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

[thomm225]

Originally Posted by CatNewBee

No, they are not, better to measure Isc.

even damaged cells can output Uoc, but the current they output in the circuit is near zero. same for batteries, generators, alternators etc. Power sources can only be vetted under load.

I don't think you understand how to troubleshoot with voltage measurements alone

I was trying to make it simpler. You can monitor the voltage as the batteries charge.

If you charge the batteries from the same starting point with each panel separately and monitor the voltage closely each should charge the battery at about the same rate/in about the same time all else being equal. On my PWM Controller after the battery charges totally, panel side of controller will approach VOC

Like I said, trained troubleshooters don't need so much data to figure out a problem. We used to troubleshoot TTL Circuits with voltage measurements alone back in the day and this problem is much simpler than that

And untrained person can use the common sense knowledge gained from these simple tests to find the problem

Btw, I've had two (2) controllers fail and three (3) batteries over the last 7 years and zero (0) panel failures.

And I just ordered one of those Victron 75/15 MPPT Controllers so I can "play" with it. I'm a tech manager now and don't get to have fun troubleshooting as much. It's all about hiring, firing, interviewing, reports, medical, paychecks, timesheets, …….blah, blah...

[CatNewBee]

Originally Posted by thomm225

I don't think you understand how to troubleshoot with voltage measurements alone

I was trying to make it simpler. You can monitor the voltage as the batteries charge.

If you charge the batteries from the same starting point with each panel separately and monitor the voltage closely each should charge the battery at about the same rate/in about the same time all else being equal. On my PWM Controller after the battery charges totally, panel side of controller will approach VOC

Like I said, trained troubleshooters don't need so much data to figure out a problem. We used to troubleshoot TTL Circuits with voltage measurements alone back in the day and this problem is much simpler than that

And untrained person can use the common sense knowledge gained from these simple tests to find the problem

Btw, I've had two (2) controllers fail and three (3) batteries over the last 7 years and zero (0) panel failures.

And I just ordered one of those Victron 75/15 MPPT Controllers so I can "play" with it. I'm a tech manager now and don't get to have fun troubleshooting as much. It's all about hiring, firing, interviewing, reports, medical, paychecks, timesheets, …….blah, blah...

As I said before, measure the current. Panels connected to a battery, especially small ones to a big bank do not do much to the Voltage, you can sit for hours until there is a small change.

You can of course measure the current indirectly as voltage drop over an known resistor (shunt), but if you have problems understand what you want to achieve, take the easy path and do it directly on the currents. You can test 4 Panels in 2 Minutes, instead of waiting for hours until one panel has charged the battery, then discharge it to the same SOC, then take the next panel and hope the sun radiation is stil the same etc. No way you will get any reliable results.

[Paul Elliott]

Voc (open-circuit voltage) ans Isc (short-circuit current) are the parameters you want to check. If you can only check one, Isc is likely to give you the best idea about the health of the panel.

The OP did test Isc and Voc for his four panels, and if the tests were done properly, it looks like all the panels have a reasonable Voc, but the Isc varies significantly among the panels. #3 looks pretty good, #2 is bad, and #1, #4 are questionable:


But panel #2 shows up twice in the table, with significantly different Isc values. So, were these panels tested under similar light conditions, and were these measurements correctly performed? The panels should be tested individually, with no outside connections beyond the voltmeter (for Voc) and ammeter (for Isc).

[thomm225]

Originally Posted by CatNewBee

As I said before, measure the current. Panels connected to a battery, especially small ones to a big bank do not do much to the Voltage, you can sit for hours until there is a small change.

You can of course measure the current indirectly as voltage drop over an known resistor (shunt), but if you have problems understand what you want to achieve, take the easy path and do it directly on the currents. You can test 4 Panels in 2 Minutes, instead of waiting for hours until one panel has charged the battery, then discharge it to the same SOC, then take the next panel and hope the sun radiation is stil the same etc. No way you will get any reliable results.

Nope, you don't need to check the current. you can tell by the voltage. Just watch and see if the battery charges at the same rate on each panel.

You are on a different level.

If the controller were bad, I could repair that as well if I had the schematic and the manual and if the components could be soldered by hand. I haven't seen the circuit board so I can't say

He just needs to do the simple charging test with each panel with a battery discharged to the same level and record the data to determine the problem. (if all the panels are the same it's the controller, wiring, battery, or a connection.)

