Hybrid Car – More Fun with Less Gas

Home solar system - Page 2

register ::  Login Password  :: Lost Password?
Posted by ads on June 21, 2018, 6:37 am
On Sun, 17 Jun 2018 16:02:16 -0400, "Jim Wilkins"

I have two EPEver/EpSolar Tracer 3210A 30amp MPPT controllers.  Bought
on ebay from sure electronics about 6 months apart.  One was $7; the
other was $7 including shipping.

The difference between the DC-DC converter in the link and the MPPT
controller is that the MPPT controller does a regular scan (45-60
minutes when looking for enough power to enter MPPT mode; don't have
the timing when it's in operation) to check for a better transfer
configuration (searching for the maximum power point).  If there is
sun, then clouds, then sun, then clouds the controller automatically
adjusts as needed.  The 3210A doesn't seem to respond to changes as
fast as some of the more expensive brands (Outback, others) but I can
accept that limitation when I'm looking at a three or more times price
difference (MidNite solar 30A MPPT is $00).

These units idle at 0.015 amp (15ma) if the load control is turned off
or about 0.035 amp (35ma) with the load turned on (15 ma is less idle
power than some basic PWM controllers).  The units go into MPPT mode
when there's enough solar power to run the DC-DC conversion and charge
the battery.  It needs about 8 or 9 watts from solar to trigger MPPT
mode.  Once in MPPT mode, it stays there as long as it can - typically
down to about 1 watt of solar input.  

If there's not enough power to enter MPPT mode, the 32XX controillers
run in PWM mode so they are always trying to get the battery to your
selected charge/float voltage if the solar voltage is greater than the
battery voltage.  Using the software, you can specify the minimum
(during discharge) and maximum (during charge) battery voltages
allowable and what the boost and float voltages should be.  The
controller does an equalization charge on the 28th of each month and
you can set the voltage and number of hours for that charge, among
many other parameters.  If you're using multiple controllers to handle
a large number of solar panels, you only need to configure one
controller's parameters because you can save the parameters and load
them into another controller.  

I've been doing some testing of solar panel placement, so I've been
using one panel (100 watt or 250 watt) in various locations.  The
3210A consistently performs the same, with the 100 watt panel taking
longer to charge the battery bank than the 250 watt panel but the
difference matches the 2.5 to 1 difference in panel power.

The 32XX series controllers are available in models up to 40 amps and
they are 12/24 volt capable with auto select based on battery voltage.
The free software talks to the 32XX charge controller via an RS485
adapter (USB-to-RS485 dongle is $ on ebay) and the connection uses
two wires of a standard CAT5 cable.  Using the software, you can
configure the various charge controller parameters for lithium battery
packs as well as lead acid batteries (I'm using AGM's because I got
them for a very good price).  The software displays the controller's
voltages, currents and power (battery, solar, load) in near-real-time
(default is 30 second updates).  You can print these graphs to have
hard copy of daily status.  They also offer the eLog01 device that can
record several months of ALL the data (volatge, current, power,
temperature, power consumed, power generated, etc, etc and you can
save that data in .csv format which any spreadsheet can read.  This
can provide even more information about the ongoing opeartion and
status of your system.

The original 32XXA series use common positive (positive ground)  The
newer 32XXAN (introduced this year and about the same price) is common
negative (negative ground).  I would prefer the negative ground
version but it's not worth replacing tested units with a newer model.
The software works on either version but some of the accessories
(MT-50 remote meter, others) are not compatible between the A and AN

This email has been checked for viruses by Avast antivirus software.

Posted by Jim Wilkins on June 21, 2018, 12:39 pm
<ads> wrote in message  


So you like how they perform.

My measurements with a DC Volt-Amp-Wattmeter and rheostat load suggest  
the power loss from directly connecting the panel to the battery  
(=PWM) is no more than 10% after the battery voltage reaches 13.8V,  
supposedly at SOC of 70% to 80%. For example on June 19 at 3PM with a  
hazy sky the max from a 100W Grape panel was 77W at 16.8V, decreasing  
to 70W at 14.0V. The current did increase as the voltage lowered. The  
only parasitic loss was the 6.2mA drain of the wattmeter.

The current needed to hold that voltage on a 105Ah battery soon  
decreases below the full output of one 100W panel. At that point I  
separate the panel downleads and distribute them among the batteries  
to charge and monitor each independently. Panel output is no longer a  

When the sky becomes clear enough I plan to measure down to 11.5V and  
compare the current from a direct connection with the highest current  
available from the DPS5015, manually adjusted to the MPP. This appears  
to be the charging region where MPPT holds the greatest advantage over  
PWM, however I need a generator and high current chargers anyway for  
overcast days.

I think I'd need to add at least two more $9 100W panels to reach the  
cost break-even point for an MPPT controller, where the gain over a  
PWM controller equals the gain from adding another panel.

Perhaps my best investment into this power outage backup sytem would  
be a storm shelter for the generator instead, such as a folding A  
frame roof over an HF steel service cart. I can run the genny at meal  
times for food prep and a fast charge and use any available solar in  
between. A roll-around tool cart is very useful for outdoor carpentry  
such as repairing storm damage. My HF inverter genny fits the lower  

The data I've seen on lead acid battery life suggests the replacement  
cost cancels any savings from free solar input. Lower depth of  
discharge may give longer life but the total KWH delivered over its  
life is almost constant.


Posted by ads on June 21, 2018, 10:20 pm
 On Thu, 21 Jun 2018 08:39:02 -0400, "Jim Wilkins"

I'm very pleased with the 3210A performance.  

 In addition to the 100 watt panels I started out with,, I picked up
some used Jinko 250 watt polycrystalline panels for $50 each from a
local solar installer.  They're pulls, but only a few years old and
they deliver 235 watts in backyard sun (94% of the rated 250 watts
under Standard Test Conditions).  The $/watt are much better than the
100 watt panels but they are more than twice the size and they are
twice the weight of the 100 watt panels.  The sizing is obvious, as
solar power depends on surface area.  2.5 times the power needs 2.5
times the surface area.

I have two backup gas generators.  a 3000 watt Champion 120 volt only
and a 5000 watt Generac 120/240 volt.  My third option is waiting for
me to assemble the pieces, with a lawnmower engine and an 85 amp CS130
alternator.  It should be a little more efficient to produce the DC
via the alternator (I know, it's 3 phase AC and goes through a diode
bridge to get DC) but it's still more efficient than running a 3000
watt gen to power a couple of battery chargers to get 850 watts of DC
to charge the batteries - and alternator provides a faster charge
(still 65 amps when the alternator is hot from being loaded).  At full
output, 85 amps at 14 volts is almost 1200 watts.  

This email has been checked for viruses by Avast antivirus software.

Posted by Jim Wilkins on June 22, 2018, 1:41 am
 <ads> wrote in message  

You want your battery clamps well insulated when you grab and remove  
one at full charging current.

When electronic fuel injection came out in the mid 70's I built a  
custom battery fault simulator for [a major carmaker] that could  
generate 100V load dumps in rapid succession, like driving down a  
rough dirt road with a loose battery clamp making intermittent  

The customer's engineers were eager to try their new toy, whose power  
greatly exceeded what they could do with standard lab equipment. They  
hooked up a prototype fuel injection controller and told me to start  
at full power. On the third pulse the prototype's magic smoke escaped.

I had been building test stations for General Electric's 1200V, 1000A  
"hockey puck" SCRs for Buddliners, so this project wasn't unusual.


This Thread
Bookmark this thread:
  • Subject
  • Author
  • Date
please rate this thread