SteveBz

So my SDR was indeed shorted.  The new one seems to work in that the LNA light comes on.  Good news.  However, my signal is amazingly noisy.  I have no idea why.  There is no signal whatsoever, even though I'm apparently in the middle of Cygnus:

If I ramp up the Median Smoothing I still get nothing.  I can't believe the antenna doesn't work, it's so simple!  Noisy cable, maybe?

I'll just replace it and see.

Kind regards

Steve.

8 hours ago, ZiHao said:

Hello Steve,

The genuine rtlsdr just arrived few days ago, so you asked at the right time! I was using a clone rtlsdr which has a lot of noise before this, and the genuine one shows a lot of improvement. I am using a very simple python script with pyrtlsdr for data acquisition, this is similar to Victor's method. My setup is LNA-->5m LMR400-->Rtlsdr-->Rpi. Still working on a cleaner baseline for easier data processing and hopefully able to mount the dish on my equatorial mount soon.

 

Hi Zi Hao,

So impressive. I love the ultra light dish.  Did you make it yourself or buy it? The waveguide is also very nice. I like how you've mounted it with the collar.  I might try the same idea.  You have the same NooELec H1 as me, so I'm pleased to see that it works. You're definitely ahead of me. The fact that the LNA comes out at a right-angle could put extra torque on the port.  I've put an elbow on mine to avoid that.

I also seem to have short circuited my rtl-sdr. There is no dc voltage on it at all, even after I switch it on🙄, which I didn't realise you had to do.

I've ordered a second one.  I just added a baby antenna direct to the sdr without an LNA -  for testing the software. Bad idea.

Can I ask you, have I understood that right? Did you need to active the bias tee?  No one else has mentioned this at all.

Good luck.

Steve.

On 07/02/2022 at 15:36, ZiHao said:

Nice work.
There's this excel sheet provided here http://www.setileague.org/hardware/feedchok.htm that calculates the feedhorn dimensions. I have built a 1.4m geodesic dish recently, using a feedhorn with diameter 145mm and length 180mm without choke. I didn't follow the exact length as given in the excel spreadsheet to minimize the weight as much as possible. The length of the feedhorn should be twice or three times as long as the distance from the probe to the end cap as a rule of thumb according "The Radio Sky and How to Observe it" by Jeff Lashley. 

feedhorn.xls 22.5 kB · 1 download

Hi Zi Hao,

I forgot to ask, rereading your post,  did you manage to get any results from this or have you not deployed it yet? I'd love to see your build and setup including software.

Kind regards

Steve.

So here's the dish all plumbed in and ready to go:

And here's the first connection. 

It seems to be the same as with no aerial, so I think maybe there is a broken connection.  I'll try a continuity test. But as it's dark, I'll just cover up the LNA and do it at the weekend.  I might even buy a shorter cable as this is old and 25 m long.  Maybe a new 2 m cable would be better.

It goes:

Dipole=>LNA(NooELec H1)===========25 m cable===============>RTL.SDR=>Powered USB hub=>RPi with H-line-software.

Regards,

Steve

1 hour ago, Victor Boesen said:

Thanks Steve, appreciate it! The re-write is mostly for my own sake I made some sneaky solutions back then that I would like to fix/tidy up, before I would go ahead and advance to new features. I would also like to increase performance slightly. I know the duration of observation is mostly due to the actual sampling by the SDR itself, but when performing digital signal processing on a Raspberry pi I am quite confident I can reduce the execution time by a significant amount.
Please get in touch if you experience any major issues or have a great idea I should try to implement I may not be able to fix/add it immediately, but I will add it to my TODO-list.

Victor

For me the performance is slowed down by uploading the whole sample file from the Pi to my pc. If it just returned the answer requested (Eg the max S/N ratio) that would be much quicker over the network. Or indeed the integral.

I might play with it and let you know.

Tx

Steve.

11 hours ago, Victor Boesen said:

the software is only able to compute the signal to noise ratio of the hydrogen line peak in dB.

That's perfect.

If I run

python3 H-line.py - t

on my RPi, then I can run a query from my pc every 5 minutes or so and plot the result.

I don't really think it needs a rewrite, it's quite clear, but I'll happily chat about improvements if you like.

