Solar panels in series/parallel

Discussion in 'The Projects Forum' started by DM59, Jul 7, 2010.

  1. DM59

    Thread Starter New Member

    Jul 7, 2010
    5
    0
    I recently bought two 15 watt amorphous solar panels to charge a 100ah deep cycle battery. The kit (Sunforce) was cheap, and I accept that quality is equivalent. I noticed that output voltage from one panel is down about 10-15% from the other, and they connect in series. The charge controller simply limits to about 14volts maximum, so excess power is lost. Panel output voltage may run 16-24v.

    I was thinking that I could use the comparators in a 555 timer to drive a double pole relay to go from series to parallel connections when the output voltage of the low panel rises above, say, 15v. When the voltage drops, it would release the relay to series. That should provide some improvement in power utilization without having to go to a DC-DC controller.

    Now, a couple questions. What happens when two panels at slightly different voltages are run in parallel? Literature says panels have a blocking diode to prevent reverse current. And if current is also low in one panel, that would seem to limit series output.

    This is a non-critical installation, just want to maintain battery charge for incidental daily use (10-20 ah) or power outage. I haven't measured output yet to see if solar can handle it. The amorphous panel does indicate it is providing some charging current in light levels even before sunrise. Just experimenting.

    Also trying to work out 555 circuit. Figure on using 5v zener for vcc, and voltage divider to pins 2,6 to change state. Thinking I may have to isolate sample voltage from the low panel so that strong panel doesn't prevent drop out when two are in parallel--if that is possible. Perhaps I need to run sample voltages to different pins. I would want to pull in the relay above an adjustable 14-15v, but also cut it out below around 14v, and I'm not sure how to set up close voltages to change state. Perhaps a zener reference would help, or raising 555 supply above ground? This circuit needs to work over a ride range of input voltage, since the panel is the supply.

    I appreciate the opportunity to post these questions. Just signed up on list after google brought up some past posts--but not quite on topic. Thanks for any help.

    David
     
  2. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,170
    395
    Is 16-24V output for each panel or when connected in series?
    Does each panel have a blocking diode?
    If the panels are each 15W & 16-24V , the output is about one A each, or in parallel 2A; if so then connect directly to battery & monitor V ocasionally. Battery type, other thad deep dishcharge unknown to us, so use Mfgs data for V vs state of charge. Twenty A would be a reasonable charge rate for 100Ah battery, so 2As might take awhile.
     
  3. nsaspook

    AAC Fanatic!

    Aug 27, 2009
    2,907
    2,165
    30W will just trickle charge a 100Ah battery but you need that charge controller in the circuit to limit the voltage to the battery when it's fully charged. You'll need 5 to 10% of the Ah rating as a charging current to keep the battery healthy. If your panels are both 12 volt types you can safely parallel the outputs to the charge controller.

    Do your panels look like the smaller ones here?
    [​IMG]
     
  4. DM59

    Thread Starter New Member

    Jul 7, 2010
    5
    0
    Further checking today under full hazy sun, output is running 21.8v and 21.1v on the two panels. My mistake, they are connected in parallel. With controller and battery as load, combined voltage at panels is running 14.2 volts, and battery has charged to 13.7v from 12.7 yesterday. Peak charge current, panels in parallel, was about 1.1 amp at battery. So, it looks like all the specs are good.

    Yes, the panels look similar to the smaller ones in your picture. There are 16 parallel ribbons about .75" wide running vertically a with traces at each edge. They are described as thin film amorphous, silicon deposited on glass, not especially critical as to sun orientation.

    Looks like together they are putting out about 16 watts maximum, but that was in late afternoon hazy sun, and battery near full charge. I plan to put them on a flat roof elevated at 30 degrees, latitude here 42 north. Can tilt them up later in season. For $100, its not too bad a start, and it will be interesting to see how much of a daily load I can recharge.

    Thanks for your reply. Sorry my initial concept was incorrect.
     
  5. nsaspook

    AAC Fanatic!

    Aug 27, 2009
    2,907
    2,165
    For $100 it's a good buy to get started. I've got five 15W and two 80W panels but plan to upgrade to about 1kW next year.

    This is a solar energy charging plot using data from my battery monitor as it switches to several batteries during the day.
     
  6. DM59

    Thread Starter New Member

    Jul 7, 2010
    5
    0
    Do you change batteries based on time or charge level? Its nice to have a plot.
    I notice the higher voltage output on second bank. Are you using a DC to DC controller?

    I installed shunt resistors on battery charge and drain, .001 ohm, just #10 copper wire 11.8" length, or 7.4" #12, so I can measure 1mv drop = 1 amp. Kill-a-volt monitor will do on inverter output, to estimate Ah used.
     
  7. nsaspook

    AAC Fanatic!

    Aug 27, 2009
    2,907
    2,165
    It's mainly switched on charge (charge controller in float mode) or the calculated state of charge but has a hard timeout depending on the Ah size of the battery. I use a Xantrex C40, it's a PWM type that adjusts the higher input voltage to the correct charging voltage.

    This plot shows the charge voltage also. The larger charge currents are on two 100Ah cells, the others are two 12Ah cells. I use Hall-Effect sensors to measure current scaled to range by the uC.
    http://www.ampsense.com/AMP SERIES 25-100.pdf
     
    Last edited: Jul 8, 2010
  8. DM59

    Thread Starter New Member

    Jul 7, 2010
    5
    0
    Those sensors are cool, didn't know they existed. So how do you generate the plot?
    It seems the Kill-a-watt does not do well with the output of my inverter, recognizes it as DC, no total wattage, but current agrees with shunt reading. So for now, will be monitoring daily battery voltage to see how much of a load I can sustain. It only dropped to 12.49v overnight from 12.8v
     
  9. nsaspook

    AAC Fanatic!

    Aug 27, 2009
    2,907
    2,165
    The data is collected from my DIY monitor system using RS-232 to download the data from the SD card, I then use openoffice CALC to make the plots. The hardware system drawings are out of date but the software is mostly current.

    http://code.google.com/p/solar-monitor/

    Power harvest plot.
     
Loading...