New install of the Discovery Dish with the L-band weather satellite LNA. So far no signal detected from GOES 16. I made the following measurements for tech support and awaiting further instructions:
Using a USB power meter the RTL V3 without feed attached voltage shows 4.943vdc, 0.150A, 0.743W.
With the feed attached: 4.895-4.919vdc varying, 0.150A, 0.735-0.742W varying.
With the bias-tee enabled: 4.716-4.724vdc, 0.389A, 1.840-1.859W
Still no signal detection shown in the spectrum and waterfall in Satdump.
Have not made any measurements at the LNA without approval.
The current draw is a bit higher than expected, which is a little odd, but often USB current meters have inaccurate readings so it could be ok.
If you have AGC’s turned on, make sure to turn them off. With the AGCs on, you won’t notice any difference in the noise floor as the AGC will compensate for the feeds gain.
Post a screenshot of your SatDump settings so we can check it.
I tried a lot of different setting combos in Satdump, only change was in the noise floor. Spectrum/Waterfall was not showing any signals either. Gain was set to Max manually. Installed SdrSharp to see what could be found on there, I could hear audible noise and there was and clicked on a spike at the upper end of the spectrum which was just a faint steady tone. I don’t have a live screen shot as I received an email yesterday to return the LNA and I repackaged it, but here is a screen shot of my settings:
I built my own GOES station back in the 90s with a 10 foot dish and a coffee can feedhorn with LNA, used it for WEFAX and EMWIN acquisition.
Waiting for a return ship label for now:
"It seems that the feed is drawing power, but maybe there is something else that is wrong. Could you please send it back to me? What is your shipping address? I’ll send you a label. "
If you still have it, I would suggest upping the samplerate to 2.4 MSPS or 2.56 MSPS. Satdump is a little picky, often it won’t decode at lower sample rates.
1.536 MSPS might be too small to cover it with the rolloff, and you also won’t be able to see the signal hump as the GOES signal is about 1.2 MHz and so it would cover most of the visible bandwidth and just look like noise.
Also, it’s possible that the FFT graph was just out of scale. If the scale is too big, then the signal won’t be easily visible because it will only be a few dB above the noise floor. You can compress the scale by playing with the FFT MAX and Min sliders.
Also not sure how accurate the dish was aimed, but I assume you had it aimed at the satellite, no obstructions, and had set the skew correctly? Often a little bit of trial and error with the aiming is required.
Thanks for your input. Syed had me ship the feed to him, so I have nothing to use for now. I tried various settings for two days, the last you see in the screen shot were ones I saw other people using which didn’t help either. Before the GOES 16 position shift I was using 131.3 magnetic AZ and 34.9 Elev. The 1694.1mhz setting only put a thin line down the middle of the waterfall. Changing FFT settings only turned the waterfall all one color with no signals embedded in it. I would think even a faint signal from GOES would show something in the spectrum and waterfall. But I could not replicate reception that others have demonstrated on Youtube.
Not sure what you are referring to in regards to “skew”, I did try vertical and horizontal poliarization of the LNA if that is what you mean. But no in-between rotations.
I am contemplating when my L-band feed is returned changing out the 6 meter coax to the LNA to just a 1 meter section to reduce strain and then going the rest of the distance to my PC with KMR400 COAX and then back to a short section of RG58 to the SDR in order to minimize line loss. Any problem with changing the coax at the LNA end? From what I could see from another individuals photo the LNA board connects with SMA to BNC with an adapter. Would that be a warranty issue?
–Thanks, Bill
There’s really no need to change the cable, the gain on the LNA is quite high, so any coax losses are very negligible. You simply won’t notice any changes with higher end coax unless you’re doing a really long run like 15 meters+. In fact the LNA needs a little loss on the output to work optimally.
The circuit warranty won’t be voided, but we won’t cover anything that goes wrong with mechanical issues due to the modifications.
If I set the Gain to a low value like 16db or less then the BPSK Sign turns back to Yellow:
If I set Gain= 0 and then Enable AGC then BPSK Signals go back to Red and the Viterbi State finally Syncs but no Sync in the Deframer:
I have rechecked the dish pointing, adjusting Skew makes little difference to the AvgSNR and the best I have achieved in that readout is about 3.1db.
