to these defects, spurs) than RTL.SDRv3. But at least it allows to see 8 MHz bandwidth on the waterfall, this is nice. Unfortunately this pulsing defect spoils the whole picture... :( If it were possible to fix it - it would be great.

artm
Posts: 16
Joined: Mon Jun 24, 2019 3:51 am

Re: MSi.SDR sound crackle pulses and spurs

Post by artm » Tue Jul 02, 2019 5:48 am

radiolistener wrote:
Tue Jul 02, 2019 4:28 am

no, I don't agree. With low-IF there are a lot of images (mirrors from borders of the Nyquist zones). And most of spurs are present in both modes, you can test it on wideband spur at 30 MHz. According to my tests, zero-IF almost always works better and gives you more control. Probably low-IF mode was added to hide some LO behavior under the hood for those who don't familiar with LO tuning. If you're familiar with SDR and LO tuning, low-IF mode will be completely useless.
True, images especially with strong signals. On zero-if signal mixed by near spur causes strong spikes around. Noise floor and dc spike higher too on zeroif.
Without spurs and xtal harmonics would be better. I dont have 30Mhz spur at all.
In overal my MSi.SDR works worse (due to these defects, spurs) than RTL.SDRv3. But at least it allows to see 8 MHz bandwidth on the waterfall, this is nice. Unfortunately this pulsing defect spoils the whole picture... :( If it were possible to fix it - it would be great.
Yes v3 is better. With spektrum it's more smooth than msi.sdr with Andrews analyzer.

Guess only solution is buying real RSP.. ? if want wider waterfall..

artm
Posts: 16
Joined: Mon Jun 24, 2019 3:51 am

Re: MSi.SDR sound crackle pulses and spurs

Post by artm » Tue Jul 02, 2019 11:55 am

One possibility that most spurs are 24Mhz and it's harmonic products?
Msi's spektrum looks terrible. Bottom rtl v3 own spurs. Scale is near dBm, not calibrated. 10dB attenuator between dongles.
msi_rtl spurs.jpg
msi_rtl spurs.jpg (52.31 KiB) Viewed 397 times
For comparison, I also test that your 30Mhz test.
compare2.JPG
compare2.JPG (40.02 KiB) Viewed 397 times

Post Reply