micro BladeRF ADSB

Discussions related to embedded firmware, driver, and user mode application software development

Moderator: robert.ghilduta

Post Reply
animan
Posts: 3
Joined: Tue Nov 20, 2018 11:52 pm

micro BladeRF ADSB

Post by animan » Wed Nov 21, 2018 12:44 am

Hi,

I recently purchased the BladeRF micro XA4. I was able to install GNURadio and create an FM tuner to receive stations with the BladeRF micro.

I tried to follow the instructions for the BladeRF-adsb project but ran into several issues.
- Build failure.
After modifying the makefile and modifying the API’s to version 2.0. I am able to build and run the bladeRF-adsb app.
- The adsbxA4.rbf FPGA file produces no output.
However, if I use the hostedxA4-latest.rbf FPGA bitfile (which I believe is now supporting adsb) I do get what seems visually valid raw output spewing out of the bladeRF-adsb app. However, the dump1090 app rejects most of the data as invalid (crc errors).

This is as far as I could go in my attempt to get adsb working on the BladeRF XA4.I would like to know:

- When will the bladeRF adsb GitHub site will be updated to support the new micro devices?
(It seems that there some TODOs, e.g. update of the Makefile, from 2016)

- Does adsb need some sort of an RX LNA like the BT200?
I have an ANT500 (75Mhz - 1GHz) antenna attached which should be quite capable or receiving adsb data, but I am getting nearly all CRC errors in the raw data.

Thanks.

EmilK
Posts: 10
Joined: Sun Nov 18, 2018 12:57 pm

Re: micro BladeRF ADSB

Post by EmilK » Fri Nov 23, 2018 5:28 am

Hello animan,
no blade device here but I have some experience in adsb area - I do have 2 stations sending data to flightware.
The only part where I can express an opinion.
You are saying that the dump1090 app rejects most of the data as invalid (crc errors)
I would say that the signal is saturated, I don't really know where your antenna is placed but I got this with my airspy r2 using a 10 db adsb antenna and an lna.... Of course the opposite situation can be also possible that (signal way too weak).

Can you try dump1090 with the --no-fix option? This way dump1090 should disable the single-bit error correction.
I would love to say " hope it helps"

Emil

animan
Posts: 3
Joined: Tue Nov 20, 2018 11:52 pm

Re: micro BladeRF ADSB

Post by animan » Fri Nov 23, 2018 9:14 pm

Emil, thanks for the information. I adjusted the gain up and down but it did not make any difference. In fact, I believe the the adsb app is spewing out a lot of random data. The data output from the app does not stop when there is no signal... when I remove the antenna.

Stats from dump1090 below:
Statistics: Fri Nov 23 13:10:58 2018 PST - Fri Nov 23 13:23:52 2018 PST
Messages from network clients:
0 Mode A/C messages received
14757877 Mode S messages received
3526360 with bad message format or invalid CRC
11231516 with unrecognized ICAO address
1 accepted with correct CRC


Sample output from bladerf-adsb app:
*f7fffbff04000100ffff04000200;
*fbff06000700fbfff9fffeff0700;
*02000000000002;
*fbfffafffeff04000b00fffff0ff;
*ebff07001200f6fff3ff0d000500;
*01000d00f7fffb;
*0a00fafffcff06;
*0f00f6ffefff09;
*01000100070000;
*0200fbfffdff05;
*0600fdfffbff06;
*fcfffaff04000500fdfffbff0100;
*00000300fcfff8;

Even though Nuand claims adsb support for the hosted FPGAs image... it does not seem to be working for me. The dedicated adsb FPGA image (adsbxA4.rbf) produces not output at all.

@Pallai, has adsb been working for you on the BladeRF 2.0 boards?

Thanks.

EmilK
Posts: 10
Joined: Sun Nov 18, 2018 12:57 pm

Re: micro BladeRF ADSB

Post by EmilK » Fri Nov 23, 2018 10:20 pm

My main hobby is astrophotography - having my own self build observatory. Sometimes I have faulty cables or software problems, lets call them generally problems. Even if my wife doesn't understand a bit about this I speak with her about those problems, sometimes speaking about your problem brings you on the right idea.
So I gonna do the same in your case, keep in mind that I don't own a bladerf (stuff not answering my emails anymore.... I don't know why).

1. Output without antenna
Thats should not happen, are you sure the downloaded image (adsb) is error free?

2. Do you have some gsm tower near you? Can you post a waterfall of the 1090 Mhz area?

Emil

animan
Posts: 3
Joined: Tue Nov 20, 2018 11:52 pm

Re: micro BladeRF ADSB

Post by animan » Sat Nov 24, 2018 2:03 pm

Thanks Emil. Yes talking through the issue with someone usually helps. Its cool that you have built your own observatory. I have dabbled into astrophotography and even own a Stellarvue SVQ100 astrograph but lately have been a bit busy and not quite as active on that end.

