Hi all

I’ve scoured forums and unable to find something similar, wonder if someone might be able to shed some insight.

I’ve had and been successfully using a Kodak OG-Plus 6.2kw with a Magneto 48V 200ah battery connected to the grid. We have now moved it the offgrid location where it continued to work fine while we mounted the panels. BMS working smoothly.

Panels switched on (4x JA 460W), and they’re getting the wattage and voltage I’d expect (183V and 800W+ in late afternoon). It’s a completely clear day, and battery is at 40%.

However, PV input is switching on and off about every second. I can hear the inverter winding up and down, and see on the battery display it oscillated between +A and -A and the charging light flickers, pauses, flickers, pauses. On the inverter I don’t see the wattage go down to zero, goes from 800w to about 250W (and fluctuates between) and voltage fluctates 167-183V. But judging from the sound and the battery, it is going to 0W (there is only ~100W load) maybe for half a second and then bouncing back. I’ve only dared leave it on for 10 minutes to see if problem resolves, which it did not.

What might be the culprit here? As simple as a loose connection or something more problematic?

Help would be greatly appreciated.

Unfortunately I haven’t gotten Watchpower working on my devices, so unable to show you a graph right now.

Its the protection diode before the voltage measuring circuit.  May be bad soldering too

2 hours ago, martinmcg said:

Hi all

I’ve scoured forums and unable to find something similar, wonder if someone might be able to shed some insight.

I’ve had and been successfully using a Kodak OG-Plus 6.2kw with a Magneto 48V 200ah battery connected to the grid. We have now moved it the offgrid location where it continued to work fine while we mounted the panels. BMS working smoothly.

Panels switched on (4x JA 460W), and they’re getting the wattage and voltage I’d expect (183V and 800W+ in late afternoon). It’s a completely clear day, and battery is at 40%.

However, PV input is switching on and off about every second. I can hear the inverter winding up and down, and see on the battery display it oscillated between +A and -A and the charging light flickers, pauses, flickers, pauses. On the inverter I don’t see the wattage go down to zero, goes from 800w to about 250W (and fluctuates between) and voltage fluctates 167-183V. But judging from the sound and the battery, it is going to 0W (there is only ~100W load) maybe for half a second and then bouncing back. I’ve only dared leave it on for 10 minutes to see if problem resolves, which it did not.

What might be the culprit here? As simple as a loose connection or something more problematic?

Help would be greatly appreciated.

Unfortunately I haven’t gotten Watchpower working on my devices, so unable to show you a graph right now.

There could be several problems causing the charge controller to switch on and off:

1. Faulty charge controller.

2. Earth fault on one of the 4 panels.

3. Loose connection on mc4 connector or on the combiner.

As soon as the panels start delivering current the charge controller switch off hence the reason for the fluctuating voltage. I once had a Sunsynk 8kw that   would not switch on any of the mppt's even with adequate voltage from the string, after investigation found 1 panel in west facing string was leaking dc to earth and the inverter detected the earth fault.

9 minutes ago, TaliaB said:

. I once had a Sunsynk 8kw that   would not switch on any of the mppt's even with adequate voltage

On Sunsynk inverters if you ever get F24 DC insulation impedance failure then it is a solar panel earth fault either the pos or neg is touching ground or the panel is internally leaking to earth( frame)

18 hours ago, TaliaB said:

There could be several problems causing the charge controller to switch on and off:

1. Faulty charge controller.

2. Earth fault on one of the 4 panels.

3. Loose connection on mc4 connector or on the combiner.

As soon as the panels start delivering current the charge controller switch off hence the reason for the fluctuating voltage. I once had a Sunsynk 8kw that   would not switch on any of the mppt's even with adequate voltage from the string, after investigation found 1 panel in west facing string was leaking dc to earth and the inverter detected the earth fault.

Thanks Talia, Chris.

