I received a USRP B100  today and I am experiencing some
issues with the Basic Rx.  I am getting the following warning from
GNURadio:

UHD Warning:
The hardware does not support the requested RX frequency:
Target frequency: 98.300000 MHz
Actual frequency: -29.700000 MHz

I
have also tested this using ExtIO_USRP and I notice that any requests
to change the LO result in the hardware changing the frequency between
-32MHz and +32MHz.  This is the output:

Req:        160.470884 MHz
H/W:        -31.529116 MHz
Target LO:    160.470884 MHz
Actual LO:    0.000000 Hz
Target DDC:    -160.470884 MHz
Actual DDC:    31.529116 MHz

Req:        240.470884 MHz
H/W:        -15.529116 MHz
Target LO:    240.470884 MHz
Actual LO:    0.000000 Hz
Target DDC:    -240.470884 MHz
Actual DDC:    15.529116 MHz

Req:        140.470884 MHz
H/W:        12.470884 MHz
Target LO:    140.470884 MHz
Actual LO:    0.000000 Hz
Target DDC:    -140.470884 MHz
Actual DDC:    -12.470884 MHz

This results in the same portion of 64MHz of spectrum looping during tuning.  Is the driver broken???

Regards,

Mark McCarron

Ian,

Thanks for that, this is something Ettus Research fails to make clear on their website.  So, what filtering options exist and where can I find them?

Regards,

Mark McCarron

Subject: Re: [USRP-users] USRP B100 Tuning Issue
From: ianb@ionconcepts.com
Date: Tue, 7 May 2013 14:22:22 -0700
CC: usrp-users@lists.ettus.com
To: mark.mccarron@live.co.uk

Mark, The B100 runs at a 64MHz sampling rate. Basic RX has no LO, it's really just simple amplifiers to buffer the antenna connections and no filtering.Thus the only available tuning elements are the DUC/DDC in the FPGA, and there tuning range is naturally bounded by the sample rate.You can target the frequencies in your example because the bandwidth of basicRX is so high, but you must supply external filtering to prevent other frequency ranges aliasing onto them.-Ian
On May 7, 2013, at 2:03 PM, Mark McCarron mark.mccarron@live.co.uk wrote:I received a USRP B100  today and I am experiencing some issues with the Basic Rx.  I am getting the following warning from GNURadio:

UHD Warning:
The hardware does not support the requested RX frequency:
Target frequency: 98.300000 MHz
Actual frequency: -29.700000 MHz

I have also tested this using ExtIO_USRP and I notice that any requests to change the LO result in the hardware changing the frequency between -32MHz and +32MHz.  This is the output:

Req:        160.470884 MHz
H/W:        -31.529116 MHz
Target LO:    160.470884 MHz
Actual LO:    0.000000 Hz
Target DDC:    -160.470884 MHz
Actual DDC:    31.529116 MHz

Req:        240.470884 MHz
H/W:        -15.529116 MHz
Target LO:    240.470884 MHz
Actual LO:    0.000000 Hz
Target DDC:    -240.470884 MHz
Actual DDC:    15.529116 MHz

Req:        140.470884 MHz
H/W:        12.470884 MHz
Target LO:    140.470884 MHz
Actual LO:    0.000000 Hz
Target DDC:    -140.470884 MHz
Actual DDC:    -12.470884 MHz

This results in the same portion of 64MHz of spectrum looping during tuning.  Is the driver broken???

Regards,

Mark McCarron_______________________________________________
USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

Marcus,

That was an excellent explanation.  You should write the user manuals. :)

Is there a complete filter bank that spans 1-250MHz that can be interfaced either through USB or the B100?

I had been reading other posts of yours because I am trying to create a proper antenna setup.  I see that the Basic Rx/Tx have different input specs from the WBX.  2Vpp and 0.2Vpp respectively.  I didn't see any specs on the maximum input current.

So far, I have come up with this basic list:

Basic Tx
Wideband Tx Antenna (either 1-250MHz or to the bottom end of the WBX @50MHz)
5000A/8000A wideband lightening/surge protector (low insertion loss @0.2dB or better)
RF Wideband Power Amp to match antenna
Low pass filter - cutoff 250MHz

Basic Rx
Wideband Rx Antenna (either 1-250MHz or to the bottom end of the WBX @50MHz, possibly Active Antenna)
5000A/8000A wideband lightening/surge protector (low insertion loss @0.2dB or better)
Wideband LNA to match antenna (or use active antenna)
Low pass filter - cutoff 250MHz

WBX (Tx)
Wideband Tx Antenna (either 50-2200MHz or to the top end of the Basic Tx @250MHz)
5000A/8000A wideband lightening/surge protector (low insertion loss @0.2dB or better)
RF Wideband Power Amp to match antenna
Low pass filter - cutoff 2.2GHz

WBX (Rx)
Wideband Rx Antenna (either 50-2200MHz or to the top end of the Basic Rx @250MHz, possibly Active Antenna)
5000A/8000A wideband lightening/surge protector (low insertion loss @0.2dB or better)
Wideband LNA to match antenna (or use active antenna)
Low pass filter - cutoff 2.2GHz

To prevent transient pulses from damaging the B100 that are not enough to trigger the surge protector (I think these kick in at 90Vpp), I require something that will attenuate input signals.  I think you recommended -25dBm for the WBX.  I didn't see a recommendation for the Basic Tx/Rx.  Any good suggestions?

