Anyone doing hydrogen line observing?

[Coto]

Alright, so no success yesterday, even though I was very confident with my setup etc. My suspicion is the bandpass filter. I removed that, soldered the 12V power wires to my in-line amp, and the + to the center pin of the first-stage LNA, connected a simple wire (as an antenna) to the first-stage LNA and wow, I can see a ton of RFI (absolutely none of which was visible with the filter, not because it was filtering harmonics or something, but because it didn't seem to pass the signal to my SDR at all - I've had problems with the filter before so this is my suspicion). This is what it looks like without the filter (LNA + in-line amp): 

 

That's some next-level RFI coming from... the power supply I'm assuming. Is it really possible for a 10 Amp 12V power supply to be sooooo noisy? Could I do something about this? Luckily I had already ordered a new power supply (12V, 5A) so I'm waiting for this one to arrive. I really hope it's not as noisy as my current power supply... (I'm also curious if using a window for the FFT could perhaps eliminate some of that RFI, especially if they're harmonics (non-sinusoidal waveforms))

[Carl Reade]

What sort of supply is it?

I would also suspect the simple wire antenna was it grounded?

Was the pi on for the first time?

Suspect everything that changed since your last successful detection not just the filter. 

No the FFT won't deal with that.

Carl

[Coto]

"What sort of supply is it?"

It's just a 10 Amp 12V AC to DC converter (taking power from the plug).

 

"I would also suspect the simple wire antenna was it grounded?"

What do you mean by "grounded"? It's just a wire attached to the center hole of the SMA port of the LNA (all the spikes go away if I remove the wire antenna).

[Carl Reade]

1 hour ago, Coto said:

"What sort of supply is it?"

It's just a 10 Amp 12V AC to DC converter (taking power from the plug).

 

"I would also suspect the simple wire antenna was it grounded?"

What do you mean by "grounded"? It's just a wire attached to the center hole of the SMA port of the LNA (all the spikes go away if I remove the wire antenna).

There's your answer it's the wire. The spikes would remain if it was the power supply.

Try it with the feed horn. I would stick to what works adding a wire is a random element.

If you change something keep the rest the same. It makes life easier.

[Coto]

Well, I connected it to the feed, tried a 30 min observation while MW is right in the beam, nothing... 😕

[Carl Reade]

7 hours ago, Coto said:

Well, I connected it to the feed, tried a 30 min observation while MW is right in the beam, nothing... 😕

First thing I would check is voltage getting to the LNA and all connections good.

[Coto]

I'll check with to see if power increases once I plug/remove the first stage LNA. As soon as I remove the feedhorn from the LNA, the spikes disappear. I assume the LNA (if it doesn't get any power) can still pass the signal to the in-line amp without amplifying it. (the in-line definitely gets power since I see its LED light up)

[Carl Reade]

3 hours ago, Coto said:

I'll check with to see if power increases once I plug/remove the first stage LNA. As soon as I remove the feedhorn from the LNA, the spikes disappear. I assume the LNA (if it doesn't get any power) can still pass the signal to the in-line amp without amplifying it. (the in-line definitely gets power since I see its LED light up)

No there would be no signal your looking for just random stuff whatever the circuit board metal would pickup, like a bit of tin foil at the end of the RF chain.

A volt meter at the LNA will reveal all.

[Coto]

Gonna try some troubleshooting soon when I get back to my dish. Just a question that came up to my mind: I installed the LNA a little bit too deep into the box, so the SMA male to male connector wasn't long enough to reach my feed's connector. Is it wrong that I used 2x right-angled SMA connectors before reaching my feed's SMA to N-Type connector? Could the loss of these connectors be significant enough to degrade the signal strength of the hydrogen line that I would otherwise be able to detect without these connectors, or am I just being paranoid? 😛

[Carl Reade]

2 hours ago, Coto said:

Gonna try some troubleshooting soon when I get back to my dish. Just a question that came up to my mind: I installed the LNA a little bit too deep into the box, so the SMA male to male connector wasn't long enough to reach my feed's connector. Is it wrong that I used 2x right-angled SMA connectors before reaching my feed's SMA to N-Type connector? Could the loss of these connectors be significant enough to degrade the signal strength of the hydrogen line that I would otherwise be able to detect without these connectors, or am I just being paranoid? 😛

Hi shouldn't make any difference, however not so sure about having a PI so close to the amps but that will just need testing.

[Coto]

13 hours ago, Carl Reade said:

Hi shouldn't make any difference, however not so sure about having a PI so close to the amps but that will just need testing.

Why? The Pi gets pretty hot and you think could make the LNA noisy? (I don't think that's the case - they aren't in contact so it's not like the LNA would feel very warm or something)

[Carl Reade]

2 hours ago, Coto said:

Why? The Pi gets pretty hot and you think could make the LNA noisy? (I don't think that's the case - they aren't in contact so it's not like the LNA would feel very warm or something)

LNAs always work better cool. It's more RF interference from the PI I was thinking of.

