Connect with us

Solar Charger Question

Discussion in 'Electronic Design' started by RST Engineering \(jw\), Nov 1, 2007.

Scroll to continue with content
  1. I'm using one of those Harbor Fright 1.5 watt solar panels. Testing in the
    late October sun shows an open circuit voltage of 21.5 and a short circuit
    current of about 60 mA. Maximum power was at an output voltage of 12.7 and
    38.5 mA...about half a watt. Probably double that in the noontime summer
    sun, but don't ever believe that a HF device will meet the published
    specifications.

    My calculation shows that I can model this little rascal as a 21.5 volt
    voltage source with an internal series resistance of 220 ohms or thereabouts
    if that is what I want to do.

    The recombinant gas "12 volt" battery I'm using likes to be taken up to
    13.6, then allowed to self-discharge back to 12.6, then up to 13.6 again...
    ..

    I had originally planned on using a switching power supply to get the last
    milliwatt of power out of the little panel, but (a) the voltage is so close
    to the desired voltage that it seems a waste to go that sophisticated when a
    simple linear PNP pass using an LM324 as the error amplifier plus another
    section of the 324 as a comparator with hysteresis to shut the current off
    when the voltage hits 13.6 and then back on at 12.6. A 3-silicon diode
    reference to attempt to "track" the desire of a lead-acid battery to charge
    to different voltages summer and winter.

    This is my first shot with a solar controller, so if anybody has any pearls
    of wisdom they'd like to lay on me before I start reinventing the Edison
    valve, I'd appreciate hearing from you in this ng.

    Thanks,

    Jim
     
  2. Why not just discuss it here or give a url to the information so that
    EVERYBODY that wants the info can have it rather than a one-on-one?

    Thanks,

    Jim
     
  3. LVMarc

    LVMarc Guest

    Nothing wrong with the HD solar panel.

    you will find the VOC and the ISC, as well as the R thevinin, changes
    dynamically with solar flux. If you are going to maj\ke a commmercail
    device and want to optimize the power pumped out of this or any cell.
    PLease contact me as our group has optimize the solar cell to battery,
    (or any other load) interface.

    best regards,

    Marc Popek
     
  4. Because you can't commercialise an answer that's already been distributed
    via usenet. :)

    I'd be wary of trying to 'optimise' this type of solar array, not that
    there's anything wrong with it, just that it's small. Unless your switcher is
    REALLY efficient and has virtually no overhead, you're going to lose more in
    the conversion rather than just pump and dump into your SLA battery and do
    very crude voltage limitation.

    Something like a schottky diode to prevent current going back into the array
    at dark, and a zener for crude voltage limitation. Many implementations do
    away with the voltage limitation, as they find there is enough buffering
    within the battery to not overcharge anyway (on small arrays of course).
    Many small panels in fact do away with reverse current protection at all,
    and recommend you babysit the thing during daylight hours (which saves the
    drop over the diode to help charge an itsy bit faster.

    The larger array controllers use either a FET as the switching mechanism, or
    a relay over the semiconductor switching mechanism to try to keep the
    impedance down, but there's overhead of the control electronics to cater for,
    and for a small system, it's really not worthwhile - losing even a little bit
    is too much compared with what you have to start with (very little).


    In my opinion, "proper" regulation (of any type) doesn't become economically
    worthwhile until you're dealing with tens of watts.
     
  5. Joel Koltner

    Joel Koltner Guest

    What he's talking about is (most likely) a so-called "maximum power point
    tracking" charger. If you Google the acronym, MPPT, you'll have lots of hits.
    Essentially you purposely change your operating point between the solar cell
    and the load to see whether you're not getting more or less power -- this is
    usually done with a microcontroller, as when using discrete op-amps/logic/etc.
    it can be difficult to avoid building oscillators. :)
     
  6. Joerg

    Joerg Guest

    Essentially an oscillator _is_ what you are building but one with
    limited excursions and riding around the peak. The kids call that
    dithering I guess. Dang, now I revealed that I am over the hill.
     
  7. Perspective gentlemen. He has 38mA to play with. How much of that do you
    think is going towards the micro and support logic?
     
  8. mpm

    mpm Guest

    I agree with this post completely.

    The only thing I would add is that you check the zener rating for
    conditions where the battery is NOT connected. (such as changing out
    the battery during full sunlight conditions and where the load is off,
    or drawing minimal current.) This could damage the zener. A small
    ohmic value series power resistor should fix it though. -mpm
     
  9. Yeah, I thought of this, however, when you're looking at about a half watt
    total power of the array, you're not going to kill any zeners that way.
     
  10. PhattyMo

    PhattyMo Guest


    Exactly.

    I've got a ~10W panel,and don't feel the need to bother with a charge
    controller.
    Aslong as the charge current is like C/10,or under,you don't even NEED a
    charge controller. The battery can take a constant trickle charge.

    This,of course,assumes you have enough battery capacity to deal with the
    solar panels output,which shouldn't be an issue with a small panel,and a
    SLA of any reasonable size.
     
  11. Martin Brown

    Martin Brown Guest

    You might well be better off putting a pair of similar sized mirror
    plastic sheets either side to form a hexagonal solar flux concentrator
    around it - like a trough \__/. Pointed at the sun it roughly doubles
    output. No fancy electronics can turn half a watt into anything
    worthwhile - but mechanical hardware improvements can get you 1W from
    the same photovoltaic array at little extra cost but some increase in
    windloading.