I have downloaded the FPGA images from the Nuand site a couple of times. Like I mentioned earlier the hosted A4 image is outputting some ADSB data but the data seems quite random and the output is received even without an antenna attached. The ADSB A4 FPGA image does not output any data at all, I have modified the adsb app to try Automatic gain and various other gain settings.

Not sure if I have a GSM tower nearby, but my RTL-SDR dongle with the small indoor antenna can receive data from quite a few aircraft overhead, as much as 10-15 aircraft at a time.

Hope to hear from the BladeRF support folks... I paid $500 for this board and ADSB was one of the apps I really wanted to run on this and quite disappointed that with the results I am getting so far.

EmilK
Posts: 10
Joined: Sun Nov 18, 2018 12:57 pm

Re: micro BladeRF ADSB

Post by EmilK » Sun Nov 25, 2018 7:12 am

Mate, I hear you loud and clear.
ADSB is also a point of interest for me, playing with the idea of getting something better as my rtl-sdr v3 or airspy R2 device. I was looking for the bladerf because is the only fpga based receiver that can and has an adsb application developed (excepting radarcape that can work only as an adsb receiver).
I also tried the limesdr mini and the bigger brother lime sdr but unfortunately the mini died and the lime sdr has a nasty spike on middle that I wasn't capable to remove (tried multiple calibrations but that crap was still there making any kind of reception imposible).
Im very interested if the problem that you are describing can be fixed and depending on the result I will or not push the buy button.
My idea is - regarding blade device - is to use it as an adsb, vdl2/acars receiver so long is idle (user sleeping or doing other stuff). When the user is alive :-) I will use it for noaa, iss, signal spotting and similar stuff, even I playing with the idea to do some radio astronomy, I have a 1.2 m prodelin dish that I can use, for now is used for the meteosat reception (as early weather warning system - tier 1) after that as tier 2 and 3 I use a lightning detector and a cloud detector but that is already somehow off-topic.

So, dear nuand stuff, I will really appreciate if some of you could jump in and "show som love" to the end user and also clear the crap spam that is posted daily here.

EmilK
Posts: 10
Joined: Sun Nov 18, 2018 12:57 pm

Re: micro BladeRF ADSB

Post by EmilK » Tue Nov 27, 2018 10:54 am

Stupid idea,
record using the rtl device the adsb messages got in a few minutes. Then use the blade transmission function to resend the same data that you captured and try to receive again this data (from wiki I understand that you need a 20 db attenuator when you connect TX with RX).
I don't know if is makeable but is worth a try, if you receive the same data then the software is running as it should and you have to debug somewhere else. If is receiving garbage then.... it's clear.

EmilK
Posts: 10
Joined: Sun Nov 18, 2018 12:57 pm

Re: micro BladeRF ADSB

Post by EmilK » Wed Dec 05, 2018 1:50 am

General question
i was looking at this document
https://github.com/flightaware/dump1090 ... er/bladerf

it is stated following:


bladeRF custom FPGA bitstream
-----------------------------

The package includes a custom FPGA bitstream for the bladeRF which can
improve ADS-B reception with dump1090.

It adds a Fs/4 mixer step to avoid the DC offset of the bladeRF, and
decimate-by-8 downsampling step to improve sensitivity. Some of the bladeRF
parameters need to be adjusted so that it produces correctly tuned/downsampled
data in the form that dump1090 is expecting.

To use it:

$ dump1090-fa --device-type bladerf --gain 66 \
--bladerf-fpga /usr/share/dump1090-fa/bladerf/decimate8-x40.rbf \
--bladerf-decimation 8 \
--bladerf-bandwidth 14000000 \
--freq 1085200000 \
(other options ...)

The bladeRF will be configured for a 19.2MHz sampling rate and will tune to
1085.2MHz so that the 1090MHz signal appears around 4.8MHz in the baseband
samples. The FPGA Fs/4 mixing step shifts the baseband signal so that it is centered
around 0, and the FPGA downsampling step produces samples at 2.4MHz, which is what
dump1090 is expecting.

This FPGA bitstream is built for the 40kLE bladeRF.


Considering this info is there a possibility to load internally from piaware the nuand adsb decoder?

In the bladerRF_adsb.c I see the following text

struct module_config rx_config = {
.module = BLADERF_MODULE_RX,
.frequency = 1086000000,
.bandwidth = 14000000,
.samplerate = 16000000,
.rx_lna = BLADERF_LNA_GAIN_MAX,
.vga1 = 10,
.vga2 = 6

Can I here set the gain to a lower gain because if I use it the way it is, the message rate drops from about 1100/second as manually configured in piaware to about 300/second using the nuand decoder.

Also where is defined the dump server? Piaware uses a newer version of the old dump1090 mutability.

Thank you and take care

Emil

Post Reply