Haven’t yet had a chance to inspect earth & MC4, but have taken a multimeter to the breaker to check voltage. It is constant around 178V, and then when breaker is switched on (supplying inverter) the voltage is varying massively and quickly, 50V-185V (which does tally with what I saw on the inverter, but I figured that was downstream of the charge controller). Once switched off, goes back to constant (up/down 0.1V at most). In my mind this suggests 2 or 3 rather than a faulty charge controller, but I know electricity is more complex than that. Might this be indicative of where the problem may lie or definitely does not lie?

Thank you all for your help.

Edited May 9, 20242 yr by martinmcg

1 hour ago, martinmcg said:

Thanks Talia, Chris.

Haven’t yet had a chance to inspect earth & MC4, but have taken a multimeter to the breaker to check voltage. It is constant around 178V, and then when breaker is switched on (supplying inverter) the voltage is varying massively and quickly, 50V-185V (which does tally with what I saw on the inverter, but I figured that was downstream of the charge controller). Once switched off, goes back to constant (up/down 0.1V at most). In my mind this suggests 2 or 3 rather than a faulty charge controller, but I know electricity is more complex than that. Might this be indicative of where the problem may lie or definitely does not lie?

Thank you all for your help.

There is 1 test you can do to rule out the solar array(4panels). You have already done the Voc test on the array by switching the breaker off and measuring the voltage. Now this part you need to take care and use the correct isolator. Isolator rated not less than 250v DC 20A DC. Bridge the top and bottom contacts on the isolator then disconnect the pos and neg from the mppt. One wire to the top of the isolator and the other wire to the bottom. You will need a dc clamp meter to measure the Isc of the panels the reading in good sunlight need to be around 13A but should you get a slightly lower reading do not worry the important part is that current reading needs to be consistent and close to 13A(depending on solar irradiance)as possible. Should you take a voltage reading across the isolator the voltage will be 0v or close to it. Should the current not be consistent you have a bad or loose connection on one of the mc4 connectors or in the combiner.

Four panels might be too low for this model to get started. Although the MPPT range is  given as 120-450 VDC, it doesn't say what the start voltage is. It's often a bit higher than the minimum MPPT voltage, say 150 V±5 V. Your panels are showing around 167-183 V, but under load they may sag to around 152 V. If the starting voltage is at the upper end, say 155 V, they might not have enough voltage under load to reliably start the solar charger.

It seems unlikely, but possible.

On 2024/05/09 at 2:15 PM, TaliaB said:

There is 1 test you can do to rule out the solar array(4panels). You have already done the Voc test on the array by switching the breaker off and measuring the voltage. Now this part you need to take care and use the correct isolator. Isolator rated not less than 250v DC 20A DC. Bridge the top and bottom contacts on the isolator then disconnect the pos and neg from the mppt. One wire to the top of the isolator and the other wire to the bottom. You will need a dc clamp meter to measure the Isc of the panels the reading in good sunlight need to be around 13A but should you get a slightly lower reading do not worry the important part is that current reading needs to be consistent and close to 13A(depending on solar irradiance)as possible. Should you take a voltage reading across the isolator the voltage will be 0v or close to it. Should the current not be consistent you have a bad or loose connection on one of the mc4 connectors or in the combiner.

Thank you Talia. I’m not able to do this test until in the week as don’t have an isolator, rather a set of two circuit breakers, not sure why but that was the guide we were following. Or can it be done with circuit breakers?

One thing I don’t fully follow is “bridge the top and bottom contacts” - by this do you mean run a piece of PV wire between them? What is the role of the isolator then, is it not being short circuited? Sorry if a dumb question ☺️.

 

 

 

 

25 minutes ago, martinmcg said:

Thank you Talia. I’m not able to do this test until in the week as don’t have an isolator, rather a set of two circuit breakers, not sure why but that was the guide we were following. Or can it be done with circuit breakers?

One thing I don’t fully follow is “bridge the top and bottom contacts” - by this do you mean run a piece of PV wire between them? What is the role of the isolator then, is it not being short circuited? Sorry if a dumb question ☺️.

