Sizing Victron MPPT Controller

[noelex 77]

Originally Posted by GTom

Provided you have at least 18V Voltage to work with

It depends which voltage you are referring too. Voc from the specifications definitely needs to be over 18v. Vmp from the specifications can be a little lower than 18v, but not by much. Real life Vmp could be much lower than 18v.

[john61ct]

Originally Posted by noelex 77

A panel with a specified Vmp of 16v on its own, or in parallel, is too low to use with a 12v battery bank no matter what type of controller is used PWM or MPPT.

I don't think that's true for most PWM, many accept a panel output pretty close to battery charge voltage. In fact the lower voltage may makes more amps output available since PWM just discards power from the higher voltage .

[noelex 77]

Originally Posted by john61ct

I don't think that's true for most PWM, many accept a panel output pretty close to battery charge voltage. In fact the lower voltage may makes more amps output available since PWM just discards power from the higher voltage .

This is correct, but the specified Vmp is for very bright conditions and cool cell temperatures. The real life Vmp will be 1.5 to 2v less than the Vmp measured under STC (Sandard Test Conditions) unless you are in a cold climate.

With a panel that has a Vmp of 16v under STC conditions, this puts the real life Vmp below the bulk charging battery voltage, which is not good for either PWM or MPPT regulators. Fortunately nearly all “12v” panels are produced with a much higher Vmp for this reason.

[conachair]

Originally Posted by john61ct

I don't think that's true for most PWM, many accept a panel output pretty close to battery charge voltage. In fact the lower voltage may makes more amps output available since PWM just discards power from the higher voltage .

Don't understand that - a panel direct to a battery will be at the battery voltage, with a PWM in bulk with the battery voltage below the absorption voltage set point the panel voltage will still be at battery voltage - how can the PWM regulator dump power from a voltage that isn't there?

[CatNewBee]

Victrom MPPT spread sheet to calculate the controller size

Regarding PWM:

PWM switches only the panel trough to the battery or off, and this quickly.

If switched trough, the voltage drops to battery voltage because the battery has a much lower impedance / resistance than the solar cell modules. At this voltage the solar cells provide charging current, whatever they can produce and the battery absorb.

Eventually the battery voltage rises above the point the PWM needs to take action and stop charging to not overcharge the battery - this is the usually absorption voltage set point. The energy transferred / stored is equal Current * Time (Amp hours)

With switching quickly on and off the charging happens no longer continuously so the energy transferred is then limited to the current flowing during the time when the PWM switches on only. Effectively the charging is reduced, when the voltage tries to rise above the set point. When there is no charge, the battery voltage drops and allows the PWM to turn on again...

During the OFF phase, the voltage on the battery drops slowly, the voltage on the solar panels jumps to Voc (open circuit) and the current of the solar panels equals zero, because the circuit is open..

That is how it works. After a while (timer setting) the controller assumes, Absorption phase finished, and set a new lower voltage threshold for switching - called Float. The switching starts then, when the battery voltage drops below that set point and keeps the battery at this voltage as long as power comes in.

In reality PWM controller (pulse width modulation) controls the energy flow by the ratio between the length of "ON" (the width of the pulse) in contrast to "OFF" by pulsing at a pre-set operating frequency ( fixed length of ON+OFF time). This is done to prevent interferences with other devices.

[conachair]

And some real world data -

I plugged the data coming out from a 75/10 into a database overnight. Well, just and no more, the computer was only getting updates every few minutes for some reason.
2 x 100w solar panels hanging vertical off the guard rails, sunrise round about the port quarter in a boatyard with lots shadow at that angle.

Sunrise 06:44 (due east it says! That must mean something about today... )

Mode 3 is bulk so regulator turned on at 06:47, or earlier. Batteries were on float on mains charger so victron figures of 5v above battery voltage to turn on seems to be spot on.

Temperature coefficient is -0.35 %/°C so say 15deg this morning a highest you'd likely get even in a tropical anchorage 35degC?
So 20 x 0.35 = 7% which would take 18.5v down to 17.2v. Close to 5v over a typical (for me) morning voltage of 12.6v. So in this case if the maths is right there's not much worry about the victon turning on late, you'll get next to nothing from solar in that first hour anyway.



If I can get the data srteaming problem fixed I'll do some more tests switching parallel / serial and graph the results to see what goes on there.

Ain't data just lovely!

Some from earlier>

[CatNewBee]

Originally Posted by conachair

Don't understand that - a panel direct to a battery will be at the battery voltage, with a PWM in bulk with the battery voltage below the absorption voltage set point the panel voltage will still be at battery voltage - how can the PWM regulator dump power from a voltage that isn't there?

Regarding of the "energy loss".

Solar panel are non-linear devices. while the voltage increases very little on increasing radiation up to Voc, the current increases almost linearly until saturation.

The inner impedance / resistance is non-linear too.

