diy solar

diy solar

Capacitors on PV ends to resolve fluctuation?

mnakkach

New Member
Joined
Feb 3, 2022
Messages
101
We struggled for 3 weeks at least trying to troubleshoot the reason why watts on PV output fluctuates drastically during load.

I have 3 Voltronic Infinisolar V IV hooked in parallel

28 Longi panels 545w

inverter1, 7 in series
inverter2, 7 in series
interter3, 7 in series+7 in series

During high load, i can see PV power on the 3 inverters fluctuates, 3300w suddenly drops down to 200w for 3 seconds and then go back to 3300w, drops are high, random amounts of watts and almost continuous.

The drops will reduce if i power off 1 inverter and keep only 2 in parallel, drops will vanish completely if i power off 2 inverters and keep only 1 inverter.

Reviewed all connectors, all cables, thickness, earth, etc... everything is as per recommended standards, max cables length is 30m, used 10mm DC cables.

It worth to mention I was having Growatt inverters on the same setup and cabling without a single drop.

After contacting Voltronic, they recommended adding 3.3uf 600v DC capacitors on the PV ends at the inverter's connector side. I found 400v 3.3 uf DC and installed them, they resolved around 80% of the problem, fluctuation is still there by around 20% but not completely resolved.

Anyone faced such a problem or have an opinion what could be the problem or solution?
 
I've observed similar behavior on other gear and assumed it was MPPT recalculating to make sure it's in the right spot on the power curve knee. 3 seconds sound like long time though and that's not what Voltronic said so that's just my 2 cents.
 
I've observed similar behavior on other gear and assumed it was MPPT recalculating to make sure it's in the right spot on the power curve knee. 3 seconds sound like long time though and that's not what Voltronic said so that's just my 2 cents.
yes and during the drops it will use the battery power to cover the load, when PV recovers then batteries will be charged back. batteries keep charging/discharging continuously.
 
It may be voltage slumping on the HV DC bus within the inverter. Many of the cheaper HF inverters do not have much capacitor filtering on their HV DC bus. The HV DC bus has some margin voltage above the peak of sinewave AC output voltage to allow for some amount of HV DC ripple voltage slump that occurs at twice the AC line frequency due to AC loads.

The PV SCC is direct coupled boost converter feeding the HV DC bus. Because it is boost only converter, the PV inputs are not allowed to exceed the voltage of inverter internal HV DC bus. If AC loading causes ripple on HV DC that drops below PV input voltage the SCC will shut down while PV is above HVDC. To get maximum output from PV panels the MPP loading must be constant based on available panel illumination current. If PV loading has ripple it significantly impacts the ability to pull optimum power from PV array.

Extra capacitance on PV input would likely make the issue worse as it makes the PV voltage stay higher longer. It may help if poor SCC feedback control stability is causing additional oscillation to SCC boost converter.,

Usually is worse when PV array is operating close to the maximum allowable PV input voltage limit. You could try dropping the PV input voltage by removing one series panel to see if it reduces or eliminates issue.

When running parallel inverters there is some tolerance in their output voltage matching. This can cause some of the inverters to take more or less of the percentage of AC load. Inverter taking excess percentage of AC load will be more likely to develop excessive HV DC bus ripple.

I assume you know not to parallel connect inverter PV inputs from different inverters from same PV array. Each inverter needs its own string to control its own Vmp point. Also, if you have some of the parallel inverters with no PV inputs or lower PV power it can further complicate the AC loading effect. Your best bet is distributing PV strings of equal power capability, at lower array voltage, across all the inverters.
 
Last edited:
It may be voltage slumping on the HV DC bus within the inverter. Many of the cheaper HF inverters do not have much capacitor filtering on their HV DC bus. The HV DC bus has some margin voltage above the peak of sinewave AC output voltage to allow for some amount of HV DC ripple voltage slump that occurs at twice the AC line frequency due to AC loads.

