Author Topic: LOW POWER/COST LM2596 POWER POINT REGULATOR  (Read 65635 times)

0 Members and 1 Guest are viewing this topic.

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
LOW POWER/COST LM2596 POWER POINT REGULATOR
« on: May 03, 2012, 02:58:14 PM »
LOW POWER/COST LM2596 POWER POINT REGULATOR
I have bought a number of these LM2596 regulators  from China. They are a convenient do anything regulator.   At way less than $3 each including shipping , it doesn't pay to build one.  Pin #5 of the LM2596 can be used to turn the regulator on and off and it is easily unsoldered from the circuit board.  Adding a single NPN transistor, zener, and a few other resistors will turn this into a power point regulator.  It will not track but it provides most of the benefits if adjusted a couple of times a year to correct for temperature.  If a string of diodes thermally attached to the panel replaced the zener, the regulator would actually track for temperature.

The listing claim of 3A is based on what a LM2596 could work at in a well designed circuit with adequate heat sink.   The thermal resistance of the circuit board itself limits maximum current to about 1A with a large heat sink added (as pictured) or a couple hundred mills without.  A majority of the circuit board bottom is ground plane.  Only the top corners where the IN+ and OUT+ connect and the voltage adjustment pot need to be isolated from a heat sink.  The extended leads of the pot can be filed down and a small piece of tape applied as insulation.   Then the board can be glued to a heat sink.  If kept thin, hot melt glue works as well as anything.  Keep in mind that hot melt glue sticks well to some surfaces and not others. Thermal conductivity is sufficient since you are limited by the circuit board.   The heat sink should cover the entire board because the diode and inductor also produce heat.   At least 1,000uF capacitors should be added to the input and output to handle the current.  Don't count on the board capacitors as they are sized just to to keep the IC stable.  A much  larger cap at the input would be better because it has to store panel energy between pulses.   For best results the zener should be a significant part of power point voltage.  The 87K resistor voltage divider on pin 5 prevents the voltage from ever going above the 25V max voltage .    Measure input voltage and adjust 5K pot for the approximate power point.  To obtain the actual power point , place an amp meter in series with a battery that is not fully charged.  Adjust pot for maximum current.

 One downside is the input to the LM2596 is limited to 40V.  Many panels will reach 21 or more volts each with no load.    Generally small  panels can't provide much current in this region.   A string of three 1W zeners  in series along with the optional zener  on the schematic would limit the maximum voltage and have sufficient dissipation.  Total voltage of all the zeners  should total 40V for a 24V panel.  The optional zener should be closer to 10V instead of 4.7V  View ebay pictures carefully. Some versions have an input cap rated at only 35V instead of 50V.  For higher currents multiple LM2596 assemblies  can be placed in parallel.  Using a blocking diode for each module will provide load sharing.   Pin can be connected together so there is a single input adjustment.

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #1 on: May 03, 2012, 03:01:05 PM »
THE TEST
I used a simulated  typical 12V solar panel voltage of 17.0V and an output after the blocking diode if 14.0V.  Load was a 15 ohm resistor.  That produced at output  a wattage of 13.02W at .93A.   The input  current  for this was .91A at 17.0V or 15.47W.     13.02 / 15.47W represents an efficiency of 84%.   If a blocking diode is not needed, that efficiency will increase.  Solar panels are very close to constant current devices for most of their range.  The test conditions are similar to an 18-20W panel on many days.  If a zener or ghurd shunt regulator with blocking diode is used instead a switching regulator the result would be .91A at 14.0V or 12.74W output.    A switching regulator seems like a lot of effort to gain just  a quarter watt!   Of course, things get better if you have higher panel voltage or the battery is discharged.   I included this comparison to demonstrate a point.  Numbers that look good don't always achieve the results expected.  Charging lower voltage batteries is an ideal application because it allows you to use standard 12V panels without the losses of a linear regulator.  This module can adjust the output down  to   1.2 volts.

