IV Curve with MPPT Charge Controller

[BackwoodsEE]

That is not a load you can test with. You should know that.

Huh? Of course it's a load. It's the PV array plus charge controller that's being tested. What else would it be? My device is passively inserted in the circuit between PV array and CC and watches how those two interact.

I've also tested just the PV array by inserting a second device that switches in some big electrolytic capacitors as a dynamic curve-tracing load. That works well and gives me IV curves for the array from 5V or so to Voc. But it's more interesting to see how the charge controller operates in situ.

[SunEagle]

Huh? Of course it's a load. It's the PV array plus charge controller that's being tested. What else would it be? My device is passively inserted in the circuit between PV array and CC and watches how those two interact.

I've also tested just the PV array by inserting a second device that switches in some big electrolytic capacitors as a dynamic curve-tracing load. That works well and gives me IV curves for the array from 5V or so to Voc. But it's more interesting to see how the charge controller operates in situ.

Uh. Actually a battery is a storage system and not a real load since it doesn't do any work or use energy. It just absorbs it or passes it on to a working load.

[BackwoodsEE]

Uh. Actually a battery is a storage system and not a real load since it doesn't do any work or use energy. It just absorbs it or passes it on to a working load.

Have we veered off into some semantic weeds here? I'm not trying to run Argonne National Laboratories, just observe how a test string of panels actually operates with an actual charge controller, which is connected in an actual way--to a battery with an inverter and some isolated AC loads.

I've seen the CC bulk charge the battery then float it, adjust rapidly to the inverter demanding more current, eke out the last few watts near dusk, and shut off to prevent backfeed at night. I've seen the different modules in the string produce or get bypassed under different levels of illumination. It's all been very interesting to watch, helped me choose a type of PV module for my planned "real" array, and impressed me with how well the TS-MPPT works. That's good enough for me.

[sensij]

Welcome to the forum.

It is sort of interesting that this CC does the full voltage sweep, instead of using more parametric type algorithms that just chase the maximum power point. I don't have an opinion on what is better, but not every CC would give such nice data to show what it is doing.

[NEOH]

Uh. Actually a battery is a storage system and not a real load since it doesn't do any work or use energy. It just absorbs it or passes it on to a working load.

Uh. Actually, when the battery is being charged, it is a load on the Charge Controller.
And if the Charge Controller is connected to only a battery, then the battery is THE load - a real load.
The battery is doing work - it is chemically converting Lead Sulfate & Water back into Lead, Lead Oxide & Sulfuric Acid.
The battery does use energy.
What the battery does with that stored chemical energy does not change the fact that charging a battery is load on the controller.

A secondary battery can be is either:
a) a load when it is being charged or
b) an energy source when it is being discharged

During Bulk Mode, a 50% DOD battery is huge load on the charge controller.

[BackwoodsEE]

One thing I've found interesting from watching these plots is how the CC strives to wring out every last watt in cloudy conditions and as the sun goes down. Just now it was getting a whole 2 watts out of the three series-connected panels near dusk, with the sun well below the trees. Photovoltaic cells scale their current output proportional to light intensity; I don't know how linear the relationship is, but it never seems to drop to zero.

I've even seen a few volts show up on the panels at night just from turning a light on inside the house and some of it hitting the array through the windows. Of course in that case, with the CC disconnected to prevent battery backfeed, there is a microscopic amount of PV current flow.

Capture.PNG

[BackwoodsEE]

Welcome to the forum.

Thanks! It's been extremely informative.

It is sort of interesting that this CC does the full voltage sweep, instead of using more parametric type algorithms that just chase the maximum power point. I don't have an opinion on what is better, but not every CC would give such nice data to show what it is doing.

It doesn't always do the full sweep. Sometimes it just does a small quick perturbation of voltage (and of course current) and then gets back to business. I've also noticed that the voltage often (usually?) stays constant through short-term changes in sun intensity from clouds passing overhead. The CC seems to pick the optimal MPPT voltage and then stick with it until it decides to look for another one. How it decides it's time to do a sweep in search of a new MPPT (at the cost of up to a second or so in lost power production) is probably a topic of some interesting programming inside the CC.

[Sunking]

Backwoods I do not think you are understanding what I am saying. Testing a MPPT Controller by charging a battery is not giving you accurate data.