The PV SCC is direct coupled boost converter feeding the HV DC bus. Because it is boost only converter, the PV inputs are not allowed to exceed the voltage of inverter internal HV DC bus. If AC loading causes ripple on HV DC that drops below PV input voltage the SCC will shut down while PV is above HVDC. To get maximum output from PV panels the MPP loading must be constant based on available panel illumination current. If PV loading has ripple it significantly impacts the ability to pull optimum power from PV array.

Usually is worse when PV array is operating close to the maximum allowable PV input voltage limit.
What could be the solution? a different capacitor if not a different inverter?
 
What could be the solution? a different capacitor if not a different inverter?
I think RcinFLA is spot on with suggesting to remove a panel or two from each string for testing.

If that solves the problem then you you will use a seperate charge controller for the panels that you removed.
 
It may be voltage slumping on the HV DC bus within the inverter. Many of the cheaper HF inverters do not have much capacitor filtering on their HV DC bus. The HV DC bus has some margin voltage above the peak of sinewave AC output voltage to allow for some amount of HV DC ripple voltage slump that occurs at twice the AC line frequency due to AC loads.

The PV SCC is direct coupled boost converter feeding the HV DC bus. Because it is boost only converter, the PV inputs are not allowed to exceed the voltage of inverter internal HV DC bus. If AC loading causes ripple on HV DC that drops below PV input voltage the SCC will shut down while PV is above HVDC. To get maximum output from PV panels the MPP loading must be constant based on available panel illumination current. If PV loading has ripple it significantly impacts the ability to pull optimum power from PV array.

Extra capacitance on PV input would likely make the issue worse as it makes the PV voltage stay higher longer. It may help if poor SCC feedback control stability is causing additional oscillation to SCC boost converter.,

Usually is worse when PV array is operating close to the maximum allowable PV input voltage limit. You could try dropping the PV input voltage by removing one series panel to see if it reduces or eliminates issue.
The max PV input as per the inverter is 400v, i put 7 in series, each is 49.5voc, i am getting 332v at peak sunny hours without load, during load it goes down to 270v.

Do you think it worth trying since it is way below the max?
 
The max PV input as per the inverter is 400v, i put 7 in series, each is 49.5voc, i am getting 332v at peak sunny hours without load, during load it goes down to 270v.

Do you think it worth trying since it is way below the max?
This is the related DS

1662380890600.png
 
The max PV input as per the inverter is 400v, i put 7 in series, each is 49.5voc, i am getting 332v at peak sunny hours without load, during load it goes down to 270v.

Do you think it worth trying since it is way below the max?
49.5voc is an unusual panel Voc voltage. It would make them about 76 cell panels. 72 cell panel would be more likely although that is also unusual.

Panel specs are usually stated at 20 or 25 degs C. With sun heating the Voc will be lower.

Vmp typically runs 0.81 to 0.85 x Voc voltage. When you say you have 332v at peak sunny day I assume you are talking Vmp and that would be about 400v Voc.
 
It is typical for 230vac inverter to have a HV bus of 500 vdc. Most 120vac inverter are limited to 250vdc max PV input due to their lower HV DC bus of 250-300vdc. Some 120vac inverters have an additional DC-DC converter between SCC and HV DC to converter 500vdc from PV SCC down to 250 vdc to give them 500vdc max PV input capability.
 
I would start by removing one of the 7 + 7 series arrays on the one inverter to see if matching all inverters helps. If the arrays are facing different directions you will not get similar PV power between the arrays.

With a moderate AC load, check all inverter AC output currents with a clamp-on amp meter to see how well they are matching in load distribution.
 
Last edited:
49.5voc is an unusual panel Voc voltage. It would make them about 76 cell panels. 72 cell panel would be more likely although that is also unusual.

Panel specs are usually stated at 20 or 25 degs C. With sun heating the Voc will be lower.

Vmp typically runs 0.81 to 0.85 x Voc voltage. When you say you have 332v at peak sunny day I assume you are talking Vmp and that would be about 400v Voc.
That's the DS of my panels

I meant i am seeing 332v from PV on the LCD screen on the inverter on a sunny day before putting any load, during load it goes down to 270v
1662382368606.png
 
Okay they are 72 cell panels.

Check each inverter AC output current with moderate AC load to see how well they are matching delivered current.