 

 

 

 

No problem there is never a dumb question. Just so that both contacts carry the load just in case, but make sure the cb is rated for dc. Maybe what @Coulombsuggested about the startup voltage of the 4 panels. I had a look at the panel specs and the Vmp of 4 in series is 150v might be the problem. If you could connect 5S1P( 5 panels in series will help with earlier startup in the morning as well. If you could try the extra panel just to ensure it is not a mppt startup issue. The symptoms does point to it as soon as the mppt draw current from the panels the voltage will drop so maybe you on the boderline of the starup voltage.

Below quick drawing should you do the Isc test.

 

 

On 2024/05/09 at 3:52 PM, Coulomb said:

Four panels might be too low for this model to get started. Although the MPPT range is  given as 120-450 VDC, it doesn't say what the start voltage is. It's often a bit higher than the minimum MPPT voltage, say 150 V±5 V. Your panels are showing around 167-183 V, but under load they may sag to around 152 V. If the starting voltage is at the upper end, say 155 V, they might not have enough voltage under load to reliably start the solar charger.

It seems unlikely, but possible.

 

On 2024/05/09 at 2:15 PM, TaliaB said:

There is 1 test you can do to rule out the solar array(4panels). Should the current not be consistent you have a bad or loose connection on one of the mc4 connectors or in the combiner.

So I have done some further tests that may offer some clues to the puzzle.

I found out that it does startup (rather than on/off) when the power is less than 600W. So I emulated that the next day by covering half a panel, it started up, then immediately uncovering and it kept working, going up to 1500W and staying working the whole (clear sky) day.

I left it on overnight and it kept working with the sunrise etc. At around 9:30am though, roughly 700W, it stopped working (fault on/off on/off) but this did coincide with the sun passing through the first clouds we’ve encountered.

To me this pointed to the startup voltage issue Coulomb pointed out, but, if I cover entirely one panel and keep it covered (lower VoC), it starts up and works. If I cover two panels, it doesn’t start up at all. But four panels, the highest voltage, it has the on/off on/off fault. The higher the peak wattage, the longer it stays on for, so at peak of the day it’ll be 2.5 seconds on, 0.5s off. Later it’ll be 1s on, 0.5s off. So it gets to peak watt before shutting off.

Something else is that with the multimeter on the PV input, even when the system is working 100% fine for hours and charging up the battery at 25A, the voltage is still fluctuating 40V-180V every second or so, is this not normal?

I also inspected and worked around with all the MC4s and didn’t notice any issues.

Thank you for all your help!

1 minute ago, TaliaB said:

No problem there is never a dumb question. Just so that both contacts carry the load just in case, but make sure the cb is rated for dc. Maybe what @Coulombsuggested about the startup voltage of the 4 panels. I had a look at the panel specs and the Vmp of 4 in series is 150v might be the problem. If you could connect 5S1P( 5 panels in series will help with earlier startup in the morning as well. If you could try the extra panel just to ensure it is not a mppt startup issue. The symptoms does point to it as soon as the mppt draw current from the panels the voltage will drop so maybe you on the boderline of the starup voltage.

Below quick drawing should you do the Isc test.

 

 

Thanks Talia, yes CBs are DC rated and 20A rated. Unfortunately a fifth panel will be a month away (we are in a remote location) when we return, but see my explanation above indicates a start-up issue but doesn’t seem to be voltage related.

2 minutes ago, martinmcg said:

Thanks Talia, yes CBs are DC rated and 20A rated. Unfortunately a fifth panel will be a month away (we are in a remote location) when we return, but see my explanation above indicates a start-up issue but doesn’t seem to be voltage related.

Is is possible to give us a graph of your pv voltage and current for 1 day that is if you have monotoring software running like Solar Assistant. Below is basically what the voltage graph should look like.

1 hour ago, TaliaB said:

Is is possible to give us a graph of your pv voltage and current for 1 day that is if you have monotoring software running like Solar Assistant. Below is basically what the voltage graph should look like.

I had been battling away trying to get the Watchpower wifi to work and having no luck, thinking I didn't have any of the right cables to connect to my PC, but fortunately my powerbank charging cable has just the right one and I've managed to connect it now.