This specific application information was requested.  The LM341 adjustable zener could also replace  both the zener and the transistor.   There are other methods to control the input voltage of switching regulators by fooling the feedback sense pin with a higher voltage.  Current limiting the output can also be achieved in the same way.  Parallel regulators are something else that should be explored.  There is a reason why cars don't come with an engine with just one big cylinder.



vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #2 on: May 03, 2012, 10:12:45 PM »
Interesting.  Can you explain how the pin-5 circuit works?  I don't seem to get it.  If the panel voltage is high enough, it'll start up the converter.  But then what?  If the current draw pulls the panel voltage too far down, it'll shut off the converter.   So in actual operation, is it constantly cycling between on and off?  At what rate?  How does that achieve higher output then leaving pin 5 grounded for constant operation?  Or can an intermediate voltage on pin 5 keep the converter in a not-quite-fully-on state, constantly drawing just enough current to keep the panel voltage intermediate?

Also, since the converter aims for a specific output voltage, how does it react to the battery voltage being below the setpoint?  Does it push as much current through as it can?

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #3 on: May 04, 2012, 06:25:20 AM »
There is a whole catagory of regulators out there now that I call "Wham Bam" regulators like the TNY series.  They have a control pin that is either on or off.  The pulse duty never changes, just the number of pulses.  If is much more efficient  to design a transformer for a fixed duty cycle than for PWM. With 120,000 or more pulses per second,  an operation with 50 pulses on and 100 pulses off doesn't look any different to the end user than PWM.  I believe any recent switching regulator operating over 30KHZ uses this method. 

If the battery voltage is 12 instead of 14 it does push out everything it can till the input drops below 17V.
« Last Edit: May 04, 2012, 06:40:09 AM by OperaHouse »

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #4 on: May 04, 2012, 09:24:41 AM »
Download a datasheed for the TNY 263 for a full explanation.   These are designed to work at line voltage and they make versions 263 to 268 I think.  One of them has a FET with an on resistance around 25 ohms.  Shockingly high but enough to drive a common mode line filter and with proper wave shaping a power FET.  That could make good size regulator.   Problem is they don't work at low voltage.  Here is a secret.  Attach a 20K rexsistor to bypass pin and input supply+ and they work fine.  These chips can be found in many wal wart adaptors.  One of those data sheets also has the trick to current limit the output.

vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #5 on: May 04, 2012, 10:21:44 AM »
Thanks, but I still don't quite get it.  These regulator chips are designed to regulate the output, and a fast on/off "wham-bam" followed with filtering is OK.  But in this case you are also trying to regulate the INPUT, in the sense of running the PV panel at some intermediate voltage & current close to its max power point (MPP).  If the PV panel is always either feeding a too-low impedance or an open circuit then it's never at MPP, no matter how fast the switching between those two states?

The saving grace here may be the largish capacitor on the input side (not drawn in the circuit diagram here, but is part of the regulator module).  Is it large enough to smooth out the current coming out of the PV panel?

Example: 1000 uF & 1A.   That will "slew" at 1V/msec, if I got the math right.  If the regulator actually does something like "50 pulses on and 100 pulses off" at 150 KHz that's 1 msec per grand cycle, so it would keep the PV voltage within a 1V band.  Not bad.  Is that really what happens?  What determines the time it stays on and the time it stays off?  Where does the hysteresis come from?

Completely separate question: suppose you have a small 12V PV system and you want to run a 6V device efficiently.  E.g., I have a nice-sounding transistor radio that runs on 6V.  Will a LM2596-based step-down regulator be significantly more efficient for that use than the other options, e.g., a linear regulator (LM7806)?

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #6 on: May 04, 2012, 12:18:20 PM »
Perhaps that wasn't the best example.  I thought it would be good visually for people to understand.  If you look at the TNY 263 data sheet it gives a scope example of it turning on and off every 2 or 3 pulses.  Voltage swings on this board can't be seen with a scope without going to the lower scales on the input or pin 5.  It looks more like a little noise.  Your UNO only does a PWM of 490hZ,  more than enough.  Order emerges out of chaos.   Pin 5 could could use a .1uF cap to get rid of noise, then again a little dither probably helps.  Just don't make pin 5 a big antenna.

Keeping the power point above a set point increases the efficiency when heavy load is connected.  A light load it just doesn't matter if the input voltage rises.  And those caps are not big at all.  Use just the ones on the board and they will be bad in a year.  All my TV and other home electronics were just about free because manufacturers saved a penny on power supply caps.

JUst wanted to note that there are HV versions.  The LM2596HV can input up to 60 volts.  These are a lot more money.  The cheap boards are so handy my next ebay order will be for 10 of them for $18 shipped.

" Where does the hysteresis come from?"

