Amplifier Trouble

[TomW]

Still running my first Cantenna, with pretty good results (great compared to stock, junk antenna). My amplifier arrived. I completely lost any planes when I installed it..... took it out, and went back to the original performance. During the test, I had:
NooElec Dongle - mcx to f pigtail - diplexer with tv out terminated - 1 foot f cable - power injector - 1 foot f cable - amplifier - 15 foot f cable - cantenna.

On the amplifier I had the antenna connected to "in" and "out" on the radio side. My power source outputs 12.18 volts (measured).

I should also mention that during the whole time, I have the diplexer in the system.

I'd appreciate any thoughts as to why amplifier doesn't work.

 

[caius]

A few things to check:

Make sure power is actually getting to the amp - measure the voltage at the connection right by the amplifier.
Check that there are no shorts in your receive chain, especially at the antenna as the amplifier will pass DC through to it.

As a suggestion, use one of the short cables to put the amplifier right next to the amplifier. That way it will be helping to overcome cable loss, and will give a better signal to noise ratio. If you have the long cable before the amplifier, any weak signals lost to noise along the coax won't be recovered by the amp.

That amp has 30db gain which is quite a lot - it's possible that it is overloading the dongle, especially as it is so close to it. Try turning the gain on the dongle down to 0, and gradually increase it and see if you start to receive anything. Another thing to try, is run rtl_tcp and connect to it with sdr# (or just run that directly if the dongle is connected to your PC) tuned to 1090MHz. If you do this with your set up that you know works, you should be able to see what the ads-b signals look like - they are short horizontal lines across the spectrogram - play with the colour settings on the right to get good contrast. Then add the amplifier and try that again, and see what the signal looks like. If it's a mess of noise, then almost certainly the dongle is overloaded. If it's blank, then it's possible that no signal is reaching the dongle - it's possible that the DC injector or the amplifier are faulty.

 

[TomW]

A few things to check:

Make sure power is actually getting to the amp - measure the voltage at the connection right by the amplifier.
Check that there are no shorts in your receive chain, especially at the antenna as the amplifier will pass DC through to it.

As a suggestion, use one of the short cables to put the amplifier right next to the amplifier. That way it will be helping to overcome cable loss, and will give a better signal to noise ratio. If you have the long cable before the amplifier, any weak signals lost to noise along the coax won't be recovered by the amp.

That amp has 30db gain which is quite a lot - it's possible that it is overloading the dongle, especially as it is so close to it. Try turning the gain on the dongle down to 0, and gradually increase it and see if you start to receive anything. Another thing to try, is run rtl_tcp and connect to it with sdr# (or just run that directly if the dongle is connected to your PC) tuned to 1090MHz. If you do this with your set up that you know works, you should be able to see what the ads-b signals look like - they are short horizontal lines across the spectrogram - play with the colour settings on the right to get good contrast. Then add the amplifier and try that again, and see what the signal looks like. If it's a mess of noise, then almost certainly the dongle is overloaded. If it's blank, then it's possible that no signal is reaching the dongle - it's possible that the DC injector or the amplifier are faulty.

Caius,

I'll do what you suggest. Can you clarify "put the amplifier right next to the amplifier"? Also, how to adjust gain on dongle?

EDIT: It was the power injector!. I replaced it and now getting lots of flights! END EDIT
Thanks
Tom

 

[caius]

I mean put one of the short 1 foot cables between the antenna and the amplifier:

antenna - 1 foot cable - amp - 15 foot cable - diplexer - 1 foot cable - DC injector - 1 foot cable - dongle
                                                                               |
                                                                         Power supply

The order of the DC injector and diplexer isn't really that important, since the diplexer can pass DC through it.

Are you using dump1090-mutability? If so, then the gain is set in the file /etc/default/dump1090-mutability. You should see a line that mentions receiver gain - you can set this to "max", "agc", or a specific number. Edit the file to the gain setting you want, save it and restart dump1090. You should see in the dump1090 log output what gain the dongle has been set to.

Here are a couple of examples to show what you might see in SDR#.

I used a temporary set up to do this - an R820T2 dongle, an LNA, and a crude antenna made from coax with the end stripped to 1/4 wavelength.

Here is what it looks like with the amp in line, but powered off. No signal is being received at all by the dongle, and the spectrum is clear of much in the way of noise.

Here, the amplifier is powered on and the dongle gain set to a reasonable level. You can clearly see the ADS-B packets appearing as the short duration wideband lines on the spectrogram, with the peak at 1090MHz. Don't use SDR# to decide exactly what gain to use, since dump1090 is a lot better at picking packets out of the noise than you can see by eye. Only use it to make sure that the dongle is not overloaded. If you see something approximating this, then you are on the right track. Obviously the number of packets showing up will depend on how many aircraft are in range, so don't worry about how many lines you can see too much, only that the ones that are there are clear and distinct.

 

[caius]

Here is what it looks like with the amplifier in line, and the dongle gain set to maximum. This setup lacks any kind of filtering, so the problem is exacerbated here, but you can clearly see the vertical lines in the spectrum that indicate that this mass of noise is an image of a nearby mobile phone tower that is overloading the dongle. You can't see any trace of the ADS-B signals, as they are swamped by the noise:

 

[caius]

And here is the troublesome mobile phone signal appearing off band. It's an LTE transmitter in the 800MHz band. Here, the amplifier is still in line, but the dongle gain is set to 0db. A diplexer or filter would reduce this a great deal.