[Coto]

So after a bit of investigating, it is indeed the RPI causing all this RFI, and not the power supply I was suspecting. Also, it seems like the in-line amp and the bandpass filter just don't seem to work. I really don't know what I'm doing wrong, but as soon as I insert the in-line amp after the LNA, I see no HI line. Same goes for the hydrogen filter: I insert it after the LNA, no HI line. It's as if the in-line amp and the HI filter are acting as attenuators... I'm not sure. Either way, using the LNA alone works fine for me so I'll go with that (I made a fine observation last night).

 

As for the interfering Pi, I'll try wrapping the project box in aluminum foil (forming a faraday cage) and making sure the aluminum foil makes contact with the feedhorn's ground (the waveguide side = the N-type connector's outer "shell"). If this doesn't help I'll have to get the Pi out of the box and put it away elsewhere, or maybe change it's CPU clock speed (cause it's originally rated at 1.4 GHz...).

[Carl Reade]

4 hours ago, Coto said:

So after a bit of investigating, it is indeed the RPI causing all this RFI, and not the power supply I was suspecting. Also, it seems like the in-line amp and the bandpass filter just don't seem to work. I really don't know what I'm doing wrong, but as soon as I insert the in-line amp after the LNA, I see no HI line. Same goes for the hydrogen filter: I insert it after the LNA, no HI line. It's as if the in-line amp and the HI filter are acting as attenuators... I'm not sure. Either way, using the LNA alone works fine for me so I'll go with that (I made a fine observation last night).

 

As for the interfering Pi, I'll try wrapping the project box in aluminum foil (forming a faraday cage) and making sure the aluminum foil makes contact with the feedhorn's ground (the waveguide side = the N-type connector's outer "shell"). If this doesn't help I'll have to get the Pi out of the box and put it away elsewhere, or maybe change it's CPU clock speed (cause it's originally rated at 1.4 GHz...).

I think you will need to move the PI away from the rest. Not sure what's going on with the line amp and filter perhaps a power issue when they are in the chain. Could it be that the LNA isn't getting power when they are connected?

I prefer the satellite inline amps as they and the LNA are just fed off the coax.

Keep posting your results.

[Coto]

Separating the Pi out of the box would be quite problematic for me as I had already planned everything to nicely sit in a box... Also, putting it a few meters away would also be susceptible to interfering with my dish. I think I'd still need to make a faraday cage for. Currently planning to try this with my current box, although I'm not so sure how it should be done... Do I just cover the whole box with aluminum foil? Do I then need to connect this aluminum foil anywhere (as a ground/redirecting all the EM energy the Pi emits elsewhere away from the dish)?

[Carl Reade]

13 hours ago, Coto said:

Separating the Pi out of the box would be quite problematic for me as I had already planned everything to nicely sit in a box... Also, putting it a few meters away would also be susceptible to interfering with my dish. I think I'd still need to make a faraday cage for. Currently planning to try this with my current box, although I'm not so sure how it should be done... Do I just cover the whole box with aluminum foil? Do I then need to connect this aluminum foil anywhere (as a ground/redirecting all the EM energy the Pi emits elsewhere away from the dish)?

A metal box would be ideal. You could line the box in foil and attach the RF ground to it and also foil up the pi seperate, ideally the pi would be best away from the RF amps. Be a matter of testing.

[Coto]

5 hours ago, Carl Reade said:

A metal box would be ideal. You could line the box in foil and attach the RF ground to it and also foil up the pi seperate, ideally the pi would be best away from the RF amps. Be a matter of testing.

So I guess my question is, where does this energy go? Like the Pi emits radiation, it is picked up by the foil (as an antenna?) and transferred to the feedhorn's ground? Where does the energy go from there? Does it get radiated outwards (into the sky and also into the dish reflector as well)? Where exactly does this energy go?

[Carl Reade]

54 minutes ago, Coto said:

So I guess my question is, where does this energy go? Like the Pi emits radiation, it is picked up by the foil (as an antenna?) and transferred to the feedhorn's ground? Where does the energy go from there? Does it get radiated outwards (into the sky and also into the dish reflector as well)? Where exactly does this energy go?

You want to be doing two things.

All of your RF chain should be shielded in metal enclosures. There's not much point in having shielded coax with the line amp open to RFI. So in the RF chain it's to keep RFI out. That would leave RFI only getting in from the antenna and where filters come in.

With the PI your trying to keep its RFI in an enclosed area so it's not interfering with anything else. Bit like a microwave oven.

So doing both greatly reduces the PI and external sources interfering with the signal your after.

However another issue is cabling which can carry RFI so cables ie USB should also have a ferrite core on as well.

Computers simply are not screened for RF.

Amplifiers are dumb if they see it in their range they will amplify it, hence the use of filters.

[Coto]

Well, I tried covering the entire box with aluminum foil, making sure it touches the antenna's ground, but I still get RFI... Why is that?

[Carl Reade]

2 hours ago, Coto said:

Well, I tried covering the entire box with aluminum foil, making sure it touches the antenna's ground, but I still get RFI... Why is that?

Is the PI on?

