Bladerf gains

Discussions related to schematic capture, PCB layout, signal integrity, and RF development

Moderator: robert.ghilduta

Post Reply
klaus10000
Posts: 2
Joined: Fri Apr 27, 2018 6:30 am

Bladerf gains

Post by klaus10000 » Fri Apr 27, 2018 9:17 am

In my latest project of an RTL SDR receiver, I've replaced my old NooElec usb key, with a brand new BladeRF board. I've started using the gnuradio-companion examples found in the bladerf folders.
The first think I noticed is that the bladerf is much less sensitive than the NooElec key receiver. Also increasing all the available gains I don't succeed in receiving the same signals I do with the old NooElec.
Looking into the osmocom gnuradio-companion block I see there are three possible gains, the RF gain, the IF gain and the BB gain. The osmocom source uses just only the RF and the BB ones. But also setting them at the maximum value, RF at 6dB and the BB gain at 60dB I still don't succeed in reaching the same sensitivity.
Some infos ont he system:
the command bladeRF-cli info returns

Serial #: 5bd6b19656b61627c92d158e21e14a13
VCTCXO DAC calibration: 0x9398
FPGA size: 115 KLE
FPGA loaded: yes
USB bus: 2
USB address: 2
USB speed: SuperSpeed
Backend: libusb
Instance: 0

the version is

bladeRF-cli version: 1.5.1-2017.12-rc1-1-ppaartful
libbladeRF version: 1.9.0-2017.12-rc1-1-ppaartful

Firmware version: 2.0.0
FPGA version: 0.6.0

The system runs on a gnuradio:amd64/artful 3.7.10.1-2ubuntu2 uptodate.

It is also a bit confusing the relationship between the gain of the osmocom block, and the bladerf gains LNA, RXVGA1 and RXVGA2. Some explanation would be helpful.

In the next step I will try to use an external LNA and filter, to see if I can compensate the low sensitivity.
Best regards

Claudio

User avatar
rtucker
Posts: 59
Joined: Sun Jan 25, 2015 10:38 am

Re: Bladerf gains

Post by rtucker » Mon Apr 30, 2018 1:25 pm

Greetings,

For the osmosdr bladeRF source block, the "BB gain" is RXVGA1+RXVGA2. It is partitioned such that all gain goes to RXVGA1 until it is max'd out, then RXVGA2 gets the remaining gain. "RF gain" is directly mapped to the LNA.

What is your setup for comparing the RTL SDR versus the bladeRF? I'd like to try reproducing it here if possible.

Thanks! -rt
Rey Tucker (she/her)
Systems Engineer, Nuand LLC
Rochester, NY, USA

klaus10000
Posts: 2
Joined: Fri Apr 27, 2018 6:30 am

Re: Bladerf gains

Post by klaus10000 » Thu May 03, 2018 3:25 am

The setup is quite basic. We have a RAC P151 antenna on a ground plane, connected via one meter (more or less) of a Huber Suhner S_02132_B cable to the receiver input.
I had prepared a pdf file with some picture of the setup and fft result, but it is too large to be attached.

The fft of the NooElec GnuRadio receiver is quite good. We have used a known carrier around 400Mhz, with gains set to medium values. With the BladeRF, with the gnuradio receiver provided in the BladeRF examples, the same fft is not as good as with the NooElec. With the basic gain setup, the carrier is lower. If we increase all the gains to the maximum, we have just only distortions I guess (or something like an inter modulation).

I’ve tried to do something similar with Gqrx, after having found a carrier at 300MHz, and the result is quite similar.
I must say the fft shape is more convincing on the NooElec scope, more detailed, while the BladeRF seems to contains less details, it is more flat, but this is just a qualitative view.
Maybe I should play with some other parameter, like the sample rate or with the bandwidth, to obtain a better result, but I would not know in which direction to go.

Having say that, I think there is something wrong in the board, or in my setup.
I welcome any help.
Regards
Claudio

AMendy
Posts: 2
Joined: Wed Jun 13, 2018 2:37 am

Re: Bladerf gains

Post by AMendy » Thu Aug 02, 2018 6:52 am

What is your sample rate? What is the length of your acquisition? What is the length of your FFT?
The more the length of the FFT is high the more the frequency domain resolution is high.
Try may be to change these parameter. I use the NoElec before and the BladeRf have a better dynamic range (12 bit ADC against 8 bit ADC) and gain of 60dB it is to amplify the power by 10^(60÷10)

Post Reply