Wednesday, 3 November 2021

Obtaining Accurate frequency spots from the RSP1a and skimmer server

 As my Anan-10 is often used as an HF multiband panadaptor (to see which bands are open) rather than a skimmer I needed to deploy another SDR as a cw skimmer. I had a spare RSP1a so decided to try that, even though it is in a plastic box!

Firstly the sdruno and api were installed from here, which made sure the rsp1a drivers were installed. Next the skimmer server was installed from here along with the needed SDRPLAYintf files to allow skimmer server to speak to RSP1a. The aggregator software was also installed from here to allow spots to be uploaded to the reverse beacon network.

The cw segment on 40m (outside contests) is narrow so only a 96kHz bandwidth (as is done on the anan-10) was chosen in the skimmer server settings and the system left running overnight

The following morning I examined the spots but found they had a random frequency error compared to other stations spotting the same station (eg ka9zap was -11.6 kHz, k3wjv was +2.1khz, k4ees was -0.3khz)

I wondered if the issue was band sensitive, so the following evening i skimmed 80m. The errors were similarily random.

I did a web search to see if anyone else had suffered the same issue, but found nothing. However I did find a page in Spanish by EA2NN showing he had succesfully used skimmer server with an RSP1a. I did however notice that he had set the skimmer server bandwidth to 192kHz so i thought it was worth a try.

After making this change all spots were within 100Hz of everyone elses spots on 160, 80 and 40m. The problem had been solved!

Note that in the skimmer server ini file you can define the frequency range that cw signals can be decoded (less than 192kHz!)

I also discovered that if you have the skimmer looking at 3 bands (the maximum it would appear from an RSP1a)  the default is that it skims 2 minutes on the 1st band then 2 minutes on the 2nd band then 2 minutes on the 3rd band before going back to band 1, so if you generate a local test signal locally it may take 6 minutes before it finds it!

Now to put the rsp1a in a metal box. I may even add an RF relay to the box so the input can be terminated on transmit!

Monday, 1 November 2021

47 GHz and 76 GHz DATV outing


Having failed to work 76km on 76GHz from the Blorenge G8GTZ and I decided to try some shorter paths

The first one was from Hackpen hill IO91CL to IO91GI Combe Gibet. I had never been to Hackpen, but I had previously worked it on 47 and 76GHz narrowband from Cleeve Common so I went there. When I arrived there was mist on the horizon towards Combe, but it slowly cleared and QSOs were had on DATV on both bands over the 26km path

I then travelled to Combe masts IO91GI61 and met with Noel and I worked G1EHF/P using fleapower on 47GHz. . With sunset being around 1615, and wanting to try another path of 35.6 km we forsook a pub lunch and had a sandwich before Noel went off to Cheesefoot IO91JB. I returned to the Combe masts. To save time we started on 76 GHz and a 250ks qso was had, even though there was a lot of mist on the path.