Do you see any problems or anything I have missed?

Regards,

Mark McCarron

Date: Tue, 7 May 2013 19:20:15 -0400
From: mleech@ripnet.com
To: usrp-users@lists.ettus.com
Subject: Re: [USRP-users] USRP B100 Tuning Issue

On 05/07/2013 05:43 PM, Mark McCarron wrote:

  
  Ian,

    

    Thanks for that, this is something Ettus Research fails to make
    clear on their website.  So, what filtering options exist and
    where can I find them?

    

    Regards,

    

    Mark McCarron

You aren't the first person to be somewhat confused about the
capabilities of the BASIC_RX series cards, all I can say is that I
know that the data-sheets

  are currently being worked over, feverishly, to clarify things
like that, and provide more detailed characterization across all of
the various

  daughtercards.



There are filter manufacturers all over the place, and for
frequencies below 200Mhz or so, you can build your own from discrete
L-C components,

  without too much hassle.  I've often used this website to help me
design filters from parts I happened to have lying around:



http://www.wa4dsy.net/filter/filterdesign.html



The thing to understand about these BASIC_RX cards is that they're
just designed for direct-sampling into the ADCs.  The analog bits of
the card itself are

  good to about 250MHz before they start to attenuate, and the ADCs
are also good to fairly high frequencies, which means that so-called

  bandpass-sampling can be achieved:



http://en.wikipedia.org/wiki/Undersampling



So, any signals that are presented to the card above the 1st Nyquist
zone (32Mhz for USRP1/B100/E1XX, 50Mhz for N2XX) get *aliased* to

  other frequencies that are within the 1st Nyquist zone.  The UHD
drivers know this, and will set the tuning (which is all-digital in
the case

  of the BASIC_RX) based on alias calculations.  But in order to
make this work, you have to have bandpass filters of reasonably
high-order

  to limit the frequencies presented to the ADCs to the "Nyquist
Zone" of interest.  Otherwise the "real" frequencies and "alias"
frequencies

  will both be presented to the ADC, which will cause nothing but
grief.



For example, if you wanted to use BASIC_RX just for listening to the
broadcast FM band, you would  provide a bit of gain, and a filter
that

  stiffly presents only the 88-to-108Mhz chunk of spectrum to the
BASIC_RX.  Notice that the bandwidth of this--20Mhz, is less than
the

  1st-Nyquist bandwidth of the USRP1/B100 (32Mhz), so that will work
reasonably well.   In fact, it turns out that good-quality FM-band

  filters are cheaply available, such as these:



http://radiohobbystore.com/gfwb3.html



Similar strategies can be used to do bandpass sampling at other
frequencies, but once you have a biggish filter-bank setup, you
might consider

  a card with actual on-board analog downconversion like WBX or
TVRX2, etc.







-- 

Marcus Leech
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

Marcus,

Absolutely correct on the hardware side of things, I write DSP software and I'm doing a deep-dive on the physical setup.  Thanks for all your help.  As I am assembling all of this over the next month or so, no doubt I will pick your brains.

Right now, I have a noise issue.  I am using the WBX board.  All throughout the sub-1GHz band, I find the following noise pattern with limited gaps of 20MHz or so now and again.  They are narrow-band AM signals that are ever-so-slightly FM modulated.  See this image:

http://i.imgur.com/OSVhxp8.jpg

My area should be RF quiet and by that I mean absolutely dead unless cows can broadcast.

I went to the 2-meter band and I found this:

http://i.imgur.com/0tbZt3x.jpg

Any idea what this is???  Or better yet, how to get rid of it?

Regards,

Mark McCarron

Date: Tue, 7 May 2013 21:19:48 -0400
From: mleech@ripnet.com
To: mark.mccarron@live.co.uk; usrp-users@lists.ettus.com
Subject: Re: [USRP-users] USRP B100 Tuning Issue

On 05/07/2013 08:54 PM, Mark McCarron wrote:

  
  Marcus,

    

    That was an excellent explanation.  You should write the user
    manuals. :)  

    

    Is there a complete filter bank that spans 1-250MHz that can be
    interfaced either through USB or the B100?