Kind regards,

Steve.

And here's the finished product. I guess I'll need to dismantle it to get it through the door 🙂.  Then it needs a bit of a tidy-up.

The next question is software.  I've managed to get @Coto's Virgo software running on my PC, but not on the RPi3b+.  I don't want to damage my current setup, so I have a 128GB SD card on order from China to see if I can get it working with a fresh install.  The other option is @Victor Boesen H-line-software, which I got running on the RPi3b+.  In both cases I'm not quite sure how to turn the sampled output into an integrated signal (ie the total flux in 1420 +/- 50 MHz), for each sample.  I'd like to get a skymap of the 21cm line intensity in the way both the above authors have on their respective pages.

On 11/02/2022 at 17:15, Lucas Barclay said:

Good evening all,

I am currently tearing my hair out at why I keep seeing these peaks along the hydrogen line and I so far have had no success in observing the actual hydrogen line peak.. I was wondering if someone had an answer to the issue at hand here.. Is this normal to see or is there something completely wrong? Any help is greatly appreciated!

Hi Lucas,

How did you set up the radio? I'm just waiting for parts at the moment and I'd love to see how other people are doing it.

Steve.

I can't remember where I found this paper,  but it covers all the calculations very nicely and compares antennae.

Scroll to section 7.5.

Steve.

Hydrogen_1.pdf

And here is the finished feedhorn/cantenna:

Monopole connected to port.

Monopole/port mounted 9.3 cm from base.

Final product with NooElec SAWbird H1 Premier LNA and RTL.SDR.

Now comes the challenge of mounting it at the focal point.

S.

1 hour ago, ZiHao said:

Nice work.
There's this excel sheet provided here http://www.setileague.org/hardware/feedchok.htm that calculates the feedhorn dimensions. I have built a 1.4m geodesic dish recently, using a feedhorn with diameter 145mm and length 180mm without choke. I didn't follow the exact length as given in the excel spreadsheet to minimize the weight as much as possible. The length of the feedhorn should be twice or three times as long as the distance from the probe to the end cap as a rule of thumb according "The Radio Sky and How to Observe it" by Jeff Lashley. 

feedhorn.xls 22.5 kB · 0 downloads

Hi Zi Hao,

I hope you're well.  Thanks for your input.

Yes, that's about what I calculated/found.  28 cm overall length, 9.3 cm distance from monopole to back, 15cm diameter, 4.9 cm length of monopole.  Nealy finished.

I'll post a photo of the antenna when it's done.

Kind regards,

Steve.

Hi People,

What does a focus star look like?  I have a Newtonian F/5 200mm/1000mm plus a Canon DSLR and I'm trying to autofocus.  I'm thinking 15 secs on a Mag 6 star at ISO 1600 is good and then an HFR calculation what do you think?

Up until now I've been maually selecting the star in advance, but sometimes I forget to check the Dec, or RA and for large constellations these can be below the horizon (this is usually sub-optimal 🙂).  So I'm thinking a little SQL and selecting DEC of DSO target +/- 10 degrees and RA of DSO target +/- .5 hrs and Mag 5.5-6.5 for the focus star.  I then loop round, adjusting the fine focus programmatically until the HFR is a minimim.

Will this work?

Kind regards

Steve

Hi People,

I've decided to start work on refurbishing my profile photo.  This is quite a long and arduous voyage as obviously I have to start with the rust.  I've decided to drop the EQ5, as it's not really strong enough and revert to the original motorised mount with it's RA motor rotated by 90 degrees.  Ie the Earth is the RA motor and the RA motor is the Dec motor (RA is the new DEC, I could say). I spent most of this week with the wire brush attachement to my drill cleaning up, coating with Hammerite rust treatment, undercoating and finish coating the mount while waiting for parts from Amazon US and UK, AliExpress, Ebay to arrive.  In particular:

• 150 mm x 500 mm aluminised pipe + separate 150 mm end cap
• RTL.SDR software radio dongle
• SAWbird + H1 LNA
• SMA Connector and various others.

The dish is 130 cm diameter with a 47 cm focal length, making it f/2.8 if that means anything here.  😀

Here are some photos:

Adjusted, painted, motorised mount.