There is far less than I would have expected.  There isn't any need for it when making 150,000 decisions a second.  he capacitor time delay only happens if ALL power is removed.  The capacitor is only making up for what the panel can't supply.  These switchers can be thought of as automatic transmissions.  The efficiency does drop slightly as the output voltage because for the same outpput current the diode loss is constant.
« Last Edit: May 04, 2012, 02:40:46 PM by OperaHouse »

independent

  • Full Member
  • ***
  • Posts: 132
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #7 on: May 04, 2012, 03:54:45 PM »
Thank you very much for posting this. Really appreciated

vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #8 on: May 09, 2012, 11:32:01 AM »
I see on fleabay there are several kinds of this general device.  Most of them are similar, with one active device (the LM2596) and one potentiometer (output voltage adjustment).  But there are a few others.  Some promise a constant-current mode in addition to the constant voltage.  They have 2 more active devices, 3 pots, and LEDs that tell you what it's doing.  And cost more (a whopping $5).  Seems that these may be ideal for charging small batteries off of larger ones.  E.g., if you have a PV system with a large 12V battery and you want to efficiently charge a small 6V battery from it.  Anybody tried it?

For charging NiMH AA batteries (and other single cells) another option is a "smart" charger that runs off of 12V - those that come with a detachable 12VDC AC adapter.  I don't know how efficient they are though.  My favorite smart charger for AAs is the BC700, and it seems very efficient (batteries, charger and AC adapter all run cool), but it runs on 3V not 12V.  The BC1HU is my second-favorite, and it runs on 12VDC, but is probably less efficient?

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #9 on: May 09, 2012, 12:04:24 PM »
There is also a high voltage version, the LM2596HV, which operates up to 60V.  These are more expensive ($6) but would be easier to implement on a 24V panel systemover 20W since there would be no need to add the shunt regulator.

I have the three pot version but havn't traced it out or tiied it.  LM359 dual Norton opamp and 78L05.  They use a short length of circuit trace in neg return to sense current.  Only a 35V cap on the input.

I've been driving a high side fet two ways with the UNP PWM, opto isolator and small transformer out of a modem.

independent

  • Full Member
  • ***
  • Posts: 132
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #10 on: May 09, 2012, 03:27:06 PM »
The 3 pot lm2596 are marketed as power LED regulators or Li-ion cell chargers. I've used one for over a year as a single cell Li-ion charger. I made it to fit into Nokia n900 external battery charging case (a 3 or so dollars from dealextreme). Using a 5v solar panel which provides up to 800mA and 6.5v. The cell is 1320mAh. They work very well for this purpose. It charges the cell to 4.2v fuller (nokia only charge to 4.15v) and quicker than a 1.2A wall charger. The current seems to taper off the closer to the set voltage as it approaches. The only downside is the voltage on outputs rises slowly over time so I have to keep an eye on that. I don't know why it rises like that, every week or two it needs resetting. I wonder if the multiturn pot is not sensitive enough.

High amp low voltage solar panels are difficult to obtain and charging small batteries is a pain without them. The solar panel mentioned will not last long, it is bubbling under the plastic coating visibly after a year of very part time use. If there is a true MPPT solution for say, 3.6v and 6v devices it would make charging electronic devices easy from more commonly available 5w & 10w 12v solar panels.

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #11 on: May 09, 2012, 03:50:58 PM »
I don't think that is enough voltage on the input for reliable operation with a 6V panel.  Both regilators would be operating on the ragged edge.  The sense resistor is the circuit board abd would be vey sensitive to temperature change.  I just started playing with the one I have.  Notice they don't tell you what pot is what.  One seller did but they appear to be gone now.

independent

  • Full Member
  • ***
  • Posts: 132
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #12 on: May 09, 2012, 05:42:32 PM »
Maybe so, but a year is longer than some turbines last..

vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #13 on: May 09, 2012, 10:39:08 PM »
This image might help identify the pots and LEDs?

I wonder if that module is useful for constant-current operation at lower currents, e.g., 200mA?  Or is the current adjustment too coarse at the low end?

Regarding the use of low-V hi-A solar panels for charging low voltage batteries, isn't the idea of these modules to enable the efficient use of higher-voltage panels for the purpose?  E.g., a "12V" (17V MPP) 3W panel would yield about 150 mA and after conversion to 4V by the LM2596 would be over half an amp?

