Pulsar Observation Phase 2

Introduction

After the first experience with pulsar i gave up because the increasing number of interference on the 23cm band reached such a level not compatible anymore with my setup.
The limitations were mainly the narrow available window due to the filters inside the transverter with 4Mhz bandwidth and also my Perseus SDR with its 2Mhz max sample rate did not allow an high sensitivity in pulsar signal detection.
Next big limit was my home made software far away from the sophisticated algorithms used by professional observatories.
The original setup was therefore deeply revised, as much as possible, in terms of receiver chain and data post processing.

Receiver Chain

Antenna, Feeder and LNA

Same as before, The 10m dish and the LNA are usable on a wide band and even if the feeder is optimized for the 23cm still it is usable for at least a 20-30 MHz window without a big gain loss.

Receiver

As receiver i have choosen to try a direct connection from antenna to a cheap Airspy SDR with a good UHF coverage up to 1.6 Ghz and 10Mhz max bandwidth excluding transverter and IF receiver.
The week points of the Airspy are the very poor NF and the saturation because of out of band strong signals.
The NF limit was overcome by adding a 35db gain mmic wide band amplifier in front of the SDR.
An interdigital 5 cells filter with low attenuation and about 50Mhz bandwidth protects from strong nearby interference.

Data Acquisition

Before describing the new environment i have to introduce my friend Andrea (iw5bhy) because it is thanks to his contribution if i could get new achievements in pulsar detection.

IW5BHY

I met him for the first time on the moon path making a 23cm EME QSO in 2014.
He was interested in radioastronomy and made his own path for detecting the B329+54 with a small antenna first on 23cm and then on 70cm where he built a real time observatory.
The results are visible on WEB and , in my opinion, it deserves a visit:
IW5BHY Observatory

It certainly is the first, and only, amateur installation of this kind.
In addition, Andrea provided for a group of amateur radio astronomers a sophisticated but easy to use software suite tailored for pulsar reception and detection.
An example of amazing achievements can be found by visiting the WEB site of Hannes (OE5JFL):
OE5JFL Radio Astronomy

I have to mention also another contribution to the group by Mario (I0NAA) and his useful MURMUR utility which allows,for a defined location and setup, to make provision for pulsar detection.
The applicatin can be freely downloaded from:
I0NAA MURMUR

Data Recording


A pulsar observation may last for several hours and be able to collect an I/Q base band signal at 10MHz sample rate is not a trivial process, it requires fast processing and fast transfer rate to disk for a huge amount of data.
Andrea (IW5BHY) choose a different approach and based on GNU open source radio components (GNU Radio) wrote a smart utility that does the following:

  • Connect to any SDR radio supported by GNU Radio

  • Receive the base band blocks

  • Covert from time to frequency domain

  • Split the frequency domain block in a desired number of channels

  • Convert back to time domain

  • Decimate to a desired ratio the data

  • Write the result to a binary file


All done in real time with a reasonable output file size.

A number of parameters are provided for setting several options.

The channelization is needed by the post processing in order to correct the frequency dispersion of the pulsar signal.

Data Conversion and Data Processing

In order to proceed with data post processing the binary collected data file needs to be converter in filterbank format (.fil) .

Andrea produced a “bin2filterbank” utility to produce the desired .fil file according to the requested options.

Post Processing

Care of the post processing is left to a nice subset porting (Andrea) of the “Presto Pulsar Suite” done by Andrea (Presto).

The Ubuntu porting of Presto is running on a Linux virtual machine supported by a Windows PC.

The processing of data consists of analysis in the frequency domain to compensate for frequency dispersion and data folding to enhance the candidate pulsar signal.

Inside the Presto feature there is also some RFI (Radio Frequency Interference) mitigation capability.

The output is provided in graphic and numeric form.

Achievements

The following is a list of the already observed Pulsars with a link to specific pulsar data and graphics.

The graphics are composed by the full Presto output plus a zoomed view of the pulsar profile.

The pulsars are listed in decreasing flux intensity order (which is measured in mJansky )and therefore in increasing order of difficulties for their detection which means in any case being able to detect few thousandths of db signal out of the noise:

Nr.

Pulsar Name

Notes

Profile

1

B0329+54

Strongest Pulsar In Northern Emisphere

2

B0950+08

3

B1933+16

4

B2020+28

Two peaks on same pulse

5

B1929+10

6

B1133+16

7

B2016+28

8

B2021+51

9

B0628-28

10

B0355+54

11

B1642-03

12

B1749-28

13

B2154+40

14

B0531+21

Crab Pulsar, Only Giant Pulses detected

15

B0740-28

16

B2310+42

17

B1804-08

18

B0450+55

19

B0823+26

20

B1237+25

21

B0809+74

22

B0540+23

Only 18 minutes of useful recording



Part of the information are taken from the ATNF Pulsar catalog:
">ATNF