The power is Voltage * Current so you need to multiply this curves at any set point. Because it is non linear there are local maximum of the product of Voltage * Current and one maximum called mp (maximum power point). unfortunately it is somehow moving with radiation and partial shading, also if more than one panel is connected there are the maximums of the other panels too. If you connect a battery directly to a panel or set of pannels the voltage is defined by the battery and the current is then defined by the point of the non-linear impedance resulting from the voltage + the radiation (Voltage / Current).

This is usually near the Ump, but slightly off. so the panels can provide more power at another voltage at this time, but because the battery either drags the voltage down (or up) off the maximum power point the panel cannot deliver its maximum potential of energy producible with the given radiation.

A MPPT controller runs along the curve to find a local / global maximum from time to time and sets the load accordingly to the maximum power point and then transforms the maximal available energy to the according current / voltage for the battery (usually a step-down converter).

The charge is then regulated very often by PWM stage after the step-down stage, because this method does not burn the excess energy and heats up the coontroller too much. Some controller (with very large heat dissipators) regulate the charge current lineary by large transistors or FET. This controllers get hot, but this is an old and expensive design. Other controller just move the input voltage regulator towards Voc to limit the incomming current, so the panels work in a mode that limits the power or switches it off before it rungs through the step down logic.

Many MPPT controller just turn to PWM only controller if the panels connected do not provide enough excess voltage for the step down conversion., so this design is very common in the industry.

[john61ct]

Originally Posted by conachair

Don't understand that - a panel direct to a battery will be at the battery voltage, with a PWM in bulk with the battery voltage below the absorption voltage set point the panel voltage will still be at battery voltage - how can the PWM regulator dump power from a voltage that isn't there?

Grossly simplified, say two PWM SCs, both panels are rated at 100W but A's panel is at 16.7V (6A) and panel B is at 50V (2A).

Now, say actual output from both panels is 90W, and currently the battery charging voltage is at 13.8V.

A will put out a 82.8W charge to the bank (A*V)

B will put out 27.6W

The higher the panel voltage above the minimum needed by the PWM controller to charge, the more power's wasted by the fact they have no DCDC buck conversion going on to capture amps from the higher voltage.

This is separate from the MPP advantage.

[conachair]

Originally Posted by john61ct

Grossly simplified, say two PWM SCs, both panels are rated at 100W but A's panel is at 16.7V (6A) and panel B is at 50V (2A).

Now, say actual output from both panels is 90W, and currently the battery charging voltage is at 13.8V.

A will put out a 82.8W charge to the bank (A*V)

B will put out 27.6W

The higher the panel voltage above the minimum needed by the PWM controller to charge, the more power's wasted by the fact they have no DCDC buck conversion going on to capture amps from the higher voltage.

This is separate from the MPP advantage.

OK, ta that makes more sense. Not really anything to the PWM regulator is actually doing, more what it's not doing - same would apply for panels connect direct in bulk?

[CatNewBee]

Yes. exactly.

[john61ct]

Similar in the abstract, but IRL the PWM is preventing a continuous too-high voltage panel from frying your bank.

[noelex 77]

Originally Posted by conachair

And some real world data -

Thanks for the data.

If the controller turns on too early the drain from the tracking circuitry will result in a negative output (net draw) until the sun climbs higher and the solar panel is generating more than the self consumption. If it turns on too late the output will be immediately significantly positive indicating the controller could have turned on earlier and harvested a little more power.

If the controllers start up parameters were perfectly set the controller would wake up and start tracking exactly at the stage when the solar panel output matched the self consumption of the circuitry. So at start up you should ideally see the panel voltage drop (because it no longer open circuit) but the current should be zero.

Turning on with a 0.09A positive current is pretty good and will not practically have any impact, but theoretically, it just a fraction late. If you have a controller with adjustable start up parameters you can use the above logic to set the controller to wake up at the perfect time, but you need to look at few cycles as the results will not always be the same.

Even when the controller has not woken up and started tracking there will be a very small negative current (draw). On the smaller Victron units this is very tiny about -0.02A but I am surprised it did not show up in the results. I suspect the circuitry was not sensitive enough to record these minor currents, but it may also be the zero point of the shunt was just a fraction out.

[conachair]

Originally Posted by noelex 77

Turning on with a 0.09A positive current is pretty good and will not practically have any impact, but theoretically, it just a fraction late. .

Bear in mind the mains charger was on so it doesn't tell much other than when the reg woke up.

[noelex 77]

Yes, the voltages suggest the regulator would have woken up about 10 mins earlier if the charger was not on.

[Pete7]

Playing around this afternoon with a pair of Victon MPPTs, a 75/15 and another 75/10 for separate panels I noticed that when connected to the lappy with a the USB widget it updated the software in the MPPTs. Originally supplied with v1.16 they updated to v1.30. So if you have the smaller Victon MPPTs it might be worth updating them if you have the opportunity.

Quite nice interface to see what is going on with them: