Battery voltage was being reported slightly lower each morning.  There was a trend occuring that would eventually lead to a problem. The refrigerator was working too hard.  It seemed to run constantly.  It used to get frosty, accumulating a layer of ice when we cruised.  That hadn’t happened the last several years, but it had been still cold.  Not so, this year.  The troubleshooting part of the manual said to add refrigerant.  I had done that to several automobile a/c units and as the process looked similar, believed I could handle the Alder Badour unit.  I found the connection and added some R-134a as specified.  The cold plate frosted up nicely!   A job well done, or so I thought.

The next morning, battery voltage was reported at 12.1, a new low!  About the 50% state of charge, meaning nearly 300 amp hours had been consumed in the past 12 hours.  It was cloudy most of the day, and stormy that evening.  When it is raining, the panel produces almost no output.  It varies on a cloudy day, sometimes barely mattering.  This was not one of them.  The wind conspired against me, keeping my radome positioned to cast a shadow on the panel (which nearly shut it down) for the portion of the day that had any usable light.  By the next morning, battery voltage was 11.7.  Discharging AGM batteries below the 50% level takes a toll on their state of health.  I try to avoid anything that would curtail their useful lives.  Time to plug her in to shore power and charge them up while I troubleshoot!

The trouble was easy to identify– the refrigerator never shut off.  The troubleshooting section of the manual says that not only can not enough refrigerant be the cause, but too much,  as well!   Adjusting the system’s pressure to around 21 psi (about 6 psig) seems to yield the frost patterns the manual is looking for.  The fridge is now cycling as expected and using the predicted 70 amp hours per 24 hours.  Still, I am installing a switch that I can control remotely that will cut power to the unit.