There are several other ways also. I was just mentioning the easiest especially if he has a few days to complete the testing

[curtis742]

Simple test, no meter required. Place panel in direct sunlight, take positive wire in right hand and negative wire in left hand. Touch both leads simultaneously to tongue. If pain is felt that panel works. Repeat with remaining panels one at a time. If a panel produces less or no pain it is defective
Check diode with ohm meter.

[Captain Bill]

If the time indicated on the solar controller is accurate, I don't see any problems here. 19 watts before 9am on a 50 watt panel is excellent. There is no way that a panel is going to be putting out even close to it's rated output at 0900 DST which is really 0800 solar time. In Idaho at solar noon (approx 1300) one would be lucky to see 90% of the rated output. Nearly 40% of rated output at that time of day is great. Panel number 2 may have an issue, but I don't see anything that alarms me about the other 3 panels. Remember obes eye adjusts to lighting conditions but the solar panel does not. What is subjectively similar lighting conditions to the eye can be quite different in actuality. The solar panels react only to the actual lighting conditions. If the Isc test was done on panel 2 even an hour before 3 and 4 conditions might have been actually quite different even if they appeared similar to the naked eye.

[curtis742]

In all seriousness, a solar panel should be facing directly at the sun for testing. The rated output is for facing directly at the sun with a temperature of ~70deg surface temperature. The higher the temperature the lower the output conversely the colder the cells are the more output will increase. If the voltage output is 90% of rated Voc or close the panel is fine. Be sure the sunlight is unobstructed during testing.

[Captain Bill]

Originally Posted by curtis742

In all seriousness, a solar panel should be facing directly at the sun for testing. The rated output is for facing directly at the sun with a temperature of ~70deg surface temperature. The higher the temperature the lower the output conversely the colder the cells are the more output will increase. If the voltage output is 90% of rated Voc or close the panel is fine. Be sure the sunlight is unobstructed during testing.

In addition it needs to be done near solar noon. Before about 11:30 dst at Idaho's lattitude there is simply too much atmosphere in the way to get anywhere near full output.

[DMacS]

Sorry thomm225, I have to agree with CatNewBee, Paul Elliott and others who stress the importance of measuring current

A nominal 12V panel consists of usually 36 individual cells in series each producing about 0.6V. A dead or dying cell may still be conductive so will not affect Voc by much but it will have an increased resistance that will pull down the voltage sharply as soon as current starts to flow. And at the end of the day it is current that charges your battery. One poor cell or bad connection between cells within the panel can bring down the whole panel and the whole string.

If the OP wants to try the direct connect test with each panel then I say good idea but put an ammeter in the circuit and measure the current. That will show how much charge is going in and saves waiting all day for the battery charge to increase only to find that the battery was goosed to start with.

I disagree that this is over complicating things. Anyone who can manage to measure Voc with a multimeter can surely flick it over to current and measure Isc. To eliminate any bad connections I would measure with the meter probes directly onto the panel connections inside the junction box on the back of the panel (unless this has been permanently sealed).

From the results Panel 2 looks like a problem but I am confused as to why there are two readings for this panel? If you leave the panel shorted out for 30 mins or so it may be possible to detect a "warm spot" if there is a resistance building up in one cell. A thermal imaging camera may be needed but that is over complicating. Also in the junction box on the back should be a diode in series with the panel to prevent the battery from discharging back through the panel. With the panel shorted or under load check the voltage dropped across the diode, should be about 0.7V for a silicone diode, or a little more depending on current, but would expect well under 1V. If it is high then diode is gone, Unlikely but worth a quick check as a new diode is cheap and from previous posts it sounds like you know how to use a soldering iron.


And as Captain Bill points out, morning, evening and high latitudes the sun comes through the atmosphere at an angle so has a lot of air to pass through. Depending how the manufacturers test their panels it may need to be high noon, near the equator, crystal clear skyes, elevated altitude and a cold panel before you see the full rated output, but all the panels should be similar.

[thomm225]

Originally Posted by DMacS

Sorry thomm225, I have to agree with CatNewBee, Paul Elliott and others who stress the importance of measuring current

It doesn't matter who agrees with whom. You can troubleshoot just with voltage readings. I've been doing it since the 1970's. (tech or tech manager for 43 years)

Knowledge of current is definitely helpful but many times troubleshooting electronics, current readings are unavailable unless you are going to cut a solder run and possibly damage a circuit board

Point is, we old electronic techs troubleshoot with voltage readings alone much of the time.

