generate signal in the wrong frequency

[weitinglu]

Recently I try to use DFS Pulse Tester. But I facing some problem.
I not sure it is my hackrf's HW fault or software not setting in the correct way.
It could generate rf signal but in the wrong frequency.

My system is using ubuntu 18.04.
install driver/software by pybombs.
everything is up-to-date.

hackrf_info :
hackrf_info version: git-1ae7c21e6c
libhackrf version: git-1ae7c21e6c (0.5)
Found HackRF
Index: 0
Serial number: 0000000000000000644064dc243c72cd
Board ID Number: 2 (HackRF One)
Firmware Version: 2018.01.1 (API:1.02)
Part ID Number: 0xa000cb3c 0x00744758

GNU Radio Companion 3.7.12.0

My friend helps me with using the spectrum to get the result. The result as follow.

DFS Channel	Spec	Measured	Delta
52	5260	5486	226
56	5280	5520	240
60	5300	5555	255
64	5320	5592	272
100	5500	5912	412

How do I fix it? or that is hackrf's limitation, it is nothing we can do?

[ecsv]

Just experienced the same problem when playing around with Debian Buster, same version (+firmware) of HackRF One and an AP on channel 120. I had to use channel 64 as pulse generating channel to see a detected radar on the AP.

Judging by the the QT GUI Sink, the generated pulse seems to be generated on channel 64 and not channel 120 (where the AP detected it). And at first glance, the pulses seem to be formed as expected in the QT GUI sink. And both QT GUI sink and the osmocom Sink are using the same center frequency.

Not sure if @adriangranados or @mossmann is the correct guy to ask for help.

[ecsv]

Seems to work a lot better for me when changing the sampling rate (lower left block) to 8e6 (1e6 * scaling_factor) and also changing the DFS_Pattern options to include the scaling factor for each vector element. @weitinglu, can you try this out and check the delta?

Just noticed weird (not really classified yet) behavior when using lower sample rates in gqrx. The device didn't end up at the right frequency (or at least the analogue device didn't end up at the same frequency as with 8Ms). So I just thought about testing this here again.

[weitinglu]

Hi @ecsv
My hackrf not by my side now. I will try your method asap once I get them.

[mossmann]

Be sure to update your CPLD to match the version of firmware installed. A mismatch can result in incorrect frequencies. Also you may have unpredictable results below the lowest sample rate supported by HackRF One which is 2 Msps. For best results I recommend operating at 4 Msps or above.

[koralilgun]

I had better results using a 5Ghz band capable antenna (an external antenna from a 5GHz wifi router for instance) on the HackRF One. The ANT500 that came in the package with my HackRF One did not work. It only seemed to work for very brief periods of time. The WiFi router I was testing for radar detection rarely detected the signal sent from HackRF One. As it says in https://greatscottgadgets.com/ant500/, ANT500 is designed for 75Mhz to 1Ghz operation. It doesn't seem suitable for DFS purposes.