High noise spikes in LimeSDR RX mode even without antenna

I am trying to use my LimeSDR but there are spurious noise spikes appearing no matter what frequency I tune into.

Attaching screenshots of waterfall tuning into 401MHz and 501MHz. Same noise spikes appear on both graphs.
For reference I’m also attaching the RTL820T which is mostly flat for it’s entire bandwidth.

What can I do fix things?

First two pictures warbling noise is your PC 99%, spikes may be internal.
Try to terminate selected antenna with good 50ohm terminator, and see.

73
Djani

Try to lower gain @ RTL, so noise is at -110, most likely spikes now hidden in the noise. Tnx

Odd looking Lime_401.PNG – like calibration is not done yet …

That aside … even if you power your computer and Lime from batteries … your still going to see some spikes …

Rtl 8bit ADC, Lime is 12 … wish it was at least 16 – even 24bit would be wonderful …

Thanks for your comments. Any help is much appreciated.

Termination did not help the two waving interferences much. I also set gain to lowest on both LimeSDR and RTL and the LimeSDR still has huge spikes.

Up to 35dB spikes over the noise floor does not seem acceptable.

I clicked on Calibrate TX, Calibrate RX and Calibrate all in LimeSuite. I noticed that my previous calibration was at 5Mhz bandwidth and now did it again at 40Mhz and attached the results which are about the same

What else should I be trying?

Hi diferential,

I did some checks. May not be directly related to what you have observed.

Two pictures with my laptop connected on the docking station and off the docking station.
I have a small telescopic antenna attached in both cases.
I see noise floor is about 5dB higher if I am on a docking station.
Also it seems I get some extra peaks coming from the PC (in case of the docking )

Without antenna I see similar things with and without the docking.

Also I connected pigtail UFL-SMA. Connected 50 Ohm termination at the SMA side. I have observed big dependency (> 10dB) of the noise floor SDR Console shows if I approach my 50 ohm termination to the LimeSDR USB connector. So we should try to keep the RF stuff away of the USB. This is on LimeSDR without enclosure.

More like

Hope the link is correct

I need to get an enclosure for my LimeSDR still … though my spikes are not as bad as yours …

If after full doublecheck of the settings and making sure they are at sane values try these steps:

Different computer.
Did the noise subside or radically change bamd placement? Then we can look at the computer a little more as one of the sources. To mitigate any issues make sure your SDR is shielded, that your cables are good and put ferrite beads on all your USB cables.
If you want to go the extra mile a small circuit that grounds RF on the USB power leads will make sure that isn’t the ingress.

I say “one of the sources” because your power supply will produce spikes unless it is a linear power supply or battery which is the next step: make sure you have a RF clean power supply, not using the computers USB to power it but a clean linear analogue power supply.
If you aren’t doing that then you will see a lot of weirdness in various spots across the ranges at very low signal levels.
Switching supplies are always spike monsters and create other noise as well.
You won’t be able to tell what is what when it comes to noise until you do at least those first three steps of shielded enclosure, blocking RF on the USB cables and clean power.

Your environment makes a mess too, do you have fluorescent lighting? Who doesn’t. Switch them off and see if that helps some of the noise.
Same for every appliance in the house including the fridge.

Which leads us inexorably to if it is not from the computer.
Then it is environmental and you will have to either play bug hunt with a field strength meter or a small SDR dongle works great for that as well. Hunting down environmental noise is a full subject unto itself, one that every ham should have passing familiarity with so if you aren’t a ham but know one contact him/her and they can probably help assuming they aren’t an appliance operator type.

But first look to your shielding, power supply and isolating your USB cabling.

1 Like

I managed to track things down a bit more. Took out the case and all the connectors and measured things again with no gain tuned to 450MHz center frequency.

I noticed that noise levels vary a lot and are related to the bandwidth. The waving patterns also change their relative frequency depending on this bandwidth. I would then assume that the clock generator on the board is the one causing the interference waving pattern.

Noise levels seem good without any cables at 15Mhz (only a spike in the middle at the tuning frequency). At higher bandwidths the spikes appear along with the waving pattern.




I see this noise problem at some frequencies and it causes the calibration to fail in certain spectral zones. I have not packaged my board in a shielded box yet so I will see if that helps.
I have found that I can set the center frequency outside a noise zone and tune into it with the NCO and not have the noise there. I am just hooked up through the two USB3 cables so I haven’t tried an external PS, but I will.

Does this huge spike move when you change the samplerate ?

It’s those wavy bands, not the spikes, that are getting me.

Everything running on a baytery … the wavy lines look like switching power supply to me

1 Like

The wavy lines on mine are monitor noise, probably the switchers on the UFL’s in them.
But I also get a LOT of various intermittent modem like noise as well when using my desktop for this, it is a cheap poorly shielded Asus setup with the noisiest USB and audio system I have ever encountered.
A lot of that disappeared when I switched over to my new high end laptop a couple of weeks ago.
Seems like cheap computers = tons of RF noise.