Troubleshooting solar setup... or troubleshooting my troubleshooting!!

[Paul Elliott]

And if you want to calculate the effects of battery voltage and battery resistance on no-controller solar panel charging current, the (very approximate) effective solar panel voltage is 375V (three hundred seventy five volts), and the effective panel series resistance is 132 Ohms. This is a linear approximation of the OP's 50W solar panel between the Isc (short-circuit current, 2.84A) and Imp (maximum power point current, 2.7A @ 18.5V) points on the V-I curve.

Battery voltage and series resistance have very little effect on charging current when using a solar panel, at least until well past the overcharge point. Look at those numbers again: 375V and 132 Ohms for a 36-cell 50W panel. This is what a constant-current source looks like.

(Actually, an ideal constant-current source has nearly infinite voltage and resistance. The ratio of these two values is what sets the current. I'm just killing time until we hear more from the OP.)

[CatNewBee]

Originally Posted by Paul Elliott

And if you want to calculate the effects of battery voltage and battery resistance on no-controller solar panel charging current, the (very approximate) effective solar panel voltage is 375V (three hundred seventy five volts), and the effective panel series resistance is 132 Ohms. This is a linear approximation of the OP's 50W solar panel between the Isc (short-circuit current, 2.84A) and Imp (maximum power point current, 2.7A @ 18.5V) points on the V-I curve.

Battery voltage and series resistance have very little effect on charging current when using a solar panel, at least until well past the overcharge point. Look at those numbers again: 375V and 132 Ohms for a 36-cell 50W panel. This is what a constant-current source looks like.

(Actually, an ideal constant-current source has nearly infinite voltage and resistance. The ratio of these two values is what sets the current. I'm just killing time until we hear more from the OP.)

This is BS. 1st, a solar panel is a nonliner device regarding resistance, where do you get the 375V from? The Voc is 20V, the Isc is 3A, both points in the curve are not usable for resistance calculation, nor for power, because one parameter is zero. you must calculate the values along the curve when there none of voltage nor current are zero. Best case get the maximum power point Vmp and Imp. so you can find out what load could absorb the maximum output. will be around R=17V/2.8A ~6 Ohm and P=17V × 2.8A = 50W.

[bluefin3]

Hi I have read with intrest all the technical posts and can not recall one were the panels were tested connected to a resistive load like a 35 watt lamp, or 50 watt lamp, to determine if panels are truly capable providing the required current. Following the above if they can provide the required voltage current then look at the charging apparatus. Check the apparatus by first connecting to the resistive load before connecting to battery. Simplifying always works best for me an ex electrician
Hope this helps

[thomm225]

Originally Posted by bluefin3

Hi I have read with intrest all the technical posts and can not recall one were the panels were tested connected to a resistive load like a 35 watt lamp, or 50 watt lamp, to determine if panels are truly capable providing the required current. Following the above if they can provide the required voltage current then look at the charging apparatus. Check the apparatus by first connecting to the resistive load before connecting to battery. Simplifying always works best for me an ex electrician
Hope this helps

Simple was mentioned repeatedly. The OP's last post mentioned he was taking a break from trying to fix his problems for a while. Could be after 100 or so posts with recommendations as to how to fix his problem with only 4 panels, one controller, wire, connections and batteries he got burnt out

If all else fails, forget all theory and start with a battery and a panel and go from there

[skenn_ie]

Originally Posted by Paul Elliott

And if you want to calculate the effects of battery voltage and battery resistance on no-controller solar panel charging current, the (very approximate) effective solar panel voltage is 375V (three hundred seventy five volts), and the effective panel series resistance is 132 Ohms. This is a linear approximation of the OP's 50W solar panel between the Isc (short-circuit current, 2.84A) and Imp (maximum power point current, 2.7A @ 18.5V) points on the V-I curve.

Battery voltage and series resistance have very little effect on charging current when using a solar panel, at least until well past the overcharge point. Look at those numbers again: 375V and 132 Ohms for a 36-cell 50W panel. This is what a constant-current source looks like.

(Actually, an ideal constant-current source has nearly infinite voltage and resistance. The ratio of these two values is what sets the current. I'm just killing time until we hear more from the OP.)

I have long wondered what the equivalent circuit looked like. How about capacitance though, particularly relevant when the load on the panel is a PWM controller. Without a substantial capacitance in the input side of the PWM, I suspect that it is largely ineffective at holding the MPP.

[skenn_ie]

Originally Posted by bluefin3

Hi I have read with intrest all the technical posts and can not recall one were the panels were tested connected to a resistive load like a 35 watt lamp, or 50 watt lamp, to determine if panels are truly capable providing the required current. Following the above if they can provide the required voltage current then look at the charging apparatus. Check the apparatus by first connecting to the resistive load before connecting to battery. Simplifying always works best for me an ex electrician
Hope this helps

Don't forget: If the expected power is anywhere near the rated lamp power, use a 24V lamp rather than a 12V one (or connect two in series), as the panel voltage may be as high as 18V. (60y/o electronics engineer)

[CatNewBee]

Originally Posted by skenn_ie

I have long wondered what the equivalent circuit looked like. How about capacitance though, particularly relevant when the load on the panel is a PWM controller. Without a substantial capacitance in the input side of the PWM, I suspect that it is largely ineffective at holding the MPP.