[Coto]

15 minutes ago, Carl Reade said:

Is the PI on?

Yes... I did a lot of experiments. I think I'll install the Pi under the dish (whose reflecting surface acts as a shield reflecting the Pi's interference away into the Earth's ground). I tried that and the RFI went away (at least sufficiently enough). I think it'll still cover it with aluminum foil or something.

 

P.S.: Do you know if G8FEK's L-Band Radio astronomy LNA is waterproof, or will I need to put it in a project box?

[Carl Reade]

13 hours ago, Coto said:

Yes... I did a lot of experiments. I think I'll install the Pi under the dish (whose reflecting surface acts as a shield reflecting the Pi's interference away into the Earth's ground). I tried that and the RFI went away (at least sufficiently enough). I think it'll still cover it with aluminum foil or something.

 

P.S.: Do you know if G8FEK's L-Band Radio astronomy LNA is waterproof, or will I need to put it in a project box?

Yea PI needs to be shielded or away from the rest of electronics.

Not sure if the LNA is water proof so I boxed mine and sealed all the connectors.

[Coto]

Just installed G8FEK's L-Band radio astronomy LNA on a fixed 3.2m dish pointing 20 deg below zenith. Here is a first result:

(The square-ish blocks are due to data intervention (had to remove some RFI spikes))

Detecting the hydrogen line is nice and all, but what about multiple HI lines in a single observation? You already did this: 

And it's been done here, by an even smaller dish: http://blog.svenbrauch.de/2014/08/29/14204-mhz-hydrogen-line-there-it-is/ (my feed dimensions are identical too)

And these guys with a 3.1m dish seem to get amazing results with similar equipment: http://www.midem-drustvo.si/Journal papers/MIDEM_47(2017)2p113.pdf

Firstly, how can I distinguish hydrogen line emissions originating in the spiral arms of our galaxy, from other types of emissions (e.g. nebulae like Cas A, emissions with extragalactic origins etc.)?

Secondly, what does it take to see multiple lines in a single (averaged) spectrum? Is pointing an important factor I'm not considering? I've managed to do it with the 1.5m (up to 2 lines though) with little observing time (2 mins in this case):

I think it'd be interesting to see more than 1 or 2 hydrogen lines, proving we live in a spiral galaxy by a single plot. 😁

 

P.S.: My chain is Feed -> G8FEK LNA -> 3m LMR-400 coax -> 1m USB extension -> RTL-SDR (without the shield 😕) -> PC

Do you think it'd increase my S/N significantly if I inserted an in-line amp right after the LNA?

EDIT: Also, do you think it'd do any difference if I took an OFF measurement and then subtracted the ON data from the OFF data? (I don't believe I could extract anything else from this plot)

Edited July 12, 2019 by Coto

[Carl Reade]

2 hours ago, Coto said:

Just installed G8FEK's L-Band radio astronomy LNA on a fixed 3.2m dish pointing 20 deg below zenith. Here is a first result:

(The square-ish blocks are due to data intervention (had to remove some RFI spikes))

Detecting the hydrogen line is nice and all, but what about multiple HI lines in a single observation? You already did this: 

And it's been done here, by an even smaller dish: http://blog.svenbrauch.de/2014/08/29/14204-mhz-hydrogen-line-there-it-is/ (my feed dimensions are identical too)

And these guys with a 3.1m dish seem to get amazing results with similar equipment: http://www.midem-drustvo.si/Journal papers/MIDEM_47(2017)2p113.pdf

Firstly, how can I distinguish hydrogen line emissions originating in the spiral arms of our galaxy, from other types of emissions (e.g. nebulae like Cas A, emissions with extragalactic origins etc.)?

Secondly, what does it take to see multiple lines in a single (averaged) spectrum? Is pointing an important factor I'm not considering? I've managed to do it with the 1.5m (up to 2 lines though) with little observing time (2 mins in this case):

I think it'd be interesting to see more than 1 or 2 hydrogen lines, proving we live in a spiral galaxy by a single plot. 😁

 

P.S.: My chain is Feed -> G8FEK LNA -> 3m LMR-400 coax -> 1m USB extension -> RTL-SDR (without the shield 😕) -> PC

Do you think it'd increase my S/N significantly if I inserted an in-line amp right after the LNA?

EDIT: Also, do you think it'd do any difference if I took an OFF measurement and then subtracted the ON data from the OFF data? (I don't believe I could extract anything else from this plot)

Glad to see your getting results. I don't think the line amp would be any advantage with only 3m of coax.

Yes pointing is important, depending where you point and arms in that direction. The above graph was a drift scan of Cassiopeia. If your getting a big jump and small bump it's safe to say Cygnus area has passed through the scope.

I use the off method for all the graphs that brings out the details and should give you results.

To seperate different objects would be difficult as the resolution of the small scope simply is not good enough and there is a lot of hydrogen up there.

Maybe using other frequencies than hydrogen would be useful to do this.

I have still to put the total power bit of the scope together to see if it works or not.

[Coto]

@Carl Reade Just made a post on the Radio Astronomy section of the forum about my radio telescope.