If the PV panel is that small, for charging a small lithium cell, do you really need the current limiting feature, or can you use the plainer LM2596 circuit?

vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #14 on: May 10, 2012, 03:20:38 AM »
Seems like I've found a schematic of the CV/CC version?

For lower output currents, and a non-adjustable current limit, when there is enough in/out voltage difference to spare another 0.7V, how about the following very simple current-limiting circuit: add a small sense resistor in the positive output wire, between Cout and the feedback loop tap.  This current-sense resistor should develop a 0.7V voltage drop at the intended current limit.  E.g., a 2-ohm resistor for a 0.35A limit.  Connect a PNP silicon bipolar transistor: emitter to the supply side of that resistor, base to the load side, and collector, via a resistor (and/or an LED as in the circuit mentioned above), to the feedback pin of the LM2596.  Disclaimer: I havn't tried this yet, and it may cause instability...

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #15 on: May 10, 2012, 04:11:00 AM »
That schematic is probably close, but with the cc pot connected to Vout it should have foldback dropping the voltage quickly.  If connected to the 5V regulator it would be CC.

boB

  • Sr. Member
  • ****
  • Posts: 389
  • Country: us
    • boB
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #16 on: May 10, 2012, 04:03:52 PM »
Looks like you could regulate the input voltage instead of the output voltage by adding a fairly simple op-amp circuit, (i.e.Some offset, some inversion and some gain reduction), between the feedback pin and the input voltage.


vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #17 on: May 10, 2012, 04:34:56 PM »
boB: I think you are referring to the original posting in this thread, how to keep the input voltage at the max-power point?

I see two separate decisions that need to be made on that:
(1) use the on/off pin (wham-bam regulation) or use the feedback pin (smoother).
(2) use a fancy op-amp circuit (like in the constant-current example) or use a crude zener-based approach (as in the original posting).

I think regarding (2) that for the proposed usage there is no need for a fancy circuit, since it won't be the exact MPP anyway, but the power curve near the MPP is fairly flat so you get most of the benefit from a simple circuit.

I don't know if the decision on (1) makes much difference, assuming that there is a way to do it via the feedback pin.  Perhaps not, since that would primarily affect the output voltage, and we want to keep that constant to fit the battery.  Just like a "real" MPPT controller, both sides need (separate) regulation.  The beauty of OperaHouse's insight is that there is a way to achieve that - with a $2 off-the-shelf module plus minor additions.

boB

  • Sr. Member
  • ****
  • Posts: 389
  • Country: us
    • boB
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #18 on: May 10, 2012, 07:07:49 PM »
boB: I think you are referring to the original posting in this thread, how to keep the input voltage at the max-power point?

I see two separate decisions that need to be made on that:
(1) use the on/off pin (wham-bam regulation) or use the feedback pin (smoother).
(2) use a fancy op-amp circuit (like in the constant-current example) or use a crude zener-based approach (as in the original posting).

I think regarding (2) that for the proposed usage there is no need for a fancy circuit, since it won't be the exact MPP anyway, but the power curve near the MPP is fairly flat so you get most of the benefit from a simple circuit.

I don't know if the decision on (1) makes much difference, assuming that there is a way to do it via the feedback pin.  Perhaps not, since that would primarily affect the output voltage, and we want to keep that constant to fit the battery.  Just like a "real" MPPT controller, both sides need (separate) regulation.  The beauty of OperaHouse's insight is that there is a way to achieve that - with a $2 off-the-shelf module plus minor additions.

Yes, but turning things on and off with pin 5 just didn't seem like a good thing to do.  Using the real PWM and feedback pin is the way.

As far as throwing lots of other circuitry in here to get this to really work, that just doesn't seem like it is worth it.
Only reason to do that here as far as I can see is completely for educational purposes except maybe for a very
small PV module and 12V battery.  40 volts (and maybe 60V) max and 3 or so amps for something that doesn't
even MPP track very well has got to be for educational purposes and I completely support that !!  It is becoming
very hard to find good techs and engineers these days and this is one way to help people learn.

If the battery and loads are small, you still gotta add over voltage protection and voltage regulation for Absorb cycles
and timers and now your starting to get into a microprocessor controlled system so then you might as well
go for the complete solution and just do it completely with a micro.