I can give you a simple example on my system (My Victron MPPT Controller hasn't arrived yet)

I can tell if my panels are working simply by monitoring the voltage on either side of my PWM controller.

Lets say I ran my fan all night and the depth finder on my boat after sailing all day with autopilot, VHF, etc on.

Voltage will be around 12.1 in the am. When the sun comes up if thing are working the voltage will start to rise.

Voltage on the panel side will equal voltage on the battery side until the voltage reaches the float setting of the controller

After the battery gets to float, the voltage on the panel side will slowly approach VOC because..........THE CURRENT IS DECREASING through the controller. But I'm only measuring voltage so how on earth would I know that???? I'm psychic maybe!

Anything different from this can be troubleshot with a common sense approach.

My idea was for the OP was to simply test each panel and make sure they charge the batteries at the same rate all conditions equal. Now if one panel if way off from the rest this is the time to get serious about getting more data. Even though it's pretty obvious the panel is malfunctioning.

At one point earlier though, OP thought all panels could possibly bad. He was trying to troubleshoot with too much info compared to his experience and getting mixed up. No offense OP

It's quite simple actually but non-tech's like to make it (troubleshooting) seem difficult

In circuit troubleshooting, if you find volt in an area near VCC or input voltage you know current flow has decreased or stop completely possibly due to an open resistor, bad transistor or chip you

This is if you are troubleshooting with a meter. If you have an O'scope you won't see any wave forms in this area. O'scope really simplify electronic troubleshooting

[thomm225]

I mean we are talking 4 solar panels, a controller, wire, connections, and some batteries right?

Don't make it harder than it is.........

These formulas will help those than don't have current readings

Watts = Amps * Voltage

Amps = Voltage/Resistance

If one variable changes, it will affect the rest

I actually sort of enjoy it when I have problems with my system....except when it was affecting my autopilot on a return trip home one time

The last time I had a wire almost totally broken in several places off one of my panels but still showed continuity.

That reminds me, don't forget to just troubleshoot your problem with just one panel at a time. Keep things simple. You can set up various tests and if you aren't sailing soon take your time and enjoy it

Here on our simulators we have it more complicated:

3 phase input power 60 and 400 hz (or a 400 hz motor generator powered by a phase of 60 hz)
visuals (as many as 10 projectors)
Image Generation (from computers to the projectors)
motion (hydraulic actuators.....computer driven)
control loading (pilot yoke feel) Roll, yaw, pitch (computer driven)
I/O (digital and analog depending on system age)
UPS Units (with as many as 48 batteries)

this times 18 systems (some with 1990's era mainframe computers and monitors)

Btw, we have no clamp on current meters on site to maintain all this

[basssears]

Thanks everyone for your input. I feel like I have done most everything that everyone's recommended, but it was over a few days and I may have missed some things or done some tests wrong.


I feel like I don't trust my somewhat cheapo clamp on meter (https://www.amazon.com/gp/product/B00K4FNSX2) but it's what I have to work with right now. Any way I could be using the clamp on wrong (I mean, seems pretty straightforward, set it to DC, zero the meter and clamp over the wire... right?)


I didn't realize I would be at such a disadvantage in the early morning.


Part of my frustration is I have a similar setup on my camper van (two 50 watt panels in series to make a 100w / 24v panel) and it has worked SO well... I got 95 watts out of it at about 7PM one evening with it pointed just right... just yesterday even though it was way of 90 degrees I was still getting 85 watts at about 4 in the afternoon.


But given the above input I am going to try one more round of testing incorporating everyone's input.


My plan is to test each panel separately (4 total) at around 1PM here in Idaho. It's going to be hot (90+) so I will try to compensate for that when analyzing my end results.


My plan with each panel is going to be:
1) Test Voc
2) Test Isc
3) Try testing the diode with the diodes function on the multimeter, otherwise test voltage drop across the diodes while shorted.

4) Connect to Victron 75/15 connected to pickup truck battery (not 4Runner, just in case there's something about that battery that's skewing things) with the headlights on, test amperage and assess what wattage the app claims I'm getting.
5) Connect to older Tracer MPPT controller in my camper van to see if indicated watts in changes significantly from what I just saw on Victron (no app for the Tracer but it does have a data screen showing what's going on).