  

Not that I'm aware of.  Such "filter banks" usually aren't done as a
separate commercial module, but usually integrated into the

  front-end of some types of receivers.






  

    I had been reading other posts of yours because I am trying to
    create a proper antenna setup.  I see that the Basic Rx/Tx have
    different input specs from the WBX.  2Vpp and 0.2Vpp
    respectively.  I didn't see any specs on the maximum input
    current.

  

Well, the input impedance is 50 ohms, so ohms law.  The WBX is a
full analog downconverter with a complete RF gain chain, mixers,
etc.  It's

  more along the lines of what you'd expect for a
wants-to-be-plugged-into-an-antenna receiver for 50Mhz to 2.2GHz, so
the maximum input levels

  are very much lower, as you would expect.  Generally in RF, we
refer everything to dBm, usually in a 50Ohm system, rather than
talking about

  P-P voltage levels.  That's just the way the RF industry "works".








  

    So far, I have come up with this basic list:

    

    Basic Tx

    Wideband Tx Antenna (either 1-250MHz or to the bottom end of the
    WBX @50MHz)

    5000A/8000A wideband lightening/surge protector (low insertion
    loss @0.2dB or better)

    RF Wideband Power Amp to match antenna

    Low pass filter - cutoff 250MHz

  

I'd again urge caution in the use of RF power amps at arbitrary
frequencies.  That's likely to get unwanted attention not only from

  law enforcement, but other properly-licensed users of the same
spectrum.






  

    Basic Rx

    Wideband Rx Antenna (either 1-250MHz or to the bottom end of the
    WBX @50MHz, possibly Active Antenna)

    5000A/8000A wideband lightening/surge protector (low insertion
    loss @0.2dB or better)

    Wideband LNA to match antenna (or use active antenna)

    Low pass filter - cutoff 250MHz

  

For terrestrial applications below 50Mhz or so, there's no point in
trying to even *design* a *true* low-noise amplifier.

  Terrestrial and background galactic noise will utterly dominate
any noise added by the electronics themselves.



Also, below about 1GHz, it's hard to find low-noise amplifiers with
enough dynamic range (expressed as p1dB and OIP3)

  that you can do without application-specific filtering.  The RF
spectrum is *really crowded* these days, and strong signals

  are everywhere that can easily drive your nice little LNA into
non-linear territory without app-specific filtering.






  

    WBX (Tx)

    Wideband Tx Antenna (either 50-2200MHz or to the top end of the
    Basic Tx @250MHz)

    5000A/8000A wideband lightening/surge protector (low insertion
    loss @0.2dB or better)

    RF Wideband Power Amp to match antenna

    Low pass filter - cutoff 2.2GHz

  

See power amp cautions as above.



The WBX already has a 2.2GHz LPF on the output, as I recall.




  

    WBX (Rx)

    Wideband Rx Antenna (either 50-2200MHz or to the top end of the
    Basic Rx @250MHz, possibly Active Antenna)

    5000A/8000A wideband lightening/surge protector (low insertion
    loss @0.2dB or better)

    Wideband LNA to match antenna (or use active antenna)

    Low pass filter - cutoff 2.2GHz

  

See cautions about linearity.    In any reasonable suburban/urban
setting, letting the front-end just "soak up whatever is out there"
is generally

  unwise, at least if you care about quality and the ability to pick
out weak signals.




  

    To prevent transient pulses from damaging the B100 that are not
    enough to trigger the surge protector (I think these kick in at
    90Vpp), I require something that will attenuate input signals. 
    I think you recommended -25dBm for the WBX.  I didn't see a
    recommendation for the Basic Tx/Rx.  Any good suggestions?

    

    Do you see any problems or anything I have missed?

    

    Regards,

The WBX already has TVS protection in the front end, and adding an
external GDT is a good idea.



If you're connecting directly to the BASIC_RX/BASIC_TX from an
antenna a GDT by itself should be sufficient.



In either case, if they're going to be sitting in strong RF fields,
you should consider attenuators, which can be purchased from a
multitude of

  vendors.  But doing so will, necessarily, reduce your sensitivity.



From the type of questions, I'm getting that perhaps RF isn't really
a core expertise for you, and I would urge you to learn about RF
engineering

  before just "throwing a whole pile of equipment together".   The
ARRL handbook, for example, offers a treasure-trove of fairly easy
to understand

  chapters on theory, and engineering considerations.  I've acquired
my knowledge through *years* of experience, having been licensed as
an

  amateur radio operator since 1986, and worked in technology since
1979.  There's still a lot I don't know, and unlike, for example,
putting

  a home entertainment system together, putting RF systems together
requires real *engineering* discipline.  Mistakes in this area lead
to

  damaged equipment and/or systems that perform very poorly, not
because any given component is "bad", but because the overall

  RF system design is not well considered.







-- 

Marcus Leech
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org