Here you can see the PV watt and volts in the on/off phase, then as soon as we cover/uncover half the panel it evens out. I'm surprised, the multimeter still shows fluctuations but seems good at the inverter. Not sure if relevant, but with the circuit breaker off, the VoC is steady at 183V on the panel side, but it fluctuates on the inverter side of the circuit 20-40V.

The voltage fluctuations during the "fault" stage are also less severe than seen on the multimeter.

 

So I can't give you a day yet, and this laptop does need to be used at times unfortunately so constant monitoring is not possible. But I can show you any of the behaviours.

Will do the panel test later today or tomorrow.

 

Edit: the battery amps are also interesting: 

they match what the inverter shows, but the battery's display clearly shows it going to zero (or negative where there is output)- which I do trust more.

Edited May 11, 20242 yr by martinmcg

52 minutes ago, martinmcg said:

I had been battling away trying to get the Watchpower wifi to work and having no luck, thinking I didn't have any of the right cables to connect to my PC, but fortunately my powerbank charging cable has just the right one and I've managed to connect it now.

Here you can see the PV watt and volts in the on/off phase, then as soon as we cover/uncover half the panel it evens out. I'm surprised, the multimeter still shows fluctuations but seems good at the inverter. Not sure if relevant, but with the circuit breaker off, the VoC is steady at 183V on the panel side, but it fluctuates on the inverter side of the circuit 20-40V.

The voltage fluctuations during the "fault" stage are also less severe than seen on the multimeter.

 

So I can't give you a day yet, and this laptop does need to be used at times unfortunately so constant monitoring is not possible. But I can show you any of the behaviours.

Will do the panel test later today or tomorrow.

 

Edit: the battery amps are also interesting: 

they match what the inverter shows, but the battery's display clearly shows it going to zero (or negative where there is output)- which I do trust more.

Can you add the pv current graph this is weird behavior obviously the mppt is trying to switch on but for some reason it is switching off again and then finally switching on and stays on just above ~148v

Edited May 11, 20242 yr by TaliaB

3 hours ago, martinmcg said:

Something else is that with the multimeter on the PV input, even when the system is working 100% fine for hours and charging up the battery at 25A, the voltage is still fluctuating 40V-180V every second or so, is this not normal?

I'm thinking that this must be some artifact of your multimeter. I would certainly not expect the panel voltage to be collapsing to 40 V (unless that's a typo, supposed to be 140 V?). To collapse the panel voltage that low, the MPPT has to be trying to pull way too much current from the panels, or the panels or their wiring is intermittent, or something is breaking down and causing  the panels to be nearly short circuited. The fact that higher voltage seems to trigger the problem vaguely suggests something breaking down.

It's quite strange.

Thank you both for your help.

2 hours ago, TaliaB said:

Can you add the pv current graph this is weird behavior obviously the mppt is trying to switch on but for some reason it is switching off again and then finally switching on and stays on just above ~148v

 

MultiSIB doesn't seem to output this, and Watchpower has PV charging current but for some reason you can't get it in the data log. Do you want the PV current (ie at PV voltage) or PV charging current (at battery voltage)? I can derive the PV voltage from the graph or tomorrow I will record it off the inverter screen. Else can record the PV charging current off Watchpower.

 

58 minutes ago, Coulomb said:

I'm thinking that this must be some artifact of your multimeter. I would certainly not expect the panel voltage to be collapsing to 40 V (unless that's a typo, supposed to be 140 V?). To collapse the panel voltage that low, the MPPT has to be trying to pull way too much current from the panels, or the panels or their wiring is intermittent, or something is breaking down and causing  the panels to be nearly short circuited. The fact that higher voltage seems to trigger the problem vaguely suggests something breaking down.

It's quite strange.

Not a typo unfortunately, really odd. May have another multimeter I can access at the neighbours, I'll try find. Any further tests you'd suggest I run to isolate the source of the problem?

Edited May 11, 20242 yr by martinmcg

7 minutes ago, martinmcg said:

Thank you both for your help.

