|
| From: | Md. Atiqur Rahman |
| Subject: | Re: Baseband signals frequency,amplitude,data rate determination. |
| Date: | Sat, 15 Feb 2020 23:23:39 +0100 |
Please reply to the mailing list, not only to me privately. Thank you.
On Sat, 2020-02-15 at 02:17 +0100, Md. Atiqur Rahman wrote:
> Hello, I apologize in advance If I sound not right.
> To follow up on my inquiry:
> Well, I get to know how to find the bandwidth, as our constellation
> modulator uses some root raised cosine filter, it needs to have a
> roll-off factor which in another way excess bandwidth. Bw=(1+alpha)*T
> = (1+0.7)*(62500 symbol rate)= 106.25Khz. As the sampling rate is
> 2Mhz, SPS 32 and in QPSK 2bit inserted at once then the data rate
> would be 106.25Khz /2 = 53.125Khz am I right? if I am right then for
> the BPSK data rate would equal to bw as 1 bit per sec?
>
> For BW calculation I think it is right but date rate I not sure.
> Thank for your all support.
>
> On Fri, Feb 14, 2020 at 10:29 PM Md. Atiqur Rahman <
> address@hidden> wrote:
> > Hello,
> >
> > Yes, you got me totally. The 'time sink', I am referring to as
> > timing diagram and about showing frequency as the signal changing
> > over time, hence signals have a certain frequency.
> >
> > ''Are you perhaps mixing up bandwidth and frequency? Or maybe
> > symbol
> > rate?'' yes I am confused in this two-term, bandwidth, of course,
> > the range of frequency where the frequency of that generated
> > signal. I am mostly like to calculate things to understand a
> > certain topic but understanding them practically I have a gap on
> > that. As like how pulse shaping really determined the bandwidth of
> > the signal.
> >
> > ''Regarding bandwidth: Bandwidth of a linear modulation (without
> > Bias) is
> > always fully defined by the pulse shaping filter. You have one of
> > these
> > in your system!''
> > The first GRC flow graph I attached by followed the PSK
> > demodulation documentation, but all I really need to understand is
> > the baseband signal with its associates(data rate sampling rate,
> > symbol rate, bw and frequency)by digital modulation system. The
> > root raised filter is used for 'polyphase clock sync' but how I
> > insert a pulse shaping filter as associated with a roll of
> > factor(excess bw). As Excess BW is 0.7, symbol rate 62.5 k, sps is
> > 32 then how can I determine BW?
> >
> > I am going to read your suggested sections and will come back.
> >
> > Thank you so much for your support, I really appreciate it.
> >
> > On Fri, Feb 14, 2020 at 9:42 PM Müller, Marcus (CEL) <
> > address@hidden> wrote:
> > > Hello,
> > >
> > > > I need the basic baseband signal (low frequency) but it would
> > > be a
> > > 0Hz frequency?
> > >
> > > Yes, as said, it would be a band-limited signal around 0 Hz.
> > > That's the
> > > definition of baseband signal.
> > >
> > > I'm not sure what you mean with
> > >
> > > > timing diagram is showing the signal has some certain frequency
> > >
> > > A timing diagram (not _quite_ sure what that is) shows no
> > > frequency.
> > >
> > > Are you perhaps mixing up bandwidth and frequency? Or maybe
> > > symbol
> > > rate? (But you already have some calculations of symbol rate, so
> > > I'm
> > > really not sure what you mean.)
> > >
> > > Regarding bandwidth: Bandwidth of a linear modulation (without
> > > Bias) is
> > > always fully defined by the pulse shaping filter. You have one of
> > > these
> > > in your system!
> > >
> > > Best regards,
> > > Marcus
> > >
> > > PS:
> > > https://www.analog.com/media/en/training-seminars/design-handbooks/Software-Defined-Radio-for-Engineers-2018/SDR4Engineers.pdf
> > > might be a good place to start; you have gaps in understanding of
> > > chapters 2.1–2.3, 2.7, and 4.1–4.2, it seems. It'll be very hard
> > > for
> > > you to understand what your equalizer and synchronizers do.
> > >
> > >
> > > On Fri, 2020-02-14 at 21:24 +0100, Md. Atiqur Rahman wrote:
> > > > Thank you for the clarification. I will update it to 3.8
> > > soon.
> > > > However, baseband is the main signal which will be converted to
> > > RF signal by means of upconverter. I have a SDR device(red-
> > > pitaya), in which two digitally modulated baseband signals(I-Q)
> > > will come out separately and later on with RF front end it will
> > > be mixed with high frequency. That is the main idea. Hence,
> > > first, I need the basic baseband signal (low frequency) but it
> > > would be a 0Hz frequency? That I have no idea of. The timing
> > > diagram is showing the signal has some certain frequency, then it
> > > would be 0?
> > > > Would you please give me some details. Thank you so much.
> > > >
> > > > ps: WX block insertion was a mistake. I didn't realize it until
> > > you mentioned it. As I am going to write a report on my work as
> > > well, it is great to know that now the graph can be export. Is
> > > there any updates also for extracting GRC flow graph from
> > > file>screen capture as well?
> > > >
> > > > On Fri, Feb 14, 2020 at 9:01 PM Müller, Marcus (CEL) <
> > > address@hidden> wrote:
> > > > > Hi Md. Atiqur,
> > > > >
> > > > > On Fri, 2020-02-14 at 20:43 +0100, Md. Atiqur Rahman wrote:
> > > > > > For a QPSK modulation technique, I choose to set
> > > constellation point 1 to -1, hence 1.4142 would be amplitude
> > > > >
> > > > > Ah, you're confusing the magnitude of the complex number with
> > > the
> > > > > amplitude of the I and Q component.
> > > > >
> > > > > So, yes, the magnitude of (1 + 1j), (1 - 1j), (-1 + 1j) and
> > > (-1 -1j)
> > > > > are 1.41, but the real and imaginary parts are still +-1.
> > > > >
> > > > > > but in the timing diagram, it looks less than that.
> > > > >
> > > > > So, that's correct :)
> > > > >
> > > > > > Another point is, as my sampling rate is 2Mhz, and samples
> > > per symbol are 32, hence 62.5K is the symbol rate, what will be
> > > my baseband signal frequency?
> > > > >
> > > > > 0 Hz. That's what "baseband" means.
> > > > >
> > > > >
> > > > > Best regards,
> > > > > Marcus
> > > > >
> > > > > PS: As someone occasionallly advising students, I'd recommend
> > > > > you clean up your flow graph; there's blocks in there you've
> > > commented
> > > > > out that you can never use (WX can't coexist with QT GUI).
> > > Also, try to
> > > > > make it as linear as possible, so to make it really easy to
> > > follow the
> > > > > "sample flow" optically. This is the Nr. 1 debugging hint for
> > > GRC
> > > > > beginners, and it really helps a lot when talking about a
> > > flow graph.
> > > > >
> > > > > Also, by the way, you're using a legacy GNU Radio. We've
> > > released GR
> > > > > 3.8, and you're still on 3.7.; the new GNU Radio has fewer
> > > bugs, and
> > > > > vector export for GRC figures, which is important to anyone
> > > writing a
> > > > > thesis with GRC flow graph pictures ;)
> > > >
> > > >
> >
> >
> > --
> > Sincerely,
> > Md Atiqur Rahman
> > Hochschule Bremen
>
>
| [Prev in Thread] | Current Thread | [Next in Thread] |