Solar Panel Voltage Regulator

On Sunday, August 7, 2016 at 10:46:44 AM UTC-4, snipped-for-privacy@gmail.com wr ote:

e:

wn or reduced output with passing clouds so the switcher has to intelligent ly recover from undervoltage/power shutdown. MPPT is meaningless for a char ger because its output voltage and thus power is constant for the most part , and that power draw has to be less than the MPP if the charger is to avoi d backing off its charging rate. For any given power draw there are two pos sible operating points for the panel: higher current at lower voltage, or l ower current at higher voltage, the two converging to a single solution at the MPP, as a function of insolation of course. This is the perfect setup f or what used to be called "hunting."

so

harger

be a

nel insolation no longer supports full power output, at which point it star ts tracking MPP. The schematic shown is an abomination, the explanation of circuit operation is total horseshit, and it's full of a bunch of trims and offset adjusts that misrepresent it as a precision when in fact it's a tou chy pos good for little more than a lab demo- if that.

It bucks down the input voltage while continually hunting for the duty cycl e that maximizes battery charging current, i.e. extracts maximum power from the solar input.

How is that not MPPT?

Cheers, James Arthur

Actually, the loss in efficiency is only 20% since the panel current increases about 8% as you go under MPP in voltage. Well, it's more like 25% but adding the switcher loss of 5% makes it 20%. And if Harry changes the orientation of the panels twice a day to get a better angle of the sun, he can make up that 20% and maybe more. So, I'm still in the relay club with the addition of a diode for temp compensation. Gotta keep it simple (KISS).

Your relay scheme cutting off the moment you reach Vtopping only uses ~80% of the battery's capacity, which reduces the battery's service life, and may waste up to 20% of potential charging time too.

Second, won't you need considerable hysteresis to avoid oscillating? That loses more charging time, until the battery has discharged to the lower trip-point, and it cycles the battery (discharge, then charge) unnecessarily, reducing life.

You *could* use (or make) a relay with a calibrated trip point to sense when the battery current falls, like the old, old days, then disconnect. Steampunk charger!

Cheers, James Arthur

Trackers are great as long as you can keep them working well. They have been kinda prone to breaking in the past.

ote:

down or reduced output with passing clouds so the switcher has to intellige ntly recover from undervoltage/power shutdown. MPPT is meaningless for a ch arger because its output voltage and thus power is constant for the most pa rt, and that power draw has to be less than the MPP if the charger is to av oid backing off its charging rate. For any given power draw there are two p ossible operating points for the panel: higher current at lower voltage, or lower current at higher voltage, the two converging to a single solution a t the MPP, as a function of insolation of course. This is the perfect setup for what used to be called "hunting."

" so

charger

d be a

panel insolation no longer supports full power output, at which point it st arts tracking MPP. The schematic shown is an abomination, the explanation o f circuit operation is total horseshit, and it's full of a bunch of trims a nd offset adjusts that misrepresent it as a precision when in fact it's a t ouchy pos good for little more than a lab demo- if that.

cle

Did you miss the part in the horseshit circuit description that says when i t finally achieves constant current operation the "dithering" integrator go es into saturation and stops "dithering." He doesn't even know what "dither ing" is because a deterministic ramp modulation is not "dithering."

ote:

s a watt,

ther panel

4 to 120

re than

see how he

more

ad-wise.

u're in a

reas where

kid, now

h is more

teries at

at 120 or

t. But he

eat. Not

st, so he

regulators

c for a

age 17.

0627-TechManual-Lo.pdf

ight

n Energy Star rating for it either.

wn or reduced output with passing clouds so the switcher has to intelligent ly recover from undervoltage/power shutdown. MPPT is meaningless for a char ger because its output voltage and thus power is constant for the most part , and that power draw has to be less than the MPP if the charger is to avoi d backing off its charging rate. For any given power draw there are two pos sible operating points for the panel: higher current at lower voltage, or l ower current at higher voltage, the two converging to a single solution at the MPP, as a function of insolation of course. This is the perfect setup f or what used to be called "hunting."

so

charger

be a

panel insolation no longer supports full power output, at which point it st arts tracking MPP. The schematic shown is an abomination, the explanation o f circuit operation is total horseshit, and it's full of a bunch of trims a nd offset adjusts that misrepresent it as a precision when in fact it's a t ouchy pos good for little more than a lab demo- if that.

Right- it's not a problem- his little MPPT goes into effect at the tails of the operating regime, most of the time it should be CC.