Try something like this. Discharge the battery to say 50% DOD, Say your test panel is 300 watts, and you have a 12 volt 100 AH Battery. Connect an Inverter to the partially discharged panel, and use a load that is equal to or greater than the panel wattage. That is going to force your MPPT controller to go to 100% power extracting every what the panel can provide.

Then monitor both Input/output voltage and current of the controller. Do that and you will see what it can really do. Example efficiency; Output Power / Input Power x 100

[sensij]

Backwoods I do not think you are understanding what I am saying. Testing a MPPT Controller by charging a battery is not giving you accurate data.

Try something like this. Discharge the battery to say 50% DOD, Say your test panel is 300 watts, and you have a 12 volt 100 AH Battery. Connect an Inverter to the partially discharged panel, and use a load that is equal to or greater than the panel wattage. That is going to force your MPPT controller to go to 100% power extracting every what the panel can provide.

Then monitor both Input/output voltage and current of the controller. Do that and you will see what it can really do. Example efficiency; Output Power / Input Power x 100

In the bulk stage of charging, his system is extracting every available watt. What you are suggesting is pointless, it will not induce the controller to generate any more power under equivalent conditions than what has already been described.

Are you objecting to the fact that the voltage and current lines look flat? The timescale is *seconds*... The motion of the sun will look flat on that scale.

[sensij]

How it decides it's time to do a sweep in search of a new MPPT (at the cost of up to a second or so in lost power production) is probably a topic of some interesting programming inside the CC.

How often are you seeing it run the sweep (or make a perturbation)? Does it update less often in clear conditions?

[bcroe]

One thing I've found interesting from watching these plots is how the CC strives to wring out every last watt in cloudy conditions and as the sun goes down. Just now it was getting a whole 2 watts out of the three series-connected panels near dusk, with the sun well below the trees. Photovoltaic cells scale their current output proportional to light intensity; I don't know how linear the relationship is, but it never seems to drop to zero.

I don't recall ever seeing a plot of the panel load curve voltage shifting (downward?) at very low
levels. My inverters will keep running down to about 1/3% of max. The trouble with checking
their voltage down there is a lot of bypass diodes may be coming into effect. There will be a
drop off point, as has been seen for open circuit under full moon. Bruce Roe

[SunEagle]

Have we veered off into some semantic weeds here? I'm not trying to run Argonne National Laboratories, just observe how a test string of panels actually operates with an actual charge controller, which is connected in an actual way--to a battery with an inverter and some isolated AC loads.

I've seen the CC bulk charge the battery then float it, adjust rapidly to the inverter demanding more current, eke out the last few watts near dusk, and shut off to prevent backfeed at night. I've seen the different modules in the string produce or get bypassed under different levels of illumination. It's all been very interesting to watch, helped me choose a type of PV module for my planned "real" array, and impressed me with how well the TS-MPPT works. That's good enough for me.

I am impressed with your actions and graphs. I was just putting my 2 cents in when it comes to what a "load" is or isn't.

If you want to use your battery for those curves then fine. Just be aware that using the battery as your "load" may not be as accurate as another device.

[SunEagle]

Uh. Actually, when the battery is being charged, it is a load on the Charge Controller.
And if the Charge Controller is connected to only a battery, then the battery is THE load - a real load.
The battery is doing work - it is chemically converting Lead Sulfate & Water back into Lead, Lead Oxide & Sulfuric Acid.
The battery does use energy.
What the battery does with that stored chemical energy does not change the fact that charging a battery is load on the controller.

A secondary battery can be is either:
a) a load when it is being charged or
b) an energy source when it is being discharged

During Bulk Mode, a 50% DOD battery is huge load on the charge controller.

You are correct that using energy to convert chemicals during a storage process may be considered "work". Although to some people it is not the same as something that converts energy to mechanical motion or to create light.

[Sunking]

In the bulk stage of charging, his system is extracting every available watt. What you are suggesting is pointless,.

No it is not pointless. Sure you know anything about CC's? The voltage should never dip below Vmp for any length of time in Bulk. Only when it goes into PWM mode aja Absorb and Float will you see voltage go below Vmp.

[BackwoodsEE]

How often are you seeing it run the sweep (or make a perturbation)? Does it update less often in clear conditions?

It seems to be every few minutes when there is a bumpy IV curve due to shading and, yes, less often when it is clear. I've seen it go quite a long time under clear skies with just small voltage deviations--maybe half an hour or more.