 

MultiSIB doesn't seem to output this, and Watchpower has PV charging current but for some reason you can't get it in the data log. Do you want the PV current (ie at PV voltage) or PV charging current (at battery voltage)? I can derive the PV voltage from the graph or tomorrow I will record it off the inverter screen. Else can record the PV charging current off Watchpower.

 

Not a typo unfortunately, really odd. May have another multimeter I can access at the neighbours, I'll try find. Any further tests you'd suggest I run to isolate the source of the problem?

I think to rule out pv loose connections or faulty panels try and do the Isc test and when the panels are bridged out climb on the roof and check hot connections on the mc4 connectors also check combiner for hot spots. Record the Isc value around 12h30 to 13h00 in full sun.

2 hours ago, martinmcg said:

Thank you both for your help.

 

MultiSIB doesn't seem to output this, and Watchpower has PV charging current but for some reason you can't get it in the data log. Do you want the PV current (ie at PV voltage) or PV charging current (at battery voltage)? I can derive the PV voltage from the graph or tomorrow I will record it off the inverter screen. Else can record the PV charging current off Watchpower.

 

Not a typo unfortunately, really odd. May have another multimeter I can access at the neighbours, I'll try find. Any further tests you'd suggest I run to isolate the source of the problem?

The request from TaliaB was to disconnect the PV to the inverter and place a short circuit on the + and -. This test with maximum current should show varying current if there is a loose connection. Coulomb also suspect there could be some loose connection. 

During this test neither watch power or the inverter itself will get any PV voltage. 

This can only be done if you have a DC clamp meter. 

8 minutes ago, Scorp007 said:

This can only be done if you have a DC clamp meter

Yes sure as this could exceed 13A and Multimeters don't scale above 10A.

7 minutes ago, TaliaB said:

Yes sure as this could exceed 13A and Multimeters don't scale above 10A.

Some of the last few yrs do at times go to 20A on the amp scale but they still use the same test leads and I think at 20A they can get hot and also the solder on the banana socket onto the PCB can melt with high heat. 

Edited May 11, 20242 yr by Scorp007

Just now, Scorp007 said:

Some of the last few yrs do at times go to 20A on the amp scale but they still use the same test leads and I think at 20A they can get hot and also the solder on the banana socket onto the PCB can melt with high heat. 

My Fluke 87V max will allow 20 A ac or dc for 30 sec then switch off terminals go open curcuit.

15 minutes ago, TaliaB said:

My Fluke 87V max will allow 20 A ac or dc for 30 sec then switch off terminals go open curcuit.

The above pic is our 1 workshop meter. Then about 5 yrs ago I bought a cheap China meter with big display to counter my eyes going off which happen to have the same scale layout. 

My Fluke 73 lasted a good 30yrs  It was also the 1st auto scale I had and I loved the fast bar graph at the bottom of the display. 

Edited May 11, 20242 yr by Scorp007

11 minutes ago, Scorp007 said:

The above pic is our 1 workshop meter. Then about 5 yrs ago I bought a cheap China meter with big display to counter my eyes going off which happen to have the same scale layout. 

My Fluke 73 lasted a good 30yrs  It was also the 1st auto scale I had and I loved the fast bar graph at the bottom of the display. 

The good old testers Scorp do you remember the English testers Simpsons  anciant i have one on the farm i did not even know it existed until my electrician found it in a box his face priceless. I will ask him to take a picture😁

1 minute ago, Derek3 said:

The good old testers Scorp do you remember the English testers Simpsons  anciant i have one on the farm i did not even know it existed until my electrician found it in a box his face priceless. I will ask him to take a picture😁

No i never knew those testers but I started my training using the AVO brand. Also used wheat stone bridge testers which I have never seen mentioned in any topic. 

1 minute ago, Scorp007 said:

No i never knew those testers but I started my training using the AVO brand. Also used wheat stone bridge testers which I have never seen mentioned in any topic. 

They where West Test Sets determining fault distance on Telephone lines where transposition cross over where used to calculated distance to short curcuit on Arial lines. In the good old days where 2 motion auto exchange was used, the first was in Braamfontein commissioned by SA Railways.