Connect with us

Charge capacitor to a specific voltage

Discussion in 'General Electronics Discussion' started by Michael Berg, May 21, 2017.

Scroll to continue with content
  1. Michael Berg

    Michael Berg

    9
    0
    May 21, 2017
    Hello Forum,

    This is my first posting so be gentle.

    I'm looking for a way to design a circuit that would allow a solar panel to charge a capacitor to a specific voltage, and then bleed off additional extra current through a resistor in order to maintain the cap at that specific voltage.

    Let me briefly explain the background for that. Solar panels are funny, in the sense that they produce most of their power at a specific voltage. You can read this voltage on the back of a panel, under the label Vmp. Vmp is around 80% of the Voc of the panel - for example it could be around 18.5V for a 21V panel. Basically, if you are able to present the panel with a load that consumes all the current at 18.5V, you get the maximum power out of the panel. This is also why most "12V battery charging" solar panels you find on ebay are around 18V if you just measure the open circuit voltage with a multimeter - they provide most of their current at 12-14V which the battery pulls the panel down to.

    So I connect the panel to a capacitor with a sufficiently high voltage rating. As the sun illuminates the panel, the cap begins to charge, which can be measured directly with a multimeter on the cap ends. Left unattended the voltage would rise to the panel's Voc, or 21V. But I don't want this, I need the cap to charge to 18.5V and no higher, even if this means bleeding off solar power through a resistor. This way I can connect my load to the cap and have the maximum power from the panel available to me because the cap holds the voltage at the Vmp of the panel.

    My idea is to use two transistors. The first transistor decides, based on the cap voltage, if power from the panel should go to the cap or not. A second transistor is wired in the opposite way, i.e. to open when the voltage exceeds 18.5 and allow current to flow from the cap + to the cap - through a resistor. Like a safety valve of sorts. As the voltage rises, these two transistors work together to keep the cap at 18.5 by either charging it or discharging. The bleed-off resistor probably needs to be able to handle quite a bit of power.

    The important thing to keep in mind is that when I measure with my multimeter what the voltage is across the panel, it must read 18.5. This is why a simple voltage divider is unsuitable for this project.

    I have already come up with a design that uses an arduino and a voltage sensor to control a relay to achieve something like this, but it is entirely overengineered, too power hungry and just too elaborate. It feels to me as if this MUST be possible with discreet components, I'm just not good enough to come up with a design.

    Can anyone help with specific designs or just suggestions on how to approach this problem?

    PS - yes I realize the Vmp changes in different lighting conditions and that an MPPT controller handles this by continuously scanning the IV curve for the optimal Vmp. For this project, simply assume the Vmp is known and fixed.
     
  2. duke37

    duke37

    5,364
    772
    Jan 9, 2011
    A diode from the panel to the capacitor will be all that is necessary to stop power going back into the panel when the light is low.
    To dump charge, you could use a zener diode at the voltage you want to limit. Most such diodes will be low power types but can be amplified with one or more transistors. Resistors in the transistor collectors should be used to dissipate most of the heat.
     
  3. Michael Berg

    Michael Berg

    9
    0
    May 21, 2017
    Thanks for the suggestion on the diode from the panel. That's a good idea, although it doesn't matter so much since my project only needs to work during the daytime.

    The idea with the zener diode sounds interesting. I don't know much about those things, but they are basically diodes that block until a specific voltage, right? And then they begin to conduct? If this is the case then it certainly sounds like this would be perfect for bleeding off charge.

    Ẃhere can I read more about zeners and how to control them specifically for this project of mine?

    Thanks again for your reply!

    -Michael
     
  4. duke37

    duke37

    5,364
    772
    Jan 9, 2011
    Wikipedia has a chapter on zener diodes.
    You will need aditional circuitry for high powers or high accuracy. Specification required.
     
  5. Michael Berg

    Michael Berg

    9
    0
    May 21, 2017
    Well my plan is to design my circuit in one of two ways:

    1) A zener in parallel with one or more capacitors. Upon reaching 18V, the zener conducts and shorts out the cap until the voltage drops below 18V again. Load is connected to the cap and powered directly from it.

    Observations:

    a) Cap will not go higher than 18V
    b) Load will drain the cap below 18V which is not ideal but acceptable.
    c) Zener will have to absorb potentially all the power from the panels, which could be > 50 watts. The zeners I've seen aren't rated much higher than 13-15w. Could I combine multiple zeners in parallel to handle a larger current?

    2) A zener attached to the + side of the cap, and the other side feeding the load. As the cap voltage reaches 18V, the zener will open and allow the load to operate.

    a) Cap will go higher than 18V if the zener and the load can't absorb it all at once. This is not ideal but it is a compromise I can live with.
    b) Zener may provide uneven power to the load. As soon as the zener opens at 18V, the load will drain the cap and the cap will fall below 18V unless the panels can immediately compensate for the drain. If the zener closes, the cap will rise above 18V again very quickly, causing the zener to open and repeating the process. Doesn't this give a kind of "oscillating" power to the load? How might I compensate for this?
    c) Zener only has to absorb as much power as the load takes. In my case, around 10 watts. This is well below popular toleances so I shouldn't have any problems finding a suitable component.

    - Michael
     
  6. Michael Berg

    Michael Berg

    9
    0
    May 21, 2017
    I decided to take a different path. Turns out there are plenty of pre-built, cheap alternatives that allow me to maintain a fixed voltage across the solar panels. For example:

    http://www.ebay.com/itm/122513080856

    One simply sets the desired voltage and the unit will vary the load (not sure how - presumably by varying the duty cycle on the internal converter somehow) and present that to the panels, allowing them to operate at that voltage - and providing the maximum power. Assuming a fixed Vmp which is not the case since it depends on lighting and temperature, but its good enough for my purpose.

    -Michael
     
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

-