Isn't  there an open source MPPT project out there ??  Something that an do more power and higher
input voltage ?

boB



vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #19 on: May 10, 2012, 08:47:14 PM »
There are situations where this module may be truly useful.  E.g., I've had an electric fence controller running off of a 12V lead-acid battery and a PV panel of about 4 watts with a simple anti-overcharge controller (VW dealer type).  It barely kept up - sometimes after cloudy weather I had to bring the battery elsewhere to charge it.  Switching to this module might have reduced the number of times I'd had to do that, if the resulting PV charge rate was noticably higher.

Going back to the question of can the _input_ voltage be regulated via the feedback pin, without losing the output regulation, perhaps yes after all.  That is because, when charging a battery, we don't really want to "regulate" the output voltage, we just want to limit it, so the battery will not be overcharged.

Think of the following circuit: same as OperaHouse's above, but connect the collector of the NPN transistor not to pin 5, but to pin 4 via a resistor.  Arrange the resistors all around so that when that transistor is conducting, Vout will be equal to the highest voltage we want to allow the battery.  If the PV voltage is high, and the battery is full, the transistor will be "on" and the LM2596 will see its target voltage and not output any current.  If the battery voltage is less than full, the LM2596 will increase the output current in an attempt to make the output voltage higher.  That will eventually lower the PV voltage and turn the transistor off.  When off, the feedback pin sees a higher voltage, and the LM2596 will stop the output current, causing the PV voltage to increase again, etc.  Ideally the transistor will settle somewhere between all-on and all-off.  That can probably be arranged by adding a small resistor between the emitter and ground, along with a resistor between the base and ground.  Add a filter capacitor or two and no more "wham bam".  But will that be more _efficient_ than using pin 5?
« Last Edit: May 10, 2012, 09:36:39 PM by vtpeaknik »

vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #20 on: May 10, 2012, 08:58:06 PM »
Attached is a conceptual (untested) schematic to go with my idea mentioned above for a simple current limit mechanism for the LM2596.  Note that the components not related to the current limit are just representative (actually stolen from the existing schematic I linked to above).

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #21 on: May 11, 2012, 09:55:45 AM »
I demonstrated that method a while ago with a 34063 regulator.  Your imaginations have gone a little wild.  I defy you to see any difference between using the shutdown or feedback pins in operation.  Your idea of the voltages bouncing around just does not happen.

When I first received the three pot regulator I thought it had way too few parts.  Only four resistors and two of those are used by the LED indicators.   I worked for a small manufacturer of custom controls that had a really high insertion cost for each component.   I would agonize for days over adding  one or two 1/3 cent resistors.  That would often decide whether a project was built or not.  I do have a fondness for these minimal circuits where every component does double duty but there are real tradeoffs in performance.   I wouldn't recommend anyone copy that circuit exactly.

My ebay three pot circuit uses a Norton amp and also uses the 5V reference.  Most op-amps compare voltage and are high impedance.  A LM359 is a Norton  current amplifier that has current passing through the inputs.  This is a much better choice than the LM358 where the signal will be buried in noise and drift.  I tested the current limiting function at about 180 ma.  Performance was sufficient for most applications. The circuit trace used to sense current was calculated to be about .018 ohms and only created a voltage burden of about 4 mV at that current.

Efficiency  at 4-6V drops to the low 70's because of the diode loss.  Adding a transistor for current sense would likely drop efficiency another 15%.  And be subject to drift.  There really isn't a good reason to not to use an op-amp instead of a transistor.

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #22 on: May 11, 2012, 02:41:52 PM »
I'dlikerosee someone parallel five of these together.  At around $2 each this isa cheap solution for a regulator.  Don't forget how useful opto isolators can be in these circuits where a transistor is used.  Tracking is far over rated. 

vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #23 on: May 11, 2012, 05:04:19 PM »
Thanks for the info on the 3-pot model.  Probably well worth the $5 if you're looking to set up a custom small battery charger.

But as far as an MPPT-subsitute for anything but the smallest PV systems, these little modules are not easy to keep cool at the higher end of their power range.  That's because they are built so SMALL!  Compare with the suggest PCB layout in the chip's datasheet.  Perhaps somebody out there is selling a less-cramped version?

Alternatively, presumably somewhere out there there is a different cheap-enough switching regulator that can really work at a higher power level, say up to 20 amps output at 14V.  That would be a stronger candidate as a substitute for an expensive MPPT charge controller.  But how to find one?  Perhaps a PC power supply can be modified to do this, changing its input from 120/240VAC to DC in the range of 16-65V, and the 12V output to 14V?  Or perhaps those are too hard to convert?

