Help needed with TL431CP regulator

Discussion in 'General Electronics Chat' started by spinnaker, Mar 27, 2010.

  1. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002
    MikeML was nice enough to provide the TL431CP circuit below. It is intended to regulate the output of a solar panel to charge a 12V SLA battery.

    I finally got everything shipped in and wired up but it is not working.

    I checked everything over a couple of times and I am pretty sure I have everything hooked up right.

    The best I am getting is .9 V. I am getting this with a battery for a load and no load.

    If I disconnect the MOSFET and measure my voltage at the cathode of U1, as I change the input voltage from 2.5 volts, the voltage on the cathode changes until around 13.2V is reached, then the voltage drops to 2 volts.

    I assume this is normal? The diode is turning on at this point?


    I guess maybe I really don't understand the whole circuit here. What is really confusing is how my unregulated voltage is on the same bus as the regulated voltage.

    Can someone please help me understand what is going on here and help to troubleshoot this circuit?



    Note: for the FET, I am using a SPP18P06P.

    One other note: I noticed that R1 gets a bit warm. Is this normal?




    [​IMG]
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Try increasing R1 to 2k to 2.5k.
     
  3. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    I see a couple of problems:
    1. The solar cell needs a diode in series with it to prevent the battery from discharging into it at night.
    2. If you have a fully charged battery, and adjust the regulator for a lower voltage, the MOSFET will suck a lot of current, possibly destroying it.

    How can you get 0.9V when you have a battery as a load?
    What are you using to emulate the solar cell? A power supply?
    Why do you start another thread every time you have a question on this topic?
     
  4. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002

    Tried a 2.3K I have on hand.

    Still .9v out.

    I did notice that the MOSFET is getting a bit warm not hot. Could have been happening before I swapped resistors.
     
    Last edited: Mar 27, 2010
  5. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002
    Yep already in my plans.




    Battery is actually fully discharged . But now that you mention it, the MOSFET is what I call warm but not hot.

    I did not yet heat sink it, maybe I should have? I only had the thing powered up for a few minutes at a time.


    That was my question. :)


    An LM317 regulator that I soldered up.
     
  6. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    You should have it configured as a current source, and you should have the diode (the one in your plans) in series with it. The LM317 should be powered by at least 18V, and iit will get HOT if the battery is not charged, unless you run it at low current.
    Since the battery is discharged, could this account for the 0.9V?
    Also - if you have the MOSFET source and drain swapped, you will get ≈0.9V, and it will get warm or hot.
     
  7. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002
    The LM317 is temporary. Just for testing.

    You sir are a genius!

    Or as an electronics instructor from many moons ago used to say "You sir are a scholar and gentleman".


    Apparently I went over everything 2 or 3 times EXCEPT the pinout of the MOSFET! Duh :rolleyes:
     
    Last edited: Mar 28, 2010
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    There is no hysteresis in that circuit, which will result in the battery being overcharged, and a short battery lifespan. SLA batteries do not take overcharging well at all.

    Charging being limited to a lower voltage level will also result in a short battery life if it is not completely charged by the time the sun fades.

    Once the battery is charged via constant current to a certain voltage level (say, 14v-14.5v) the voltage should be reduced to a "float charge" level, say 13.3v-13.7v.
     
  9. eblc1388

    Senior Member

    Nov 28, 2008
    1,542
    102
    I would place the diode in series with the lead to the SLA battery instead of the solar cell.
     
  10. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002
    So it sounds like my plan to charge just at trickle charge is just about as bad as charging all out.

    I'm thinking I might just go back to my plan of purchasing an off the shelf charger.
     
  11. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002
    And I take it back

    Ron H is still scholar and a gentleman.

    The regulator is indeed working. I can adjust the trim pot to a nice 13.8 V or whatever I want.

    But the FET gets blazing hot!! That is a lot of energy that could otherwise charge the battery.


    There is a product out there from Morningstar Technologies called the Sunguard. It uses PWM to control the charging. It runs about $30. Unless someone can help be salvage my little circuit above, I think I will go with the Sunguard.
     