PWM controller do not care about MPP, they act as a shortcut between panel and battery, until the absorption voltage is reached, and then start to pulse to reduce power and prevent overcharging / overvoltage. They are just a smart switch.

[skenn_ie]

Originally Posted by CatNewBee

PWM controller do not care about MPP, they act as a shortcut between panel and battery, until the absorption voltage is reached, and then start to pulse to reduce power and prevent overcharging / overvoltage. They are just a smart switch.

A lot of them are sold AS MPP controllers though.

[CatNewBee]

Originally Posted by skenn_ie

A lot of them are sold AS MPP controllers though.

... the usual scam ... Some seller even do not know what MPP means, they think its a quality label like a CE Certificate or ISO9001 label...

You'll find it out very easy with a look at the data sheet.

[Paul Elliott]

Originally Posted by CatNewBee

This is BS. 1st, a solar panel is a nonliner device regarding resistance, where do you get the 375V from? The Voc is 20V, the Isc is 3A, both points in the curve are not usable for resistance calculation, nor for power, because one parameter is zero. you must calculate the values along the curve when there none of voltage nor current are zero. Best case get the maximum power point Vmp and Imp. so you can find out what load could absorb the maximum output. will be around R=17V/2.8A ~6 Ohm and P=17V × 2.8A = 50W.

Wow, where do I start? Or better yet, where should you start?

First of all, please learn to read. I said "approximate" "linear equivalent" and "over the range from Isc to Imp". This is how you can understand the constant current equivalence of a solar panel, by comparing it to an ideal constant-current source. Of course there's no actual 375V or 132 Ohms anywhere in the panel, but if you want to model the equivalent constant-current source as a voltage source with series resistance, those are the values for the voltage and resistance. Do the math yourself if you don't believe me. If you want to get a closer approximation to the actual behavior, increase that 375V, increase the 123 Ohms, and add 36 silicon diodes in series across that constant-current source. You will get a circuit that pretty well matches the panel over the full range (Voc to short-circuit).

And this has nothing to do with the maximum power available from the panel. That's a different set of calculations and you end up somewhere on the knee of the V-I curve, where the current is starting to fall off.

The only reason I'm bothering with this is that some are suggesting that charging a small battery with a big power supply shows the same behavior as you see when charging a big battery with a small solar panel. There are big differences.

As for cell capacitance and PWM chargers, I doubt that the capacitance is large enough to matter. The battery certainly won't care. I doubt that it makes much difference if the PWM is switching at 1 Hz or at 100 KHz.

An MPPT charger is a different thing, but the significant difference is on the panel side, not at the battery side. The MPPT charger draws a controlled amount of current from the panel while monitoring the panel voltage, and adjusting that current draw until V*I (power) is maximized. The MPPT controller then uses switching-regulator techniques to deliver this power to the battery at the battery voltage. This power is reduced by the controller as the battery voltage increases over time. This switching and conversion may all be done in a single switching stage, but the principle is as I described.

[Paul Elliott]

I have used series-parallel strings of incandescent lights, as well as large power resistors to test solar panels. I've also used power supplies to test the reverse-bias current in shaded panels. The panels behave as they should, and ultimately you really don't need much more than your Voc and Isc measurements and the panel spec sheet to tell if the panel is OK.

There was one time those tests didn't discover a problem though. A 100W Renogy panel was delivering full power, but when I partially shaded the panel the bypass diode didn't shunt the current around the shaded section. I discovered that the bypass diode connections were not properly soldered, but after I touched up that connection the panel worked as designed.

[Paul Elliott]

Originally Posted by Paul Elliott

Wow, where do I start? Or better yet, where should you start?

Let me apologize in advance -- I suspect that we are discussing different issues and talking past one another. But I do stand by my technical description of the equivalent constant-current source characteristics of the solar panel. You appear to be concerned with other parameters.

[CatNewBee]

Originally Posted by Paul Elliott

Let me apologize in advance -- I suspect that we are discussing different issues and talking past one another. But I do stand by my technical description of the equivalent constant-current source characteristics of the solar panel. You appear to be concerned with other parameters.

No problem. It's a complex matter. I agree, it's technically a current source. The I/U diagrams related to radiation intensity shows this quite impressivly, while the current is almost proportional to the radiation, the voltage changes very little.

I guess we lost the crowd with our posts. Better get back to "twist the wire ends to the left and lick on them to tell if it's working...", it's more fun than boring theory.

[thomm225]

Originally Posted by CatNewBee

I guess we lost the crowd with our posts. Better get back to "twist the wire ends to the left and lick on them to tell if it's working...", it's more fun than boring theory.

Yeah, you need to know enough theory to be dangerous.

The rest can be completed with proper troubleshooting technique.

I was trained in the 70's, (75-78, military) learned computers OJT, and started running a maintenance site in 1995 so I'm still relying on my memory of the lessons we had in the 70's.

One of which was from this engineer from Hewlett Packard teaching us to troubleshoot power supply circuits with voltage readings only

This guy look just like a NASA Engineer you see on the documentaries from the time. White shirt and tie, pocket protector full of pens and pencils, brylcreem hair..........he was awesome. His name may have been George Stanley.

Wow, it was George Stanley:

https://www.youtube.com/watch?v=ftLQiO1H7G0

[Paul Elliott]

Actually, I find the Turbo Encabulator to be an invaluable part of my solar panel testing toolkit:

https://www.youtube.com/watch?v=RXJKdh1KZ0w