On Monday, August 8, 2016 at 6:38:01 AM UTC-4, snipped-for-privacy@gmail.com wro te:

:

utdown or reduced output with passing clouds so the switcher has to intelli gently recover from undervoltage/power shutdown. MPPT is meaningless for a charger because its output voltage and thus power is constant for the most part, and that power draw has to be less than the MPP if the charger is to avoid backing off its charging rate. For any given power draw there are two possible operating points for the panel: higher current at lower voltage, or lower current at higher voltage, the two converging to a single solution at the MPP, as a function of insolation of course. This is the perfect set up for what used to be called "hunting."

e," so

ry charger

uld be a

e panel insolation no longer supports full power output, at which point it starts tracking MPP. The schematic shown is an abomination, the explanation of circuit operation is total horseshit, and it's full of a bunch of trims and offset adjusts that misrepresent it as a precision when in fact it's a touchy pos good for little more than a lab demo- if that.

cycle

om

it finally achieves constant current operation the "dithering" integrator goes into saturation and stops "dithering."

Yes, I missed it. This is what I read--

"Either way we get convergence toward VMPPT and maxi- mum charging current for B1. This mode of operation continues as B1 charges and its voltage rises to the ?14.1-V terminal-voltage setpoint de- termined by the R6-R7-R8-RT network. Once reached, A2 saturates with zero output and normal LTC1149 con- stant-voltage regulation takes over. RT provides temperature compensation appropriate for typical lead-acid bat- tery chemistry."

IOW, it hunts for MPPT while charging, then switches to temperature- compensated constant voltage once the target voltage is reached.

That's exactly what it should do. It wouldn't make sense to maximize charging into a battery that's fully charged.

ulation is not "dithering."

I've always called that--superimposing a triangle modulation for hunting or a/d resolution enhancement--"dithering," and the first two references I consulted agree. But Woodward's description shows his modulation waveform anyhow, so it's not like he was unclear, or his description confusing.

Cheers, James Arthur

It's not at all clear what your point is. I think everyone understands how the circuit works. When the load can draw more current than the solar panel can supply, the circuit pulls the maximum power by running at the MPP. When the load can't accept the full power the solar panel can supply, the circuit limits the power drawn and so obviously won't run at the MPP.

Why do you insist it's not a MPPT charger... it is?!

--

Rick C

Well, I don't know if this is right or not, but it looks like the open circuit voltage at 90% capacity is 25.46 which is almost 2 volts from the top of 27.2. So, it probably won't oscillate in a 2 volt range. Another idea is to add a resistor and diode to provide a maintenance charge of .01C as Fred suggested. So, the only loss would be 10% as the battery falls from the maintenance point to 25.46.

Your relay / transistor circuit is a giant improvement over the rat's existing gear. But it would be nice to use that last 20-30% of the battery's capacity(*), improving the battery's service life at the same time.

(*) I'm eyeballing this graph:

(article:

PowerSonic's boosted LM317 circuit isn't bad--pretty simple, treats the battery well, and uses the battery's whole capacity.

The next simplest thing would be a temperature-compensated linear regulator set to the float voltage. Prevent back-flows with a diode. No relays required, and no fiddling with hysteresis thresholds.

Cheers, James Arthur

I suppose he could use a LM317 to maintain the float voltage (27.2). But the

317 requires at least 3 volts more going in than coming out, so at 15 amps, you have at least 45 watts of heat which requires a heat sink..And I think he has 10 panels which would waste 150 watts. The relay solves 2 problems, it doesn't drop any voltage, and it enables the panels to supply maximum current instead of a lower constant current. I can add temp compensation with one diode in series with the 6.2k resistor and than readjust the 6.2K to fix the trigger points. I might get some more info tomorrow about what he actually wants to do. .

"billbowden" wrote in message news:57ae7455$0$43806$c3e8da3$ snipped-for-privacy@news.astraweb.com...

Here's the latest version with a transistor replacing the relay. The pass transistor drops about 1 volt or 5 watts at 5 amps. Hysteresis points are set at 28.4 and 25.2. Two diodes are added for temp compensation. All transistors are rated at 60 volts so no damage to the circuit if the battery is disconnected.