Hammerite rust treamement and some undercoat.

Fully undercoated support plate (it's VERY rusty)

Topcoat, also Hammerite.

Cantenna, cut down to 28 cm with angle grinder and then sharp edge ground down with emery disk.

Cantenna with end cap.

Cantenna undercoat.

Cantenna topcoat.

There are a number of measurements which it is hard to get right.  The length of the 'can' is variously given as anything from 14cm to 38 cm (including the choke).  The correct answer seems to be 28 cm. It's calculated as 'at least' 3/4 of the wavelength in the waveguide.  Apparently the wavelength in the waveguide is not 21.1 cm, but more like 37 cm after other interactions are included.  28 cm seems to be the right answer.

I'm waiting for the topcoat to dry before assembling the dish and mount for the next exciting installment.

Any warning, advice, observations or encouragement happily received.

Kind regards

Steve.

On 26/01/2022 at 14:21, robin_astro said:

The wavelength calibration looks about right but something has gone wrong with the response correction. I recorded a Uranus spectrum with the the ALPY 600 a couple of weeks back. It is in the BAA database. This is what the response corrected spectrum looked like 

https://britastro.org/specdb/data_graph.php?obs_id=11449

and this is what it looks like after division by the Pickles G2v spectrum, compared with the measurement from a professional paper

OK, very nice.  On Google there were very many different callibrations of Uranus and I wasn't sure which were right.

But the response is very off, I agree.  I'll look again if we get another clear night this side of the summer 😅

Kind regards

Steve.

Is this anywhere near the right calibration and response correction (not divided by 29 Ari)?

11 hours ago, robin_astro said:

There are no Balmer lines in the reflectance spectrum (Uranus is not hot enough to produce these) There will be Balmer lines from the reflected sunlight but they are not obvious in G2v spectra at low resolution. If you don't know the dispersion of your setup (eg from past measurements of hot stars), you will have to guess the dispersion, either matching the features in the spectrum of 29Ari to the G0v template or you could use the Telluric bands in 29Ari (They are probably too mixed with the broad methane bands to be used in the Uranus spectrum)

Cheers

Robin

OK, great.  I can just use the dispersion from my last star and then do a 1 line calibration on the main star.

Then in the meantime I'll just use old files to do the response correction approximately.

Then I'll try to use either a corrected 29 Ari, if I can, or a library star to do the division.

(PS I didn't realise that the colour of Uranus would be so beautiful - mesmerising - might try to get a longer and more in focus photo too) 

Kind regards

Steve.

Hello People,

Well I took up Robin's challenge.  Very low in the sky.  The shed roof cut off half the image!!  But just to show you it was the right object here's a quick snap in all her glory:

The greeny blue was quite unexpected!  My 29 Ari was a bit over exposed, so I may need to play with that a bit, but in the meantime here's the Uranus spectrum with some old biases and darks (I may refresh them if I need to) and stacked in Siril then cropped and spectrum from Bass. No response file applied.  My first question is, I don't even know where the H-Beta line is, if there is one. Have I got it right?

What to do next?  I'll work on the 29 Ari file in the meantime, but could I not use a standard library spectrum?

Thanks

Steve.

9 hours ago, robin_astro said:

Have you bagged a Uranus spectrum yet?  Well placed in the early evening, about  the same brightness as 63 And. It should appear pretty much star like in your short focal length Newt so give a sharp spectrum showing nice Methane bands.  G0v star 29 Ari is conveniently nearby too at the moment and similar enough to the G2v solar spectrum so if you take that as well and divide one by the other (no need to correct for response) you should get the reflectance spectrum of Uranus direct without the contamination from the solar spectrum features.

Cheers

Robin

Hi Robin,

I was so excited by your suggestion.  I thought it would be really interesting and quite challenging, however, I've just checked its location and I see it's almost bang on the ecliptic.  Sadly I don't think I can get that low.  I normally work on about 35-30 degrees Declination in the South.  But I guess this must vary by time of year.  Wednesday is looking good, I'll take a photo as low as I can and platesolve to see how low it is.

Kind regards,

Steve.

