I've been having a problem with my mount not tracking correctly recently.  I use Stellarmate for mount control and guidance and the first thing I noticed was when polar aligning I would get the message... Mount aborted. Please restart the process and reduce the speed... so I reduced the speed and got the same message this happened repeatedly. I eventually managed to get an alignment and tried some images When I did the first thing I noticed was the reticle on the screen of my laptop was jumping about in an EW or NS direction. Then it started drifting as the object I was photographing moved westwards. I'm guessing from what I observed on the calibration plot this also happened during polar aligning.  Checking the mount settings tab and the tracking had shut down.

The following night I set up again to go through the mount and the complete system having done a quick Internet search to find the problem may well lie with connections. I know the RPi3B+ is a bit fussy over the power supply so I reverted to my original power setup that had worked for the past two years with no problems and had a good 15 minute run.  After a second 15 minute run which showed one problem I swapped to the official RPi power supply even though the new board exceeded the spec of the official supply. After some experimenting I found the official supply appeared to give the best performance with a 20 minute tracking run and no errors.

The following night again I replicated this set up outside and immediately started to see the same errors despite now using the official RPi power supply. So after gathering some images I noted the reticle flicking about on the screen didn't appear to make a difference to the image quality. Problems only became apparent when the mount tracking was turned off which is to be expected. I then powered down and disconnected the RPi and the Lynx Astro FTDI EQDIR USB Adapter and attached the hand controller to the mount which then tracked perfectly.  So I don't believe it to be a mount issue.  The Lynx Astro lead is also in good condition.

I did note on checking that the RPi's RTC had lost the time and It has occurred to me that this may be the reason although I can't really see how.  Checking on the forums I know this problem has occurred on two other occasions but apart from some questions about mount behaviour there has never been a solution mentioned. Has anyone else been experiencing the same problem and how did you resolve it? I'm currently thinking of reinstalling Stellarmate on the RPi but is there something I may be missing?

I've attached an image which shows the effect visually and I've added a logfile for information.
Any thoughts would be appreciated.

Update - Problem found and all being well solved.

The Raspberry Pi 3B+ does not have an RTC and so I had added one to maintain the time between sessions especially as WiFi in my backyard can be a bit iffy.  I suspected from comments made about timing of Meridian flips and also electrical noise in the system that it may be something to do with the loss of synchronisation between the RPi and the laptop caused by a glitch with the RTC.  Sure enough when I checked the RPi the date was four days old.  Whilst the RTC worked fine once reset it lost the data overnight.  So I removed it from the RPi and set the RPi to connect via my home hub rather than create a hotspot.  This updated the time and I was able to do a test tracking for 55 minutes without seeing any reference to the horizon limits file nor any reference to align points as the reticle was not moving randomly  any longer the only movement of the reticle on the screen was a 1 pixel twitch where the screen was updating.

I'm not saying this is the answer in all reported cases but if you use an RTC in your RPi it may be worth checking it out.    

Update 2
Despite what appeared to be the solution the problem persists whether there is an RTC on the RPi or not.  So I'm reposting.