    If you are prepared for slightly more exotic engineering solutions a
    pair of parabolic curves with axis perpendicular to the collecting
    surface and focus placed at the opposite edge is about as good as non-
    imaging flux concentrators get. A good one can manage about 10x but is
    big and ugly - arguably the law of deminishing returns sets in well
    before then unless you are struggling to detect very weak single
    photon signals in a big volume with small expensive detectors.
    There isn't really enough power output to be worth the effort of
    regulating. Think of it more as a trickle charger and defend the
    battery against leakage current back through the solar cells at night
    with a low voltage drop diode, and a suitably rated zener to stop
    overvoltage.

    I'd be surprised if 500mW applied for half a day every sunny day would
    ever inconvenience a decent sized lead acid battery. But my intuition
    could be wrong here. 10W and above would need a controller.

    Regards,
    Martin Brown
     
  12. mpm

    mpm Guest

    You know, strictly as a convenience to anyone who may be looking for
    small wattage solar panels, in the US, there's a company called
    "PowerUp". I believe their URL is www.powerupco.com

    They carry aluminum framed, glass covered solar panels with attached
    cord (and on some models, integral blocking diode), in sizes as low as
    1-watt.

    I do not own stock in this company nor do they pay me to mention them.
    I only mention it because on a recent project, I almost pulled my
    (remaining) hair out trying to find a vendor who could actually
    deliver small quantities of commercial grade, low-wattage panels.

    We used them, and had no problems.
    Note, several manufacturers no longer make the small panels (evidently
    due to conversion of factories to supply the European demand that is
    being driven by some legal environment there?)

    -mpm
     
  13. Joel Koltner

    Joel Koltner Guest

    Easily under 10%, and pobably <1mA with some effort (i.e., lots of "sleep").
    I would agree, that, that it's probably kinda iffy as to whether or not it's
    worth the effort -- I was responding more just to the one "tease" response
    telling the O.P. how he could wring some more power out of his panels, that's
    all.

    If he already has a microcontroller anyway, of course, that definitely alters
    the balance of desirability...
     
  14. LVMarc

    LVMarc Guest


    Often, a great controller can be realized with analog control loop! Not
    every thing requires a micorporcessor. and in the case of managing
    modest sized solar arrays and optimizing the power harvested,you DO NOT
    want the excess power of a micro, unless it is absolutely necessary!

    current consumed by the power optimizer is less than 1 mA!

    Marc
     
  15. Joerg

    Joerg Guest

    If you do that with CD4000 logic or maybe a MSP430 with a micro-power
    regulator it should be feasible. At least it would be a nice challenge
    from an engineering perspective. I have to deal with this kind of
    situation a lot. No power to speak of available but the client wants
    this, that and the other thing user programmable. And since the client
    is king I do my best.

    It is not impossible to increase efficiency via MPPT even with a grand
    total of 38mA or less. Of course in the end it could cost more than
    buying $2 worth of more solar cells.
     
  16. Hereteic! A square wave oscillator is "pushing" the set point of the usual
    regulator up and down. Then a synchronous rectifier is used to move the
    regulator set point via an integrtator. the maximum power point is when the
    up/down steps reverses i.e. the controller end up "siting" on the peak of
    the power curve.

    I.e.100% analogue should work just fine.

    Should really slap a design on the web together with a design for a resonant
    power supply for the "i need EHT people". Maybe Later ;-)
     
  17. Joerg

    Joerg Guest

    This one I'd also do analog. For larger installations a uC can make
    sense because they also need time tracking, calendar, sun position and
    all that for the big motor controllers that steer the array heading.
     

  18. Joerg & Fritz ...

    Could you flesh out your thinking a bit? My control theory class was a
    LOOOOOng time ago, and I've never reduced it to practice.

    Audio and RF on the other hand ...

    Jim
     
  19. Joerg

    Joerg Guest

    The energy yield versus voltage increases, crests and then goes back
    down for higher voltages. For maximum energy yield you want to stay
    around the peak. Problem is, that peak also moves about depending on
    clouds rolling through and such. So, for example, you can sweep and then
    dither around that peak and the electronics can track it. Basically
    you'd be looking for the point where the gradient of the sweep is the
    least, meaning total power drops on either side and hence the peak must
    be in the middle. Now you steer your switch mode stage so that it pumps
    the maximum juice into the battery or grid while keeping the input at
    maximum power point. Unless your are located off grid and the batteries
    are full in which case solar panels are often shorted.

    The last one I built was not for solar and I can't disclose anything but
    except for min-max setpoints this was all analog. It had to be fast and
    also was cheaper that way.

    It's called maximum power point tracking or MPPT. With solar stuff there
    isn't all that much detail in their "How it works" section but at least
    a little can be found here:
    Probably universities have more but I had been in a rush so I just went
    ahead and designed it ...

    My favorites as well. However, sometimes we must eat pea soup ;-)

    In RF speak MPPT could to some extent be compared to automatic antenna
    tuners. There, you have to find the point of least VSWR. But in RF we
    mostly have phase information so we (usually) don't have to dither.
    Except when the "load" isn't free air but some odd object in which case
    it all needs to become more nifty (but still not necessarily digital).
     
  20. Joerg

    Joerg Guest

    Yes, they have some huge incentives. I know people in Germany who
    installed panels and rake in grid backfeed kWh rates several times
    higher than what a kWh normally costs. Of course, that is a hidden tax.
    There is no free lunch and usually the other "normal" ratepayers must
    subsidize this. Or the taxpayer does. Or both.
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-