MPPT buck converter that isn't optimized for battery recharging?

**sensij** · 05-10-2016, 08:29 AM

The power put out by the panels varies throughout the day. The power consumed by the computer varies depending on the task. The two numbers are almost never going to be the same, you need a battery to buffer the difference. Solar panels are not a voltage source that supply only the current that is needed, they will move up and down the IV curve but are usually better understood as a current source.

**SunEagle** · 05-10-2016, 09:57 AM

Unless that PC has a UPS to help "ride through" variations in the power source it could cause it to improperly shut down. Using solar panels without a battery and not connected to the grid to power something (other than a small DC motor) can cause other issues.

You have become aware of the real "costs' to use a solar / battery system. None of any quality are low in cost. Anyone that skimps on what they spend usually get a system that fails to meet their needs or worse just fails.

I am not sure why you want to power your PC using solar but if you feel it will save you money then the only system that meets that criteria is grid tie system.

**mberg2007** · 05-11-2016, 04:12 AM

Originally posted by SunEagle

Unless that PC has a UPS to help "ride through" variations in the power source it could cause it to improperly shut down. Using solar panels without a battery and not connected to the grid to power something (other than a small DC motor) can cause other issues.

Sure, I am aware of that. I'm using a decent buck converter to provide me with a constant voltage of 5V and with a total of 40 watts in panels and a pc that requires 10 watts, there is power to ride through minor overcast conditions. This is a raspberry pi that runs a program that never updates the disk, so pulling the plug to shut it down is not a problem in this case.

Building this system is a hobby project of mine. Just evaluating different configurations in terms of cost and performance. Right now the battery and charge controller account for 80% of the cost (if not more) and I don't think I'm getting very much in return for that.

-Michael

**mberg2007** · 05-11-2016, 04:24 AM

Originally posted by sensij

The power put out by the panels varies throughout the day. The power consumed by the computer varies depending on the task. The two numbers are almost never going to be the same, you need a battery to buffer the difference.

This is true if the computer must keep running throughout the day. But as I stated this is not a priority in my case. If the sun goes away, tough luck. Power off. Not a problem.

I am looking for the equivalent of a buck converter with built in MPPT tracking to extract the maximum power from the panels at various conditions. Right now I have "dumb" a buck converter that just absorbs the power straight from the panels and converts (up to) 40 watts @ 20V to 5V and (up to) 8A. But this conversion doesn't take into account the way solar panels vary in efficiency depending on the load impedance. An MPPT charge controller does this continuously, and uses the power to charge a battery. I'd like the same MPPT tracking from a "plain" buck converter, just not for battery charging but for immediate consumption.

To clarify: Charge controllers have built in "battery chemistry profiles" that tells the controller how to properly charge a lead acid battery for example. There's the bulk phase, the absorption phase and the float phase, any of which requires the controller to vary the voltage and current to reduce gassing, maximize battery lifetime - etc. I don't wnat any of that logic. I just want a module that does proper MPPT tracking (continuously) to maximize power from the panels, and provide that to me at some adjustable output voltage.

Here's an example of a module that kind-of does what I need: http://www.ebay.com/itm/Solar-Panel-...-/141933370253

But this appears to provide only a manual setting for the resistance that is required to maximize the power from the panel at some specific lighting condition. I need the unit to constantly track the power point such as what the Tracer 1210 charge controller (and many others like it) do.

-Michael

**jflorey2** · 05-11-2016, 08:31 AM

Originally posted by mberg2007

I am looking for the equivalent of a buck converter with built in MPPT tracking to extract the maximum power from the panels at various conditions.

Won't work. An MPPT controller can harvest the maximum energy from a panel only when connected to a battery. In a way it is an impedance matcher; it matches the (very low) impedance from the battery to the higher impedance of the panel, and it changes that match so as to extract the maximum possible power from the solar panel. It does this by sending more or less current to the battery.

With your computer that will not work. The computer cannot draw as much power as the panel can generate. Thus you will not be able to draw maximum power at any time unless the load happens to match the max power point. The rest will be wasted.

Right now I have "dumb" a buck converter that just absorbs the power straight from the panels and converts (up to) 40 watts @ 20V to 5V and (up to) 8A. But this conversion doesn't take into account the way solar panels vary in efficiency depending on the load impedance.

Again you don't need that since you cannot match the two impedances.

All you really need for this is an impedance limiter. If your buck converter has a separate sense input, then all you need is a circuit that drives that voltage upwards when the voltage to the input drops by about 10%. (An op amp and a reference should do it.) That will prevent your buck converter from going to too low an impedance, which is the one problem with using a buck with a solar panel.

Another option is to use something like a Midnite Kid and just set the absorb and float voltages to the same value. Not sure why you would do that though since you are not using most of the functionality of the CC.

**sensij** · 05-11-2016, 08:40 AM

You are still not understanding how a pv system works. There is substantially less logic in a battery controller than you think.

Solar panels do not produce "up to" 40 W, for example. They produce whatever they are capable of producing at that moment in time. You cannot simultaneously hold them on their maximum power point and have a load that does not use all of that power, or is trying to use more power. It has to match exactly.