Here is one person's impressive DIY effort:
http://electronics-diy.com/electronic_schematic.php?id=1023
- this thing is NOT an MPPT controller, just a large and efficient DC-DC converter, and
- it looks like too much work relative to buying an off-the-shelf MPPT controller for about $200.
« Last Edit: May 11, 2012, 06:08:48 PM by vtpeaknik »

vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #24 on: May 12, 2012, 10:24:23 AM »
Found a larger DC-DC module with heat sinks offered for about $15 - search on fleabay for "150W DC-DC 10-32V to 12-35V Converter Boost Charger".  This is supposed to be a step-up converter - any reason it couldn't also run as step-down?

boB

  • Sr. Member
  • ****
  • Posts: 389
  • Country: us
    • boB
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #25 on: May 12, 2012, 09:44:14 PM »

Going back to the question of can the _input_ voltage be regulated via the feedback pin, without losing the output regulation, perhaps yes after all.  That is because, when charging a battery, we don't really want to "regulate" the output voltage, we just want to limit it, so the battery will not be overcharged.


I think that regulating the input voltage is only going to ensure that the regulator gets as hot as it can.

To do any good here, I have a feeling that connecting the PV directly across the battery and using a shunt regulator to limit the battery charging voltage is the way to go.

But yes, you could connect all sorts of circuitry to this and regulate the input voltage and output voltage limit as well.

How about doing this same thing except using a TI (National Semi) Simple Switcher IC would would give you some real benefit, maybe.

boB

vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #26 on: May 12, 2012, 11:26:02 PM »
"I think that regulating the input voltage is only going to ensure that the regulator gets as hot as it can.  To do any good here, I have a feeling that connecting the PV directly across the battery and using a shunt regulator to limit the battery charging voltage is the way to go."

- I disagree.  Regulating the input voltage here means drawing enough current to bring the PV voltage down to the max-power point of the panel - but not lower.  That does not heat up anything, on the contrary.

As any exposition on the merits of an MPPT charge controller will explain, if you connect the panel to the battery directly you are "burning" a lot of the power heating up the panel.  That is because the panel can output almost as much current at a higher voltage than the battery voltage, that's more watts.  A circuit that can convert the higher voltage to a lower one efficiently enough will yield a higher charging current at the battery side.

I see it every day in my PV system, which uses higher-voltage panels (max power at 25V rather than the more typical 17V) - and I have two of those panels in series.  In full sun, the 6-7A from the panels translates to 20-25A into the battery thanks to the MPPT controller.

Even the simple and cheap circuit discussed here does it, see the numbers reported above by Operahouse - although the gain is minor in the worst-case conditions used for that test (low panel voltage, high battery voltage).

boB

  • Sr. Member
  • ****
  • Posts: 389
  • Country: us
    • boB
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #27 on: May 13, 2012, 02:34:35 AM »
"I think that regulating the input voltage is only going to ensure that the regulator gets as hot as it can.  To do any good here, I have a feeling that connecting the PV directly across the battery and using a shunt regulator to limit the battery charging voltage is the way to go."

- I disagree.  Regulating the input voltage here means drawing enough current to bring the PV voltage down to the max-power point of the panel - but not lower.  That does not heat up anything, on the contrary.

As any exposition on the merits of an MPPT charge controller will explain, if you connect the panel to the battery directly you are "burning" a lot of the power heating up the panel.  That is because the panel can output almost as much current at a higher voltage than the battery voltage, that's more watts.  A circuit that can convert the higher voltage to a lower one efficiently enough will yield a higher charging current at the battery side.

I see it every day in my PV system, which uses higher-voltage panels (max power at 25V rather than the more typical 17V) - and I have two of those panels in series.  In full sun, the 6-7A from the panels translates to 20-25A into the battery thanks to the MPPT controller.

Even the simple and cheap circuit discussed here does it, see the numbers reported above by Operahouse - although the gain is minor in the worst-case conditions used for that test (low panel voltage, high battery voltage).


Oooops !!   SORRY !!  I thought this was a linear regulator.

The original schematic did not have the inductor.

I was trying to say go to a simple switcher but this ~IS~ a simple switcher !!

sorry !  My bad for not noticing the second schematic !   Nevermind.

