Re: [pvusers] Dual voltage systems mind teaser

[Jeff Clearwater <clrwater@earthlink.net>]

Hi David and Sheila,

Have you considered just using one solid 24 volt pack sized to do all 
the loads and then using a Vanner battery equalizer?  The unit does 
what you have talked about doing with a MPPT controller charging the 
second pack, but this way you get to utilize all the AHs available 
for all the loads and your voltage will stay higher throughout the 
cycle.  You also get to use all your inputs for all loads rather than 
have some PV dedicated and therefore potentially underutilized.

The Vanner is basically a DC>DC converter that constantly monitors 12 
volt both halves of a 24 volt pack and uses the 24 volt tap to charge 
either of the 12 volts taps as needed, constantly keeping the two 
halves balanced.  They cost about $350, are built very solid and work 
for a long time hassle free.  It's the best solution I've found for 
dual voltage needs.

Hope that helps,

Jeff Clearwater
Ecovillage Design Associates

>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/

-- 
~~~~~~~~~~~~~~~~~~~
Jeff Clearwater
Ecovillage Design Associates
2525 Arapahoe Ave, Suite E4, #280
Boulder, CO 80302
Community and Village Scale Renewable Energy Systems
303-546-0460, clrwater@earthlink.net

Council Member - Ecovillage Network of the Americas  http://www.ecovillage.org
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

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