Version 4 SHEET 1 880 680 WIRE 16 -160 -112 -160 WIRE 272 -160 96 -160 WIRE 432 -160 272 -160 WIRE 512 -160 432 -160 WIRE -112 -144 -112 -160 WIRE 432 -128 432 -160 WIRE 352 -80 272 -80 WIRE 368 -80 352 -80 WIRE 512 -80 512 -160 WIRE 448 -32 432 -32 WIRE -16 48 -112 48 WIRE 80 48 -16 48 WIRE 512 48 512 16 WIRE 512 48 80 48 WIRE -16 64 -16 48 WIRE 352 80 352 -80 WIRE 352 80 144 80 WIRE 352 128 352 80 WIRE 512 144 512 48 WIRE 144 176 144 160 WIRE 288 176 144 176 WIRE -112 208 -112 48 WIRE 512 240 512 224 WIRE 512 240 240 240 WIRE 80 288 80 48 WIRE 144 288 144 176 WIRE 256 320 240 320 WIRE 352 320 352 224 WIRE 352 320 336 320 WIRE 512 320 512 240 WIRE -16 336 -16 272 WIRE 16 336 -16 336 WIRE 240 336 240 320 WIRE 240 336 208 336 WIRE 240 352 240 336 WIRE 352 352 352 320 WIRE 112 384 80 384 WIRE 144 384 112 384 WIRE -16 400 -16 336 WIRE 112 400 112 384 WIRE 240 448 240 432 WIRE 352 448 352 432 WIRE -16 496 -16 480 WIRE 112 496 112 480 FLAG 352 448 0 FLAG 240 448 0 FLAG 112 496 0 FLAG -16 496 0 FLAG -112 288 0 FLAG 512 384 0 FLAG -112 -64 0 SYMBOL voltage -112 192 R0 WINDOW 0 -77 59 Left 0 WINDOW 3 -253 144 Left 0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value PULSE(20 30 0 1 1 .01 2.01 3) SYMBOL npn 16 288 R0 WINDOW 0 118 322 Left 0 WINDOW 3 -47 326 Left 0 SYMATTR InstName Q1 SYMATTR Value 2N3019 SYMBOL npn 208 288 M0 WINDOW 0 85 324 Left 0 WINDOW 3 54 47 Left 0 SYMATTR InstName Q2 SYMATTR Value 2N3019 SYMBOL res 224 224 R0 WINDOW 0 -293 -125 Left 0 WINDOW 3 -31 47 Left 0 SYMATTR InstName R1 SYMATTR Value 2k SYMBOL res 128 64 R0 WINDOW 0 -209 355 Left 0 WINDOW 3 39 74 Left 0 SYMATTR InstName R2 SYMATTR Value 2k SYMBOL res 224 336 R0 WINDOW 0 -58 -229 Left 0 WINDOW 3 34 88 Left 0 SYMATTR InstName R3 SYMATTR Value 2k SYMBOL res 96 384 R0 WINDOW 0 42 40 Left 0 WINDOW 3 43 67 Left 0 SYMATTR InstName R4 SYMATTR Value 2k SYMBOL res 496 128 R0 WINDOW 0 -305 118 Left 0 WINDOW 3 41 60 Left 0 SYMATTR InstName R5 SYMATTR Value 2k SYMBOL res 336 336 R0 WINDOW 0 -56 -50 Left 0 WINDOW 3 39 60 Left 0 SYMATTR InstName R6 SYMATTR Value 2k SYMBOL res -32 48 R0 WINDOW 0 293 351 Left 0 WINDOW 3 44 52 Left 0 SYMATTR InstName R7 SYMATTR Value 6k SYMBOL res -32 384 R0 WINDOW 0 410 -14 Left 0 WINDOW 3 -50 62 Left 0 SYMATTR InstName R8 SYMATTR Value 2k SYMBOL res 352 304 R90 WINDOW 0 -137 -200 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R9 SYMATTR Value 35k SYMBOL diode -32 144 R0 WINDOW 3 208 466 Left 0 SYMATTR Value 1N4148 SYMATTR InstName D1 SYMBOL diode -32 208 R0 SYMATTR InstName D2 SYMATTR Value 1N4148 SYMBOL zener 528 384 R180 WINDOW 0 20 70 Left 0 WINDOW 3 -130 70 Left 0 SYMATTR InstName D3 SYMATTR Value BZX84C12L SYMBOL voltage -112 -160 R0 WINDOW 0 40 39 Left 0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 WINDOW 3 42 72 Left 0 SYMATTR InstName V2 SYMATTR Value 44 SYMBOL pnp 288 224 M180 WINDOW 0 53 31 Left 0 WINDOW 3 49 63 Left 0 SYMATTR InstName Q5 SYMATTR Value 2N2905A SYMBOL res 112 -176 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R11 SYMATTR Value 4 SYMBOL npn 448 -80 R0 WINDOW 0 -141 692 Left 0 WINDOW 3 36 49 Left 0 SYMATTR InstName Q3 SYMATTR Value 2N3055 SYMBOL pnp 368 -32 M180 WINDOW 0 -98 -644 Left 0 WINDOW 3 25 52 Left 0 SYMATTR InstName Q4 SYMATTR Value 2N2905A SYMBOL res 256 -176 R0 SYMATTR InstName R10 SYMATTR Value 2k TEXT -264 432 Left 0 !.tran 4 TEXT -304 208 Left 0 ;24 Volt Battery TEXT -400 -152 Left 0 ;5 Amp Solar Panel TEXT -400 -120 Left 0 ;4 Ohm Inpedance TEXT -416 -88 Left 0 ;44 Volts Open Circuit TEXT -368 -56 Left 0 ;36 Volt MPP