Hi Gents,

I had so many hiatuses with getting this Bresser 114/500 to piggy-back my mount.  I'm back now to where I was 6 months ago!  Anyhow it seems to be working.  Bresser f/4.4 mounted on my main scope on an EQ6 with a QHY5iii178mm and SA100.  It worked well last night, so here it is (63 And - B9v) 50 exposures stacked in Siril with darks and flats corrected and calibrated in Bass with an old response file from before.  Everything's changed but it seems to work.  Obviously I need to update the response file.

(NB Just updated with darks and biases.  Much smoother response)

Does anyone have some nice ideas for spectroscopy in the current season, assume we get any more clear nights?

Kind regards

Steve.

Hi Ken,

Yeah, I thought it was nicely dated😁

So the other option might be a focal Reducer, but I don't want to introduce chromatic aberration into the optical stream. Would that work (maybe to replace one of the spacers I have now between the SA and the sensor). What would you recommend?

Have a good day.

Kind regards

Steve.

On 13/12/2012 at 00:31, Merlin66 said:

I bought a stack of them for the grandkids (boys and girls 8-12yrs)

before I could even finish telling them about telescopes and how to.........they had it out the box, on the table looking at the Moon , and later Jupiter!!!

Bang per buck - highly recommended.

(I also kept one to pull the mirror out for a spectrascope project....where else would you find a good 75mm/ 300mm mirror for 25 gbp!!!!)

Hi Ken,

I just bought an etx 90 as a piggy back for my sa100, but focal length is 1250 mm and that really lowers the resolution. My cheap bresser 114/500 is much crisper. That said I imagined a 76/300 would be even better. What do you think?

Kind regards

Steve.

So now we have a minimum viable product, here:

There's a slight bounce on the close function, which I'm working on, but basically it more-or-less works.

Here is the unit.  Does anyone recognise it.  I'm trying to find the user manual to set the stop limits properly.

Now to think about a few bit to add to stop it flying off in a storm. 😅

Kind regards

Steve.

11 minutes ago, robin_astro said:

Resolution is not the only parameter in spectroscopy. You can go much fainter at low resolution  and there are lots of interesting targets for the simple Star Analyser including WR stars, bright novae and supernovae and even quasars. 

Nice.  I'll look into that.  Actually, I'm quite pleased with the results and the setup.  I don't really want to go more complex at this stage. I automated the slide-off roof on my shed at the weekend with an old garage door opener I found on ebay, which makes things a lot easier.

As I mentioned before the reason I'm doing all this is that I'm also doing a project on binary stars with a colleague.  Theory is great, but actually looking at stars really helps to keep it real.  Even just chatting here with you about Zeta Lyrae brings clarity to the issue.  I saw the description 'binary star' on Zet01 Lyr, but not Zet02Lyr and I thought oh, they missed one, but your clarification about the 4.3 days and the spectral binary really helped shed light on the situation for me.

So it's all very well being theoretical, but looking at it in real life really brings a new dimension to ones understanding. 

Kind regards

Steve.

20 hours ago, robin_astro said:

Yes 6Lyr definitely  is a known binary with a close secondary (The orbital period is just 4.3 days) but the secondary is not seen in the spectrum, most likely because it is much fainter. We know it is a binary because the lines in the spectrum from the primary move back and forth on a 4.3 day period. This is known as  a single line spectroscopic binary. See the wikipedia page  for the reference to the 1910 paper

Cheers

Robin

So it's a hierarchical system. 6 Lyr is the brighter one and it has a close invisible companion only detectable because of it's variable effect on the spectrum.  7 Lyr is fainter and may be a wide companion for 6 Lyr.  So the nice strong spectrum I have is for the brighter star (ie 6 Lyr), but I don't have the resolution to show the detail you showed earlier.  Do you think I could improve the spectrum I got, or is that pretty much as good as it gets with the SA?  I was pretty sure you had some better resolved ones.

Kind regards

Steve.

1 hour ago, robin_astro said:

There is an extremely high resolution professional spectrum of 6 Lyr in the ELODIE archive

http://atlas.obs-hp.fr/elodie/fE.cgi?ob=objname,dataset,imanum&c=o&o=6 lyr

Cheers

Robin

Actually, that's a fantastic resource. I shall look into it more.