I did add an 20 foot extension coax of KMR400 to the 6M provided coax and should have only attenuated about 2.2db to reach from dish to PC. I don’t understand why in Satdump increasing the gain causes the signal SNR to start turning Red.
So no success for me at this point.
I have a Zoom account and can easily setup a screen sharing session if needed.
Update: Well Carl I tried something different after posting the above. I removed the coax extension and in it’s place installed a USB extension so that I could attach the SDR directly to the L-feed coax. Also set the skew for 60deg and finally started getting some Syncs and partial images. So will continue to tune things some more. Even though I have processing set for HRIT I get a lot of LRIT CRC invalid errors. Not sure why I am getting LRIT messages when setup for HRIT.
Can you post another full screenshot, showing everything, especially the SDR settings and the waterfall?
Secondly, are you sure that you have the satellite alignment 100%? You can roughly point to it using a phone or compass, but you will need to go out there and tweak the dish alignment to get the best signal.
The gain should be on manual, and maxed or near max for best results. If you’re getting worse results with the gain up, it could be a case of a nearby interferer causing issues which is unfortunately very difficult to deal with with any hardware.
If you are getting an average SNR of 3.1dB, then this is enough for a decode. But from your images I think the SNR you are seeing is not real, because your BPSK demod screenshot does not show a lock. The green dot cloud should have two distinct balls if you’re actually receiving the signal.
It will be a while for a screen shot, had decent but not ideal reception, decent amount of images received, some complete others not. Never have seen two distinct balls, because average SNR would stay around .3-…4db, peake SNR would sometimes hit 3.95Db and turn yellow. Tried readjusting skew, Az/El and now I have lost everything. Been hours trying to peak this dish. My best result was running a 20M USB repeater cable to the SDR at the dish position. But then I am not sure if that is causing irroneous results for signals.
So will post a full screenshot if I ever get close to working again. Sometimes Satdump will occasionally freeze and I have to stop/start again to continue. Might be USB related, this is the best cable I have for now for long runs. I have others that control my deep space cameras in the observatory and they work fine but this one is a different brand.
Well I have spent almost 7 hours today trying to peak this dish and the results go from bad to worse. I can no longer achieve what I had at 4:00 a.m. At least I was getting a collection of images and other data in spite of the low SNR. I setup a Zoom session to my tablet and made adjustments at the dish in AZ, no matter where it was pointed on eaither side of 130degM I could not get a decent SNR. I gave up on AZ and started adjusting skew and finally Sync appeared intermittently but could not achieve the sync locks I had at 4:00 a.m.
Here’s a couple screen shots of my settings, could not capture a fresh Sync lock after this session. So far I am not optimistic about achieving the average SNR you suggest. I was going to order a spare L-feed but not after this session for now.
That looks like you have A LOT of RFI. All those small spikes in the spectrum should not be there, and I suspect that it causing SatDump to fail decoding. I can see a decent signal there, hidden under all that RFI, though.
This RFI looks like in-band noise from local electronics, like LCD screens, switch mode power supplies etc. Is there anything nearby or anything near the path where the dish is pointing that could be suspect?
Screens, computers, power supplies, TV antennas (emitting RFI from the TV they’re connected to), cameras, solar panels, improperly shielded PoE equipment like cameras etc are some suspects to think about. Anything coupling through the coax, or USB cable connected to the dongle could be something to look at too.
That makes sense since it was dark when I had the best signal. There are some ground mounted solar panels nearby. They are disconnected from the inverter but can still produce 80vdc as the sun rises even though unloaded. They are not in the line of sight of the dish and it is aimed away from them, closest group is about 20 feet away but a few feet to the right of the dish. They have disconnect switches on the back side of them, so will switch those off and see what the spectrum looks like.
I also have a portable frequency counter with antenna for locating stray RF sources that goes up to 2.4ghz I could try that around the dish and follow the USB cables back to the PC inside.
From what I can read thus far solar panels in themselves don’t generate broad spectrum RFI, the inverters are the main culprit in that regard. Just the same I will test them with a frequency counter and antenna.
I will also check out the utlity smart meter on the back of the house which is about 30 feet away from the dish:
Smart meters primarily use radio frequencies in the 900 MHz and 2.4 GHz bands to communicate data, with 900 MHz offering better range and 2.4 GHz providing faster data transfer speeds.
