[pvusers] Dual voltage systems mind teaser

["David & Sheila Knapp" <solar@aeroinc.net>]

Originally we had specified two separate systems for our new solar home. The
main house system would be 24 volts nominal and the office/ham radio system
would be 12 volts nominal.

I specified the main house system to be 24 volt because the daily load is
about 3,000 W-H.  The separate office/ham radio system came about because we
have quite a few devices that consume 12 volt nominal power much more
efficiently that using an AC power supply to convert back to low voltage DC.

Although the ham radios run off of what is called 12 volts, their best
efficiency and ultimate happiness rests with running them at 13.5 - 13.8
volts (a vehicles charging voltage).  Several other appliances including our
cell phone charger, answering machine, DC charger (AAA, AA, C, D cells,
etc.), and a citizen's band radio for the local emergency network runs well
from it too.

Initially I specified 300 watts of PV and 415 A-H of sealed batteries at 12
volts for the office system.  The main house system would use ~1500 watts of
PV and a large bank of Rolls batteries located in the utility area of our
solar earthship. The smaller 12 volt nominal office system with the sealed
batteries would be located close to the ham radios in a Rubbermaid container
for a minimum voltage drop.

It seemed like a workable solution until I examined the charging cycle over
a 24 hour period.  The 12 volt nominal system would use MPPT charging to
bring the sealed batteries up to 13.8 volts during the day when the sun was
shinning and allow it to drop to resting voltage at night.  Most of the
constant loads were small, but did add up over 24 hours.  The larger loads
from the ham radios (up to three hundred watts on transmit) could potential
bring the battery voltage below 12.0 volts at night when the radios would be
used the most when the charging would already be finished for the day.
Although the battery voltage would come back up after each transmit cycle,
the radios are not very happy working down there.  They much prefer the
automotive standard of 13.5 volts.  I have heard many backwoods radios emit
distortion when running from an RV battery during extended operation as the
battery voltage sagged.  Even during the day, the peak charging is only ~13
amps (plus any MPPT boost). The ham radios draw 20 amps on transmit plus the
smaller constant loads.  I briefly considered using AC power supplies for
the larger radios, but did not want to add an extra inverter and was
bothered by the efficiency loss of so many power conversions.

I then considered using AC from the main house inverter, but then I knew I
was stealing power away from the house system and letting it go to 'waste'
in the office system.  Then an idea was born, why not combine the two
systems into one bigger system.

The new design called for 1800 watts of PV wired up for the main 24 volt
nominal system in two array's feeding a couple of MPPT charge controllers to
a Roll's 1,900 A-H battery bank.  There are a several 24 volt nominal DC
loads (booster pump, emergency DC lighting, refrigerator, etc.), but the
majority of the DC loads are for the office and require 12 volts nominal.

I had the idea of using another MPPT charge controller to convert the 24
volt nominal to 12 volt nominal and to use a small 210 A-H 12 volt battery
bank as a buffer.  It would do this at 97% efficiency and I could precisely
control the output voltage 24 hours a day as long as I stayed under the 50
amp limit of the charge controller (which was no problem).

The big problem I have run into so far is have both systems grounded to the
same point.  This arises from the negative wires on both systems being
grounded as required by code.  This effectively messes up the function of
the MPPT controller in the office system which requires an isolated input.
The ham radio system has to be negative grounded as all of their chassis
connections are connected to the negative wire as also are all antenna
shield connections and all antennas.  Even using a Trimetric meter with an
external shunt to monitor the charge status of the office battery system
won't work if I understand its operation properly.

The NEC as adopted in Colorado requires a ground on all current carrying
wires that aren't fused, yet one side must be grounded if I understand the
NEC code properly.  Further, by having a common negative ground as I
described it, I have set up the potential where the ground could carry the
current intended for the negative wire because of the loop I had now formed.
Changing the MPPT controller to a voltage regulator (24 to 12 volts) seems
to not change anything unless I go to a regulator that has a common negative
for both input and output.  At that point, using a shunt on the buffer
battery appears to suddenly seem like that it will work okay.  It is only in
the controllers that require an isolated input that dual voltage systems get
into trouble.

I am only giving up some efficiency in the charge regulator of the
unisolated input negative type unless I can find a higher quality one that
is using high frequency DC switching. It looks like the mechanics of it will
work, now I just have to convince the CO state code inspector that the
smaller ham radio system subset is a legal system with regards to being able
to shutdown the main system and the second smaller system having it's own
battery disconnects (as well as the 24 volt breaker feed at the main
system).  It is impossible and impractical to install it all right at the
main system within six feet of each other as the office system is some 25
feet away located right at the point of use.

Once I resolve all of these issues, I'll have to document it.  I cannot
believe there already has not been an article in HP about similar
circumstances such as these.  As more people mainstream into RE, it is
likely that this will be encountered again and again.  A non-gadget guru
would not understand the problems they would face in this scenario.  Even
several dealers called it unusual as well as the MPPT charge controller
manufacturer who suggested we simply leave the 24 volt system ungrounded,
which I am not about to violate code or go to the expense of adding dual
breakers for all 24 volt negative and positive wires in a system that daily
consumes 3+ KWh.  I'm sure there are some wrenches out there that might have
solved a problem or two like this and still kept the code inspector happy.

;-)

Dave Knapp
http://www.esolarliving.com




--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*
  List sponsored by CREST/REPP  
	http://solstice.crest.org/   http://www.repp.org/