VK6KCH

As part of cleaning up the RF environment at my QTH, I ran a quick Apples-vs-Apples comparison on two switchmode power supplies.

At least down at the VLF part of the spectrum, the newer Apple A1444 power supply (the compact little brick without the ability to change the 230V pins) is significantly quieter than the older Apple A1205 SMPS (the one that allows you to swap out the 230V connector when you're roaming to another country, etc).

The image below shows the cleaner spectrum of the A1444 below, versus the slightly noisier performance when powered from the A1205. Nothing else was changed, and the comparison was made just a few minutes apart, and then re-confirmed back the other way a few minutes later.

At a subjective level, it was much easier to copy the Pearce NDB on 340kHz when using the A1444 SMPS, versus the NDB being detectable but virtually unreadable on the older unit.

[ 281kHz and 340kHz are the YPJT and YPEA NDBs, whereas 158, 238, 316, 397, 476kHz are noise coming from source(s) that I'm yet to find and mitigate. I suspect they are coming from other SMPS ]

Of course, this is just from a sample size of one, but I thought I'd share this observation in case it helps anyone else.

73, Chris VK6KCH

Hi John and Thierry,

Thanks for your replies. The KiwiSDR is behind a 4G router, so there's a stage of outbound NAT there (with no explicit port forwarding to handle inbound connections), and then on top of that, the 4G router is only assigned a CGNAT IP address, so there's another stage of NAT occurring on the cellular service provider's edge. I had assumed that only outbound TCP sessions would work, and that attempts to initiate inbound traffic (TCP or UDP) wouldn't work, thanks to the NAT barriers (especially the cellular service provider's edge)... hence the beauty of the proxy service! It was all working fine up until recently, so the overall architecture is viable.

I suspect that an outbound TCP session from the KiwiSDR to the proxy server has collapsed and not been reestablished? Or, is it even a UDP "session", rather than TCP? (I'm now wishing that I'd inspected the traffic with wireshark earlier, but I was focused on the RF side of things).

In the short term, I'll just have to hope it fixes itself, but I'll try to get back to the remote site in the next week or two and try rebooting the KiwiSDR whilst keeping the 4G router powered up. (I wonder if the issue was caused by the whole system being power cycled as part of the changes I was making, and a race condition being entered that has messed up the NAT table, especially as the 4G router has a habit of connecting to the cellular network for about a minute, then disconnecting and reconnecting again soon after... did the KiwiSDR establish the outbound NAT during the first short connection window, but it's not honoured during the currently established connection?)

In the longer term, I wonder if there's a way that the KiwiSDR can detect that the connection to the proxy service is a bit of a zombie, and then clear it down and rebuild it, or even force a complete reboot, to try to reestablish service? (If that feature is enabled... don't want to be rebooting any completely standalone installations that are only meant to be used locally!)

73,

Chris

It seems that if autorun FT-8 and/or autorun WSPR is configured to run, it isn't possible to initiate a DRM session.

It would be ideal if starting a new DRM session would automatically preempt all the preemptible autorun sessions.

Thanks,

Chris

Thanks! It works fine now.

Just a random thought; what about using WSPR receive mode, and comparing the signals received at the local station, versus a couple of nearby stations? Wouldn't that achieve pretty tight frequency calibration? (The WSPR signals should be less prone to false detection that ye olde time signal carriers, and you could try at 28MHz to start with).

73, Chris

An update on my "shy" KiwiSDR : in short, it's back up and running!

I didn't add any entries to the blacklist at all, nor make any config changes. I think it was just some strange race condition to do with powering up the entire remote site at the same time after the RFI improvements were made, and the 4G router has a strange pattern of connecting to the cellular network, disconnecting, and then reconnecting shortly thereafter. I wonder if the KiwiSDR registered an outbound connection in that first short window, but didn't realise the session was half broken and then clear it down and start again?

Anyway, the KiwiSDR has been up and running the whole time, just not remotely accessible. I could see that the WSPR and FT-8 receive traffic was being reported due to the autorun sessions, hence how I knew it was up and running ok.

I was hoping that the KiwiSDR could still call home for firmware updates, and I believe this was the case, as it has now updated to v1.809. I believe that the restart associated with the firmware upgrade caused a new session to the proxy system, and voila, it's back up and running and I'm able to access it again! Phew!

In the meantime, I had purchased an Arlec PC191HA from Bunnings, which is part of the Grid Connect family of products, and am planning a trip to the site in the next few days so I can remotely power cycle the unit in the future. However, the firmware upgrade beat me to it! :-)

73, Chris vk6kch