discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] Fundamental question on Sampling rate


From: Kyeong Su Shin
Subject: Re: [Discuss-gnuradio] Fundamental question on Sampling rate
Date: Fri, 27 Apr 2018 14:13:24 +0000

Hello Mehtap:


Sorry, but please ignore my previous e-mail; I may have misinterpreted a part of your e-mail and made poor explanations.


So, what I wanted say is that the advertised sampling rate of the boards is usually in terms of the complex sampling rate, not the sampling rate of the I-channel and sampling rate of the Q-channel added together (which is the definition that you are using). So, when they say 61.44MS/s, that is equal to 61.44MHz of bandwidth, not half of it (which is the case for the real sampling rate). That is why those 61.44MS/s USB3 boards are actually sufficient for 50MHz signal acquisition (in terms of the sampling rate - may still have other limitations). 


With a 61.44MS/s board, you can get up to the effective complex sampling rate of 61.44MS/s. As long as your effective complex sampling rate (after the interpolation and decimation) is higher than or equal to 50MS/s (the bandwidth of the signal), you can still demodulate the signal. 


Regards,

Kyeong Su Shin


보낸 사람: Kyeong Su Shin
보낸 날짜: 2018년 4월 27일 금요일 오후 10:40:31
받는 사람: mehtap özkan; address@hidden
제목: RE: [Discuss-gnuradio] Fundamental question on Sampling rate
 

Hello mehtap:


You are probably thinking about the Nyquist sampling rate with one ADC(or DAC) case - but most of these radios (not all, but in most situations) are designed to do complex sampling, and in case of the complex sampling, the bandwidth (assuming perfect filtering) is equal to the sampling rate of the ADCs/DACs (not half of it). 


So, you can actually get away with a USB3 board, if you prefer that. The boards advertised to have the sampling rate of 61.44MS/s will give you 61.44MHz of bandwidth, not half of it. In reality, the usable bandwidth is somewhat smaller than that, but 50MHz should not be a problem on most USB3 boards (maybe except BladeRF, which is a bit old now). Still, PCI-e boards may have other advantages, such as latency.


You need to use a fast PC or do some FPGA works if you want to  process the data in real-time. Also, the sampling rate of the boards are not always as flexible (can be adjusted, but in steps). So, please consider that when you are choosing your boards. (I think the transceivers on LimeSDR boards and USRP B2x0 boards are pretty flexible, but I do not remember their exact limitations). You can still re-sample your data, but that's computationaly expensive when the data rate is high.

Regards,

Kyeong Su Shin


보낸 사람: mehtap özkan <address@hidden> 대신 Discuss-gnuradio <discuss-gnuradio-bounces+address@hidden>
보낸 날짜: 2018년 4월 27일 금요일 오후 9:57:24
받는 사람: address@hidden
제목: Re: [Discuss-gnuradio] Fundamental question on Sampling rate
 
Based on your calculations, I have decided to get a LIMESDR-PCIe.
The sampling rate will be 122,88 MSPS which will give  61.44 MHz I and 61.44 MHz Q.
That will be enough to demodulate a 50 MHz wide QPSK modulated signal.
I am not sure if I can decimate the  122,88 MSPS by 2 and still being able to correctly demodulate the signal.
Any Ideas?


reply via email to

[Prev in Thread] Current Thread [Next in Thread]