Even setting aside clouds, the maximum power output of the panels varies throughout the day, with the angle of the sun and other factors. Your load needs to continuously adapt.

If the input range of your dumb buck converter covers up to the voc of the panels, it works as long as your load isn't too high. The panels are free to operate at whatever point matches the load. If you try to constrain both the input and output, you will not match and it won't work.

Try this... with your dumb converter, vary the load. Watch what happens to the operating voltage of the panels. There will be a point at which you are at maximum power... any attempt to increase the load from there will sag the pv voltage without proportionally increasing current and actually reduce the power produced.

**mberg2007** · 05-11-2016, 09:10 AM

Originally posted by sensij

YSolar panels do not produce "up to" 40 W, for example. They produce whatever they are capable of producing at that moment in time.

I don't think we disagree here. The panels produce up to 40 watts in full, direct sunlight and substantially less than 40 watts in cloudy conditions. Wasn't that what I said? I did say Watts, not Volts, did I not? Watts = Volts * Amps?

Originally posted by sensij

You cannot simultaneously hold them on their maximum power point and have a load that does not use all of that power, or is trying to use more power. It has to match exactly.

An MPPT charge controller is able to put more power into a battery than a PWM controller, because it optimizes the impedance to match the MPP of the source panel. For any given area of solar panel, the MPPT controller will be able to extract 20-40% more power from that area than a PWM, under the right circumstances. Solar panels produce different power (yes power) for different load impedances, and the controller will try increase or decrease the load impedance to a point on the A-V graph where the total power (V*A) is at its maximum at that given impedance. It actually does that by continuously scanning the panel, varying the impedance and finding the optimum impedance that yields the largest total power output.

Originally posted by sensij

Even setting aside clouds, the maximum power output of the panels varies throughout the day, with the angle of the sun and other factors. Your load needs to continuously adapt.

This is the whole point of an MPPT solar charge controller. I want a buck converter that has the same capability but without the battery charging profile logic.

Yes I realise I won't be able to use all that power all the time, but I will be able to run the computer at "even more cloudy" conditions than otherwise.

Originally posted by sensij

Try this... with your dumb converter, vary the load. Watch what happens to the operating voltage of the panels. There will be a point at which you are at maximum power... any attempt to increase the load from there will sag the pv voltage without proportionally increasing current and actually reduce the power produced.

I understand this. This is what an MPPT charge controller is doing all the time. Varying the load to maximize power under varying lighting conditions.

-Michael

**karrak** · 05-11-2016, 09:58 AM

If we look at this from the point of view of power transfered from the panel to the load.

If the load is drawing zero power the panel voltage will be at Voc. As we increase the load the panel voltage will drop towards Vmp. If the load is greater than the maximum power that the panel will produce at Vmp the voltage going into the buck power supply will collapse causing the output voltage to collapse as well, this will cause the power drawn from the load to drop which will cause the panel voltage to recover. We will either end up with a brownout and a stable voltage or the voltage will oscillate.

As jflorey2 has already said it is the load which will dictate the amount of power you are extracting from the panel and this will be maximised if the load power equals the maximum amount that the solar panel will deliver.

Simon

**sensij** · 05-11-2016, 10:05 AM

Ok, let's try a different approach. In the bulk stage of battery charging, the mppt controller does not put out fixed voltage... it sends as much power into the battery as it can, because the battery impedance is low enough to take it. Once you want to output fixed voltage in the absorb stage, you are no longer assured of operating at the maximum power point, instead, the panel voltage is allowed to drift away from it so that the current produced matches the current the battery can accept at that voltage.

What you want to do basically has the same problem as using a grid tie inverter off grid (assuming you defeat the anti islanding). Let's say you succeed in holding the panels at their maximum power point... now your *output* voltage will be unstable. If you try to draw more power the controller has available, the voltage will sag... I think you understand that. The problem is if you try to draw less, the voltage will spike. It is not possible to hold thhe input at maximum power and the output at an arbitrary load, they have to match, or only one can be constrained.

**Mike90250** · 05-11-2016, 11:48 AM

As a battery fills up and approaches full, a MPPT controller changes modes from MPPT to PWM, because the battery no longer presents enough load for the MPPT circuit to continue to function properly.

This is sort of the problem you have, With low light, you need every bit of power that can be wrung out of the panel. But with full sun, you have way too much power and need a different circuit topology,
The buck boost needs to have MPPT awareness to prevent it from crashing the PV, but with high power, low load it will run away, so you may need a 2nd regulator to be a SHUNT load to keep the MPPT loading within spec of the buck boost. That's something a $30 chi-com solar "mppt" knockoff can't do

You will have to plow through chip mfg data sheets for a compatible circuit. It's likely you will have to use a large capacitor (with proper voltage rating) to emulate a battery to get enough stability to make this work. I have nothing up my sleeve to hand you.

**sensij** · 05-11-2016, 12:01 PM

Yeah, with an appropriate super capacitor or small battery, a regular mppt controller is as good as it gets.

**mberg2007** · 05-13-2016, 08:11 AM

Thank you all for your replies.

