[Hackrf-dev] HackRF Jawbreaker Bugs

[Michael Ossmann]

On Tue, Aug 06, 2013 at 09:34:47PM -0400, Russell Hande wrote:

>

> Okay I'll have to get back to everyone when I figure out what is going

> on, maybe my clockgen is defective or something. At 1300mhz it

> glitches up even at 2msps.


The Intermediate Frequency (IF) is set to 2600 MHz by default.  When you
tune to a frequency like 1300 MHz that has 2600 MHz as a harmonic, you
are going to get a bad spur.  This can be avoided by changing the IF.
Support for IF selection was added to the firmware and libhackrf in the
latest release, but I don't believe it has been implemented in any user
software yet.


> And at 2293.474 I am picking up noaa wx radio (162.525).


This is probably due to a harmonic of the front-end Local Oscillator
(LO).  These things become more common as you get closer to, but not
within, the mixer bypass range of 2300 to 2700 MHz.

2600 - 2293.474 = 306.526 (LO)

The LO is square, so odd harmonics are prevalent.

306.526 * 9 = 2758.734

2758.734 - 2600 = 158.734

158.734 MHz is pretty close to 162.525, so that may be what you are
seeing.  If you tune directly to 162.525, you should see the same signal
more than 10 dB stronger than it is when you pick it up from the LO
harmonic.

Again, this is a problem that can be avoided (or at least the harmonics
can be shifted around) with IF selection.  In this particular case, a
better approach might be to extend the mixer bypass range beyond 2300 to
2700 MHz.  I'm sure the MAX2837 works quite well at 2293.474 MHz even
though it is out of spec.  However, I don't believe we have the option
to override this range in current firmware.