  12. k7elp60

    Senior Member

    Nov 4, 2008
    478
    69
    I wouldn't give up yet. You can float charge a SLA battery from 13.5 and 13.8V and never overcharge it. Most solar panels will provide excess of 18V in full sunlight. You could still use the LM317 on a heat sink to charge your battery, then you wouldn't have to buy another charger. If you can tell me the AH rating of the battery I can tell you if you will have to limit the current on the LM317. That is easy to do also.

    At first I thougt you could use the P channel mosfet and the TL431 as a charger, but then I realized the gate to source voltage may not be enough, thats why I recommended the LM317.

    I also recommend a diode between the charger and the battery, to prevent the battery from being dischared when the solar panel is not supplying adequate voltage to the charger. A schotty diode as a lower voltage drop, but a regular diode will also work.
     
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    One of Spinnaker's problems is that the battery and charging circuit are going to go into a somewhat small, sealed plastic box. There won't be any way to remove heat from it, except by conduction through the box itself.

    A clamp will be OK until it hits the clamp limit - then where is all the heat going to go?

    Let's say (just for grins) that the clamp voltage is set for 13.8v; we'll ignore the diode Vf for now.

    Let's also say that the solar panel is sourcing a constant 2A current, up to a maximum of around 22v (I can't remember exactly, but I think this is in the ballpark of his solar panel specs).

    So, the battery charges up and reaches 13.9v. Then the clamp starts taking effect. Initially the power dissipation won't be too bad as the battery will still be absorbing a charge - but after awhile, the power dissipation in the MOSFET will get to be more significant.

    If the battery eventually reaches full charge, and the clamp is taking the full load, then dissipation will be around (22v-13.9v)*2A, or 16.2 Watts. With no place for that power to be dissipated, the MOSFET and box will get mighty toasty mighty quick.

    An LM317 solution would not help, as it would dissipate power whether or not it was actually limiting current or regulating voltage. The LM317 has a minimum dropout of 1.7v. If it were passing 1.5A current, that's 1.7V*1.5A = 2.55W. That's not all of it though, because as Iout from the regulator increases, Vdrop increases along with it. With an output around 1.5A with the regulator at around 100°C, Vdrop will be around 2.3v; so power dissipation will be around 3.45W. Even that's optimistic, because as the regulator's temperature increases beyond that, Vdrop also increases. It won't be long before the 317 goes cycling in and out of thermal shutdown.
     
    Last edited: Mar 28, 2010
  14. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002
    I have a TL431 (actually two). They are 8 pin dip.

    My panel was putting out about 23V @.6 amps open circuit.

    The battery is 12V 5AH
     
  15. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002
    As I said the panel puts out about .6 amps open circuit. But we should have some "growing room" should I decide to replace the panel with something a bit beefier.

    I was considering using an enclosure that was plenty big enough. No way do I want to go back to try and fit everything into the light itself, like the original design.

    I as also going to try and fashion some sort of vent that would keep out rain, though a bit of a concern is that the vent might allow condensation.

    This is the off the shelf solution I was considering.
     
  16. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    So is it 23V at 0.6A current, or 23V in open circuit?
     
  17. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002

    What I meant to say is it is 23V @ .6a with no load. I did do some testing with various loads. I will see if I can find the results.
     
  18. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002
    Here we go

    No load 22.7V .6amps
    3 ohms 22.4V .37 amps
    25 ohms 22.5 .32 amps
    50 ohms 23.1 .21 amps
    100 ohms 23.1 .12 amps
     
  19. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Good stuff, Wookie.
    My comment about using the LM317 as a current source was a response to Spinnaker, who said he was using it to emulate a solar panel. I wanted him to know that using it as a voltage source will not emulate a solar panel.
     
  20. spinnaker

    Thread Starter AAC Fanatic!

    Oct 29, 2009
    4,884
    1,002
    But on the solar panel. Does not the voltage increase and decrease depending on the amount of sunlight? Or does the current continue increase (to some level) after the maximum voltage is attained?
     
Loading...