.

I can't say I understand. This circuit drops ridiculous amounts of voltage across the 4 ohm resistor wasting up to 121 Watts. Is there a reason for this?

I think maybe this is not an accurate model for the solar cell. I don't think you can achieve a decent model with just voltage source and a resistor. So in reality, there would be a lot less wasted power in the solar cell.

I don't see a float charge with this. Running the simulation I see the current drop as the voltage rises until the voltage reaches 28.4 volts. Then the current drops to nothing, small negative in fact. Shouldn't there be a constant voltage charge to top off the battery? Then a lower voltage to float charge it?

--

Rick C

The electrical power dissipated in the panel is negligible compared to the unused irradiance energy that goes up as heat, so this whole thing about MPP is a bunch of phooey for battery chargers.

How is the heat from unusable radiation at all relevant? The issue is getting as much power from the panel when needed allows the panels to be most cost effective.

--

Rick C

The panel is 44 volts open circuit. MPP is 36 volts at 5 amps, so the impedance is 8/5 =1.6 ohms. The current increases about 8% when the panel is loaded down to 24 volts. So, the new impedance is (44-24)/5.4 = 3.7 ohms. I was using 5 amps and not adding the extra 8% so I got 4 ohms.

The idea is to display the circuit voltages and currents when the charge current is about 5 amps and also when the current is zero and the panel voltage is 44 volts. The numbers are not exact, but close enough.

The batteries have more capacity than needed, so the float voltage isn't important

.

Using a higher input voltage (PV) into the charger is also good for reducing I^2XR loss in long cables and reducing the cost of copper for long runs. That cost saving alone can pay for the extra cost of an MPPT charger controller.

Another reason to place more PV modueles in series even without long cable losses is that if there is a partially shaded module in the string from poles or trees at certain times of day, the MPP voltage drops down and as long as the MPP V is higher than the battery voltage somewhat, the system will still charge, but lose that one PV module in the series. That is, as long as you have at least one extra module of nominal battery voltage in series.

Of course the best thing is to try not to place the PV array in any shade but sometimes that is not possible. Also, some off grid cabins will need to place the array FAR away from that cabin, on the other side of trees, etc, and then the wire losses can be mitigated with thie higher input voltage.

Then again, depending on your battery voltage, for instance 12V batteries, it is much easier to find PV that is higher than your nominal 12V (~21 Voc) module these days. Harder to find a 12V module because of all the grid tie only systems in place today which is most of the market. So matching is much easier with MPPT...

And, due to fact you MUST use the same ~nominal~ voltage PV array as the batteries for PWM chargers, meaning, the MPP V is just a few volts above the battery voltage, when it is HOT out, the PV heats up and the MPP V drops down to and often BELOW the battery voltage and then you aren't getting optimum battery charging. Higher MPP V (~17V for 12V battery) and low battery S.O.C. is exactly the time you NEED that extra charging ability.... Of course, these days, if you can find cheap nominal battery voltage PV panels, then you can just add more of them ~IF~ you have the space which some people don't... That's another reason for MPPT is less roof space required.

So, although it was kind of questionable say, 20 years ago to go MPPT charging, it certainly isn't so phooey any more.

This also works for small wind and micro-hydro but different tracking algorithms are necessary for those sources.

boB K7IQ

wrote in message news: snipped-for-privacy@googlegroups.com...

Here's a low dropout 13.6 volt regulator I cooked up using 5 transistors. Dropout is 1.3 volts at 5 amps. Sort of a waste of time since you can buy one at Jameco or Digikey for $2.50. (LM1084). This thing could be modified for 27.2 volt operation. It's temperature compensated with diode D2. The reference is a 3 volt white LED (D1).. Current is somewhat balanced in the differential pair Q1,Q2. V1 is a 19 volt open-circuit solar panel with a

0.82 ohm impedance. V2 is a 12 volt battery sitting at 10 volts, and R11 represents the battery impedance at 5 amps.

