Could you paste a print screen of Device Info window from LimeSuiteGUI (Modules -> Device Info and then press GET INFO button).
OK, thanks @shao. Now we are implementing phase check software to check phase jumps between channels from different LMS7 chips using LimeSDR-QPCIe board. Will update later with results.
Could you make one experiment. Feed all 4 RX channels from one signal generator via RF signal spliter and check the phase with your software. Let me know about results.
Excuse my ignorance but do you expect that upgraded TDOA functions
can be used on two LimeSDR USB units on the same clock source?
Ok. I will buy the RF signal splitter. If I use a LimeSDR-Mini as my signal generator, is it safe to directly connect it to the receiving port of the qpcie board? Do I need to use a 5dB or 10dB attenuator on the tx port of LimeSDR-Mini to avoid damaging the receiving ports of the qpcie board?
You can safely use LimeSDR-Mini (or any otyer similar board/device, even signal generator) to generate test signal for DOA application. Just to fnd why the phase is jumping so randomly. Not to mention. Do not drive RX inputs with 30dBm Power splitter can even be made quickly (voltage divideer 150 Ohm in series + 50 Ohm to the ground) x 4. Should be good to prove the concept. Fine tunning we do later. I hope you agree? Regards,
This is my test. I am using LimeSDR-Mini as the signal generator, and a mini-circuit signal splitter. I am still using attenuator (5db+10db), because I find limesdr mini would send a high level signal at the beginning of transmission for calibration, which would saturate the receiver.
In my test, I found the phase jumping is even more severe than when I was testing with walkie-talkie. But channel 0 and channel 1 is still more stable than channel 1 and channel 2.
Are you having any underflows? That would kill your coherence no questions asked.
How to determine whether underflow is happening? I am directly calling LMS API, instead of through gnuradio, so no UUUUU or OOOOO showing up.
I think for LMS API, if something like that happened, it should also print something in terminal right? But there isn’t, you can see on my terminal in the video. Plus I am using a rather fast computer, its processor is E5, and the sampling rate is only 10MHz per channel.
I think it’s possible, even if two limesdr-usb isn’t synchronized with the same clock. It’s similar as what I am doing now, only the 2 antennae in diagonal is synchronized, and I am currently doing 2 sets of 2 channel DOA, combining their result into 360 degree.
The shortage of that is without 4 channel synchronized, I can’t detect 3 incoming signal simultaneously, and only azimuth angle is detected.
Thank You Shao,
I am expecting the same result.
Such approach can be useful for my SAR application.
If you imagine aircraft flying in the sector where someone is
expecting assistance. With 3-4 full heading sample, out of simple
“S” turn on course you can narrow exact location with certain probability.