My Small Radio Telescope (SRT), version 2.0

[bojan]

Yesterday I received the Nooelec RTL module (with TCXO, 0.5ppm) from ebay. It works with rtlsdr_scan (https://github.com/EarToEarOak/RTLSDR-Scanner/releases) (PC Spectrum Analyser, also records results in scv format).

Also, I started building the new interdigital BP filter (3 stages, design is here (http://www.changpuak.ch/electronics/interdigital_bandpass_filter_design er.php)), hopefully this time the performance will be better in terms of flatness and losses, so there will be less interference (last time (http://www.iceinspace.com.au/forum/showthread.php?t=113407&highlight=small+radio+telescope) I didn't use filter).


I will post again about the progress when there is one... one of the most significant improvements to the system will be motorization of the antenna, and coupling to CdC.

[SimmoW]

Keep us updated progress, even I am starting to get curious about RA due to 'visual factors' lately!

[bojan]

Construction of BPF.
The sides were cut from Aluminium bar 20x6mm (Aluminium Trade Centre (https://www.aluminiumtc.com.au/)).
Rods are from ebay, brass hex 50x5mm spacers, cut to required length (47.5mm)
The internal surface of the resonator cavity will be covered by copper sticky tape, to ensure the best possible conductivity (according to design, losses should be no more than 0.4dB)
The filter on the image is not tuned yet, it is only test assembly.



**************
Interdigital Bandpass Filter, based on work of Jerry Hinshaw,
Shahrokh Monemzadeh (1985) and Dale Heatherington (1996).
http://www.changpuak.ch/electronics/interdigital_bandpass_filter_design er.php
Javascript Version : 09. Jan 2014
-------------------------------------------------------------------------
Design data for a 3 section interdigital bandpass filter.
Center Frequency : 1420 MHz
Passband Ripple : 0 dB
System Impedance : 50 Ohm
Cutoff Frequency : 1410 MHz and 1430 MHz
Bandwidth (3dB) : 20 MHz
Fractional Bandwidth : 14.1 MHz
Filter Q : 71
Estimated Qu : 2064.25
Loss, based on this Qu : 0.597 dB
Passband Delay : 31.831 ns
-------------------------------------------------------------------------
Quarter Wavelength : 52.78 mm or 2.078 inch
Length interior Element : 47.59 mm or 1.873 inch
Length of end Element : 47.71 mm or 1.878 inch
Ground plane space : 20 mm or 0.787 inch
Rod Diameter : 5.5 mm or 0.217 inch
End plate to center of Rod : 13 mm or 0.512 inch
Tap to shorted End : 2.64 mm or 0.104 inch
Impedance end Rod : 89.870 Ohm
Impedance inner Rod : 91.889 Ohm
Impedance ext. line : 50.000 Ohm
-------------------------------------------------------------------------
**** Dimensions, mm (inch) ****
# End to Center Center-Center G[k] Q/Coup
0 0.00 (0.000)
1 13.00 (0.512) 32.44 (1.277) 1.000 0.707
2 45.44 (1.789) 32.44 (1.277) 2.000 0.707
3 77.88 (3.066) 0.00 (0.000) 1.000 1.000
4 90.88 (3.578)

**** Box inside dimensions ****
Height : 52.78 mm or 2.078 inch
Length : 90.88 mm or 3.578 inch
Depth : 20.00 mm or 0.787 inch

[mswhin63]

Neat flat ali bar. good to see how well it works

[Robert9]

Bojan, did you ever consider making the box from double-sided PCB and using brass rod for the lines?
Robert

[bojan]

Yes, I have... and I would have done it if I had thicker PCB copper cladded fiberglass plate at hand (even single-sided would be OK).
I think the approach with aluminium should be mechanically more stable than standard 1.6mm PCB. Also, I can open the aluminium box much easier, in case there is a need for minor adjustment (like length of the rods are not adequate, or tap needs to be moved a bit to get better matching...).

[bojan]

There was an recent update of rtlsdr_scan (https://github.com/EarToEarOak/RTLSDR-Scanner/releases) (PC Spectrum Analyser) software.
I asked the author couple of weeks ago for continuous scan recording on the hard disk, and apparently it is happening just about now or very soon (release should be dated 20/12/2016... with version from September, only last 500 scans were available for saving in cvs format).
With this new feature, the application will become really useful for radio astronomy.

[bojan]

Still no 100% joy..
This filter seems to have too low coupling between resonators. Maybe a thicker rods would improve that? I will try on first occasion.
Copper sheets (RS components) did not help to reduce the losses either ...
However, this filter is better than the previous one, loss is ~2.5dB, ripple is 0.5dB in 2MHz band... good enough to be placed between LNA and amplifier.

[bojan]

I checked Dishal method (https://www.johansontechnology.com/dishal-bandpass-filter-tuning-using-lasertrim-chip-caps.html) on Johanson webage..
I think I remember now using it back in 80's...
Anyway, it seems it is not necessary to monitor a phase, return loss may be sufficient. I will try....

[bojan]

Time has come to revive this post...
In parallel with ROR observatory activities (I am still sore from last week (I had to help concreter to finish the job before dark), I am working on filter optimization. This will
Also, I checked the newest version of RTLSDR scanner, it is very easy to install by installer from Git hub.

Most likely I will move the antenna to Ballarat as well, but not before we move there permanently (maybe next year?). Also, I will motorize it properly (controlled by OnSTEP or Bartels)

[hamishbarker]

Hi Bojan,

any update?

Also, what are you using for your antenna and feed horn?

Cheers,
Hamish

[bojan]

Hi Hamish,

Not much.. except I removed antenna (ø2.3m satelite dish) from Melbourne and brought it to property in Ballarat, it is in garage now.
As for feeder, I was using helical antenna (5 turns, ø50mm)
See here: http://www.iceinspace.com.au/forum/s...d.php?t=113407 (http://www.iceinspace.com.au/forum/showthread.php?t=113407)

I still have the same problem with filter passband, but not enough time to sort it out. Currently it is dismantled, I will try to improve it during holiday season....

[AussieSky]

Hi Bojan,


Thought you might be interested in a circuit for a 24db gain pre-amp as used on the Haystack University 1420Mhz SRT.

I'm just in the process of having some boards made up to play with different bandpass filters. I have an original Haystack SRT pre-amp
Greg

[bojan]

Greg, thanks for the info, it is interesting.
I couldn't find anything on BPF itself, search for 3DR25-1420/t40-1.6 gave two results, but not on filter itself.
BTW, for LNA, I am using Mini-circuits amplifier ZX60-P162LN+ (https://www.minicircuits.com/WebStore/dashboard.html?model=ZX60-P162LN%2B), it's NF is also ~0.5dB typ @1.4GHz.

[bojan]

Ispired by this project (C:\Users\stajarb\Downloads), I made my version (it supports LCD display), using Arduino Nano board's ADC. Result (calibrated with accurate SG) on input level is sent via serial interface (USB), and it is displayed on 2 rows LCD (with I2C interface).

Since this is essentially broadband rectifier/log amplifier, it will record only signal level in the whole receiver band (determined by BPF), but it could still be useful.
If anyone is interested, I can post more details of my design.