Also, Doc asked about your brother and how he is doing?

Version 4 SHEET 1 960 680 WIRE 240 -128 176 -128 WIRE 352 -128 320 -128 WIRE 528 -128 352 -128 WIRE 624 -128 528 -128 WIRE 352 -80 352 -128 WIRE 448 -80 432 -80 WIRE 464 -80 448 -80 WIRE 624 -80 624 -128 WIRE 560 -32 528 -32 WIRE -384 32 -464 32 WIRE -224 32 -304 32 WIRE 128 32 -224 32 WIRE 256 32 128 32 WIRE 624 32 624 16 WIRE 624 32 256 32 WIRE -464 48 -464 32 WIRE -224 64 -224 32 WIRE 128 64 128 32 WIRE 256 144 256 32 WIRE 448 160 448 -80 WIRE 128 176 128 144 WIRE 128 176 80 176 WIRE 160 176 128 176 WIRE 240 192 240 176 WIRE 384 208 352 208 WIRE -224 224 -224 144 WIRE 16 224 -224 224 WIRE 256 224 256 208 WIRE 256 224 224 224 WIRE -224 240 -224 224 WIRE 160 320 160 272 WIRE 352 320 352 208 WIRE 352 320 160 320 WIRE -224 352 -224 304 WIRE 160 352 160 320 WIRE 256 352 256 224 WIRE 448 352 448 256 WIRE 80 432 80 272 FLAG 176 -48 0 FLAG -224 432 0 FLAG 448 432 0 FLAG 256 432 0 FLAG 80 432 0 FLAG 160 432 0 FLAG -464 128 0 SYMBOL voltage 176 -144 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 WINDOW 0 -75 54 Left 0 WINDOW 3 -77 84 Left 0 SYMATTR InstName V1 SYMATTR Value 19 SYMBOL npn 560 -80 R0 SYMATTR InstName Q4 SYMATTR Value 2N3055 SYMBOL res 448 -96 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R7 SYMATTR Value 1k SYMBOL res -240 48 R0 WINDOW 3 41 59 Left 0 WINDOW 0 -36 60 Left 0 SYMATTR Value 660 SYMATTR InstName R8 SYMBOL res -240 336 R0 WINDOW 0 -35 65 Left 0 WINDOW 3 41 65 Left 0 SYMATTR InstName R9 SYMATTR Value 2k SYMBOL npn 384 160 R0 SYMATTR InstName Q6 SYMATTR Value 2N3904 SYMBOL res 432 336 R0 SYMATTR InstName R1 SYMATTR Value 330 SYMBOL res 336 -144 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R12 SYMATTR Value .82 SYMBOL LED 240 144 R0 WINDOW 3 39 -38 Left 0 SYMATTR Value NSCW100 SYMATTR InstName D1 SYMBOL pnp 224 272 R180 WINDOW 0 39 44 Left 0 WINDOW 3 274 -290 Left 0 SYMATTR InstName Q1 SYMATTR Value 2N3906 SYMBOL pnp 16 272 M180 WINDOW 0 39 44 Left 0 WINDOW 3 69 21 Left 0 SYMATTR InstName Q2 SYMATTR Value 2N3906 SYMBOL res 112 48 R0 WINDOW 0 -36 53 Left 0 WINDOW 3 42 53 Left 0 SYMATTR InstName R2 SYMATTR Value 330 SYMBOL res 240 336 R0 SYMATTR InstName R3 SYMATTR Value 2k SYMBOL res 144 336 R0 SYMATTR InstName R6 SYMATTR Value 560 SYMBOL voltage -464 32 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V2 SYMATTR Value 10 SYMBOL res -288 16 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 41 61 VTop 0 SYMATTR InstName R11 SYMATTR Value .72 SYMBOL diode -240 240 R0 WINDOW 0 -48 32 Left 0 WINDOW 3 42 34 Left 0 SYMATTR InstName D2 SYMATTR Value 1N4148 SYMBOL pnp 464 -32 M180 WINDOW 0 34 46 Left 0 WINDOW 3 1 127 Left 0 SYMATTR InstName Q3 SYMATTR Value 2N2905A TEXT -480 552 Left 0 !.tran 5 TEXT -384 -96 Left 0 ;13.6 Volt, 5 Amp, Low Dropout Regulator

A situation like that requires some messing around. 3 pp3 batteries =27 v into the gate of a powermos as a source follower. And get the powermos from an old TV. Dumpster diving is an art.