I guess this is a mixed bag of results. On one hand the consensus seems to be that it just isn't possible to track the maximum power point unless the output power is going into a battery. On the other hand this sounds like it somehow isn't possible to consume power from a solar panel directly, which I know for a fact isn't true.

The question here is not what to do when the sun is full on the panel. My raspberry will consume 10 watts and the panel will produce 40 watts, which means I will loose 30 watts somewhere - probably dissipated as heat somehow. What to do with too much power is a positive problem that makes MPPT irrelevant.

The question is what to do when the sun is NOT full on the panel. Cloudy conditions, panel that is in shade somehow, sunset, off-axis relative to the sun. In those conditions I want to extract as much power from the panel as possible, which - as I understand it, means presenting a load to the panel which as just the right impedance so the panel produces at Vmp. I imagine that if I were to sit on site turning a small variable resistor up and down, and monitoring the output *power* from the connected buck converter, I would be able to find a spot where the power was higher than anywhere else. Just enough to run my raspberry. But I would have to sit there all the time, turning the resistor up and down as conditions varied during the day.

Is this simply not possible to automate? I always thought this was precisely what the MPPT controller did for me.

If I power my raspberry with a buck converter like this one (see below), is this really the best I can do in terms of keeping the computer running for as long as possible (with no battery)?

http://www.ebay.com/itm/331433779821...%3AMEBIDX%3AIT

-Michael

**mberg2007** · 05-13-2016, 08:50 AM

Originally posted by sensij

It is not possible to hold thhe input at maximum power and the output at an arbitrary load, they have to match, or only one can be constrained.

Could I not constrain the output power by attaching something like a fan (in series with my raspberry) that would spin at whatever power was available? This way I would consume all the power from the panel, whatever it was producing.

-Michael

**jflorey2** · 05-13-2016, 10:08 AM

Originally posted by mberg2007

I guess this is a mixed bag of results. On one hand the consensus seems to be that it just isn't possible to track the maximum power point unless the output power is going into a battery. On the other hand this sounds like it somehow isn't possible to consume power from a solar panel directly, which I know for a fact isn't true.

Of course it's possible. You could just connect a resistor, for example, and dissipate all that power as heat. Or you could connect an LED and convert it back into light. Or connect it directly to a Pi (if the voltage is correct) or a DC/DC converter and then a Pi (provided you guard against voltage collapse.)

I imagine that if I were to sit on site turning a small variable resistor up and down, and monitoring the output *power* from the connected buck converter, I would be able to find a spot where the power was higher than anywhere else. Just enough to run my raspberry. But I would have to sit there all the time, turning the resistor up and down as conditions varied during the day.
I would be able to find a spot where the power was higher than anywhere else. Just enough to run my raspberry. But I would have to sit there all the time, turning the resistor up and down as conditions varied during the day. Is this simply not possible to automate? I always thought this was precisely what the MPPT controller did for me.

Yes. However, your Raspberry/DCDC converter combo is NOT like a resistor. You can turn that knob until your resistor takes 20 watts, but you can never get your Pi to take 20 watts.

With your pi, as long as the voltage is high enough, it will take as much as it needs from the panel. The rest of the power simply won't exist. (It's "wasted" in a sense, but not dissipated, because the panel simply isn't generating it.) As you load down the panel (or decrease the light reaching the panel) the voltage will start to drop. It will continue to drop until the impedance of the load exactly matches the impedance of the panel. At that point you are at the maximum power point.

As you go below that point you see the problem with impedance matching we discussed before. The DC/DC will keep decreasing its impedance to try to pull more power from its source. If the source was a voltage source that would work. But it's not - once you start moving to the right side of the VI plot for a panel it becomes more like a current source. So the power goes DOWN as you decrease the impedance. That makes the voltage drop further until you "crash" - hit the lowest voltage the DC/DC will work at. It will then stay there, and not return until the power from the panel at the new lower voltage exceeds the power the load needs. In practice this means the system would work fine until a cloud passed in front of the sun, at which point you'd see a rapid drop to close to zero power, with no recovery until you were in bright sun.

If I power my raspberry with a buck converter like this one (see below), is this really the best I can do in terms of keeping the computer running for as long as possible (with no battery)?

That would take care of power conversion. You would also need an UVLO circuit to reduce power output when the panel drops below its lowest possible operating point. For a 12V panel that would be a voltage just below the MPPT voltage (to take temperature effects into account.) This will prevent those crashes described above by reducing power (and thus increasing impedance) when the panel starts to crash.

Or if you really, really want to spend the money, you could get something like a Midnite Kid. But there are three problems there:

1) that's a lot of functionality you don't need. Most of the time the CC will do nothing useful.
2) you need a separate supply to start it up. The Kid requires power on its battery terminals to start up. So you'd need a separate solar system to provide enough power to boot it.
3) you still need a DC/DC. As far as I know the output of the Kid won't go down to 5 volts. And since you don't want the two control loops fighting each other, you probably still need the UVLO.

MPPT buck converter that isn't optimized for battery recharging?

MPPT buck converter that isn't optimized for battery recharging?

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment