full day of mounted + portable harvest

Today I will be planning and storing for the next move. which will happen tomorrow morning.   Since it will be a homebody day I will set the timer to go off 2x/hour and will record the solar harvest.   This morning’s temps were ~40F and it will get up to ~70F this afternoon. Clear skies.  Winter solstice, should should be the shortest daylight of the year. 

According to this page the day’s solar events look like this(rounded to the nearest minute):

  • sunrise -  0741

  • solar noon - 1236

  • sunset - 1732

The portables will be positioned twice:  once for early morning harvest and once for afternoon harvest.  


There is no unified place to monitor the difference systems so I will be using:

  • the BT dongle app on the main mounted 40A 570w system

  • a physical MT50 monitor on the portable 10A 200w system

  • the shunted battery monitor to observe Vbatt, charge rate, and amp count.  Note that after we hit Vabs the amp counting will not accurately reflect SoC due to lead’s infamous charging inefficiencies in Absorption.

things to watch for

  • after the system holds 12.9v for ~10 minutes the opportunity load circuit will engage and loads will increase sharply
  • as ambient temps rise the fridge will run more
  • by 10-11am we will hit Vabs and acceptance will start dropping; this means we will not see what the system is capable of at solar noon.


The mounted panels (yellow line) are high enough that creosote bushes didn’t interfere. The portables (blue line) were shaded heavily until 9am and were no longer shaded by 10am.

Absorption voltage was attained at 1015, at which time the system was making ~335w; I don’t know exactly because it was between half-hour measurements. This is the big set of peaks on the left half of the graph:

As demand fell the controller for the portables stayed maxxed out at around 140w and power supplied by the mounted panels decreased. That pattern held until the sun started to set behind the mountain, at which point the portable harvest plummeted and the mounted panels attempted to rally (small yellow bump toward the end).

There is no practical difference in which panels make what power as long as needs are met. Note that in both bulk and in the sunset both controllers gave their respective max. Only in Absorption where you have more capacity than load would the preference be revealed. Still, the discrepancy is interesting.


  • the wiring between the 10A and battery v. the difference between the 40A and the battery could result in slightly different voltages being detected by the controllers, and therefore slightly different voltages being delivered to the battery. Since the higher voltage “wins” in multi-charger scenarios it’s clear the 10A was charging at a slightly high er voltage. The controllers only have a 0.1v granularity so I don’t know how far off they are.
  • there is a temp probe only on the 40A controller, so one was measuring controller temp and one was measuring battery temp. That could also result in small differences.

I was more interested to see what effect the preference for the portables would have on the seriously overpaneled controller. The little booger is overpaneled +67% (ie, 3.5x more aggressively than the 40A controller which is +18.75%). It wasn’t a fair fight, but the $40 controller stayed hammered at full/clipped output for 6 hours today and never made it up to body temperature (maxxed at 94F). Ambient temps in the low 70s. I wouldn’t do it in 100F weather, but the heavily-heatsinked Chinese controller took a lickin’.

Absorption current ceased meaningful progress around 2:30p, or 4.25 hours after reaching Vabs. It stuck at C/100 (2.2A), which is to me is right on the margin of acceptable. When the bank has been fully charged consistently it drops to C/125, or ~1.75A. It finally did that at 4:30p, or 6.25 hours after hitting Vabs. A few more days of good sun should get us back in that territory faster. Note that if I hadn’t set Vfloat==Vabs it would have dropped to float at 2.76A when the controller’s timer hit 180 minutes.

In total, the portables harvested 1,090w and the mounted ones 530w, for a total of 1,620w. Not bad for the shortest day of the year when the system had excess power from 1030 on. With ~3.0 hours of FSE available in AZ in late December I reckon the practical max harvest if all the power were consumed would have been about 1,917w.

moving on

While I was doing housework and making the above observations, a couple came by and mentioned there has been no BLM enforcement here perhaps due to covid-related closures in nearby California. I told them I had to go to town anyway to mail Dear Daughter’s BD card. And I don’t feel comfortable breaking the rules when there is other land available.

Tomorrow I’ll hit the laundromat for washing and wifi, fill up the tank, visit the USPS, and head out to Parker.