 Completely valid point.

boB

« Last Edit: May 13, 2012, 02:40:44 AM by boB »

independent

  • Full Member
  • ***
  • Posts: 132
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #28 on: June 04, 2012, 12:44:08 AM »
Hello everyone. I just built and tested a version of this circuit. I got it from Chris Graecen's article in Homepower issue 38. In his article he calls it a LCB or a Linear Current Booster and is somethiing I've been interested in for some time. In the article he has 100uf on the inputs and 1000uf on the output caps. Otherwise much the same configuration using npn transistor switching pin 5 to maintain maximum power point. For the actual switching he (and I) used an LM723 voltage regulator in DIP 14 format hooked up with a 10k pot, a few resistors, a small cap and  a number of jumpers.

The scenario:
It was getting near evening and I used a AA battery I knew was or was nearly exhausted. Short circuit current on the 15w panel was 73mA. Once the circuit was installed I was instantly able to see 350mA on the ammeter between the battery and the circuit. The light level didn't change much to my eyes but it ended up maxing out with around 450mA (nearly half an ampere woo-hoo!). Perhaps a little more light and tuning of the maximum power point did it. At no time was there direct sun on the solar panel. For the LM2596 all I had available was the 3 pot version. Mentioned above the single pot version of this circuit would perhaps be more efficient. I have immediate use for this as an AA charger. 6x AA or even a single cell Li-ion cell, the panels either 5w or 15w.  As you can see there are no shadows at all. The circuit will behave differently in sunlight. This 3 pot LM2596 has current limiting but I won't be charging a single AA cell off it with the 15w panel.

The Homepower magazine circuit is from 1993/4 and uses a LM2576. I just subbed in the LM2596 with the 3 pot version I got from somewhere in China/HK. It uses an LM723 in almost exactly the same way as OperaHouse has done above although with few more components.

I can see so many uses for this circuit from running 12v DC adapters like cellphone chargers. I can't confirm it but apparently the LM723 will shut down at 8v which might stop brownout problems with electronics which can be a problem when running them directly off solar without a battery. The maximum power point will no doubt also be helpful here. Other uses are running motors directly and battery charging as mentioned above
« Last Edit: June 04, 2012, 04:15:07 AM by independent »

OperaHouse

  • Hero Member
  • *****
  • Posts: 1308
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #29 on: June 04, 2012, 10:46:32 AM »
I was a little confused at first.  A linear LM723 or a 7805 can be used as a switching regulator.  So you use the 723 as a reference and controller for the 2596.  I see the extra heat sink.  The 2596 has thermal shutdown and is capable of higher currents.  The limit of these China modules is the switching diode heating up and the inductor changing inductance dramatically as they warm up.  If pushing these modules remember to also apply a hearsink to the entire bottom of the cirduit board.

independent

  • Full Member
  • ***
  • Posts: 132
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #30 on: June 05, 2012, 01:55:24 AM »
Yes, LM723 does the job of regulating the panel voltage via a 2n2222 to pin 5 on the LM2596 much in the same way OperaHouse's circuit does from what I can tell.

I also ran this through 6x AA cells connected in series today, regulator was set to 8.4v or thereabouts. Panel was a 15w crystalline. There wasn't any sun and short-circuit current on the panel was 138mA and into the cells was 150mA through the regulator. 8.7% gain in measured current in marginal conditions is promising, waiting for the sun and an opportunity to try this on my 5w panel.

It seems to me there is a real limit with these boards due to the size, no real mounting holes and difficulty of attaching a proper heatsink directly to the LM2596 chip
« Last Edit: June 05, 2012, 03:23:39 AM by independent »

vtpeaknik

  • Full Member
  • ***
  • Posts: 136
  • Country: us
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #31 on: June 05, 2012, 12:12:04 PM »
Instead of the LM723 could use a MAX8212 chip.  Uses less power, and wouldn't even need that 2N2222, as its output is a "sink".  That chip is also useful on its own as a LVD (alternative to ghurd's circuit).  I've gotten a 5-pack of them off of fleabay so I'll be experimenting.

independent

  • Full Member
  • ***
  • Posts: 132
Re: LOW POWER/COST LM2596 POWER POINT REGULATOR
« Reply #32 on: June 05, 2012, 04:17:13 PM »
If I'm reading this right the MAX8212 can only source 18v max. In the datasheet they mention the ICL8212 which goes to 30v. LM723 goes to 37v