6) Probably will connect my known-working 100w panel on camper van to the Victron once and see if it puts out expected wattage / equal wattage to the Tracer.


Then have a beer and hit myself in the head with a hammer... I"m sure everyone can't wait to see my "results"



-- Bass

[Captain Bill]

Originally Posted by basssears

Thanks everyone for your input. I feel like I have done most everything that everyone's recommended, but it was over a few days and I may have missed some things or done some tests wrong.


I feel like I don't trust my somewhat cheapo clamp on meter (https://www.amazon.com/gp/product/B00K4FNSX2) but it's what I have to work with right now. Any way I could be using the clamp on wrong (I mean, seems pretty straightforward, set it to DC, zero the meter and clamp over the wire... right?)


I didn't realize I would be at such a disadvantage in the early morning.


Part of my frustration is I have a similar setup on my camper van (two 50 watt panels in series to make a 100w / 24v panel) and it has worked SO well... I got 95 watts out of it at about 7PM one evening with it pointed just right... just yesterday even though it was way of 90 degrees I was still getting 85 watts at about 4 in the afternoon.


But given the above input I am going to try one more round of testing incorporating everyone's input.


My plan is to test each panel separately (4 total) at around 1PM here in Idaho. It's going to be hot (90+) so I will try to compensate for that when analyzing my end results.


My plan with each panel is going to be:
1) Test Voc
2) Test Isc
3) Try testing the diode with the diodes function on the multimeter, otherwise test voltage drop across the diodes while shorted.

4) Connect to Victron 75/15 connected to pickup truck battery (not 4Runner, just in case there's something about that battery that's skewing things) with the headlights on, test amperage and assess what wattage the app claims I'm getting.
5) Connect to older Tracer MPPT controller in my camper van to see if indicated watts in changes significantly from what I just saw on Victron (no app for the Tracer but it does have a data screen showing what's going on).

6) Probably will connect my known-working 100w panel on camper van to the Victron once and see if it puts out expected wattage / equal wattage to the Tracer.


Then have a beer and hit myself in the head with a hammer... I"m sure everyone can't wait to see my "results"



-- Bass

I looked up the specs on the renogy 50 watt panels and the maximum power point voltage is 18.75 volts. If your mppt controller is working properly the voltage on your series connected panels should be held between 35 and 38 volts on the input side if it's working properly assuming clear skies near local noon. People should be aware that the rating methodology on solar panels does not use sunshine. The panels are exposed to a light that has the specific output of 1000 watts per square meter and just gets a flash at that. There is no chance for the panel to heat up or any of the other issues that affect real solar panels out in the real world. The watt rating of a panel is useful for comparing the output to other panels and gives you a general idea of what the panel can do. It's not likely that you'll often see the rated output in the real world.

[CatNewBee]

Originally Posted by Captain Bill

I looked up the specs on the renogy 50 watt panels and the maximum power point voltage is 18.75 volts. If your mppt controller is working properly the voltage on your series connected panels should be held between 35 and 38 volts on the input side if it's working properly assuming clear skies near local noon. People should be aware that the rating methodology on solar panels does not use sunshine. The panels are exposed to a light that has the specific output of 1000 watts per square meter and just gets a flash at that. There is no chance for the panel to heat up or any of the other issues that affect real solar panels out in the real world. The watt rating of a panel is useful for comparing the output to other panels and gives you a general idea of what the panel can do. It's not likely that you'll often see the rated output in the real world.

Voc AND Isc will be enough to sort out panel problems AND pair the right panels to strings - if desired (chose for a pair panels with equal or similar Isc). If the values are too different, use all panels in parallel and not in serial strings, then you will have a better total output. Voc will be almost the same for all, because they have all the same number of cells and the same specs.

[Captain Bill]

Originally Posted by CatNewBee

Voc AND Isc will be enough to sort out panel problems AND pair the right panels to strings - if desired (chose for a pair panels with equal or similar Isc). If the values are too different, use all panels in parallel and not in serial strings, then you will have a better total output. Voc will be almost the same for all, because they have all the same number of cells and the same specs.

Voc and Isc will definitely sort out the panels, but it will not sort out the setup as requested by the OP. None of his posts so far has shown that his controller is working properly, though I suspect it is because he was doing his testing early in the morning before the panels were producing sufficient voltage under load. The information I posted about what he should expect the panel side voltage to be if the MPPT controller is working properly was to evaluate the controller, not the panels.