Since you have each inverter with its own array is the PV power variation only on particular inverters? If one inverter is taking more of the AC load current and it is also the one showing more of the PV power variation it definitely would point to the problem.

Another thought, if the inverters are freaking out with AC output voltage regulation due to the parallel inverters' interaction this could also be the cause. The AC load current on each inverter would be jumping around if this is the case.

Your setup with multiple inverters, each having PV arrays, makes a very complicated interacting feedback control problem for the inverters to handle.
 
Last edited:
Is there a battery involved in this set up? What specs for the battery?
 
Okay they are 72 cell panels.

Check each inverter AC output current with moderate AC load to see how well they are matching delivered current.

Since you have each inverter with its own array is the PV power variation only on particular inverters? If one inverter is taking more of the AC load current and it is also the one showing more of the PV power variation it definitely would point to the problem.

Another thought, if the inverters are freaking out with AC output voltage regulation due to the parallel inverters' interaction this could also be the cause. The AC load current on each inverter would be jumping around if this is the case.

Your setup with multiple inverters, each having PV arrays, makes a very complicated interacting feedback control problem for the inverters to handle.
Fluctuation is happening on all the 3 inverters, but there is a noticed pattern, before i installed the capacitor, fluctuation would start when load increases above 15A, while after the capacitors were installed, fluctuations started when load goes higher than 35A only, below this number no fluctuations is noticed with capacitors installed

AC load is split equally across the 3 inverters but PV input is for sure more on the inverter 3 since it has 14 panels while others have 7 panels only

Another important point, without capacitors, AC load at 20A, when i shutdown inverter3 (the one having 14 panels), fluctuation remains there for inverter 1 and 2, I keep load at 20A and i shutdown inverter1 or inverter2, fluctuation disappear and PV input stabilizes.
 
It may be voltage slumping on the HV DC bus within the inverter. Many of the cheaper HF inverters do not have much capacitor filtering on their HV DC bus. The HV DC bus has some margin voltage above the peak of sinewave AC output voltage to allow for some amount of HV DC ripple voltage slump that occurs at twice the AC line frequency due to AC loads.

The PV SCC is direct coupled boost converter feeding the HV DC bus. Because it is boost only converter, the PV inputs are not allowed to exceed the voltage of inverter internal HV DC bus. If AC loading causes ripple on HV DC that drops below PV input voltage the SCC will shut down while PV is above HVDC. To get maximum output from PV panels the MPP loading must be constant based on available panel illumination current. If PV loading has ripple it significantly impacts the ability to pull optimum power from PV array.

Extra capacitance on PV input would likely make the issue worse as it makes the PV voltage stay higher longer. It may help if poor SCC feedback control stability is causing additional oscillation to SCC boost converter.,

Usually is worse when PV array is operating close to the maximum allowable PV input voltage limit. You could try dropping the PV input voltage by removing one series panel to see if it reduces or eliminates issue.

When running parallel inverters there is some tolerance in their output voltage matching. This can cause some of the inverters to take more or less of the percentage of AC load. Inverter taking excess percentage of AC load will be more likely to develop excessive HV DC bus ripple.

I assume you know not to parallel connect inverter PV inputs from different inverters from same PV array. Each inverter needs its own string to control its own Vmp point. Also, if you have some of the parallel inverters with no PV inputs or lower PV power it can further complicate the AC loading effect. Your best bet is distributing PV strings of equal power capability, at lower array voltage, across all the inverters.
I have 8 MGX 5048 inverters. But i could never manage to let them work in parallel. MPP advice me to put 3.3 uF caps at the pv input of each inverter. But still I get them not running in parallel. I get F72 and other eror codes. Any idea what to do or how to solve?
 
MPP advice me to put 3.3 uF caps at the pv input of each inverter.
That sounds like they have some kind of problem of RF noise leaking out to the PV circuits that is then impacting the operation of the MPPT. If that is the case, it would be kinda weird and would be a sign of a poorly designed and tested system. (Anyone designing an inverter must expect RFI to be a potential problem. Inverters turn high currents on and off at a high frequency....and that will generate a lot of noise.)
 
Back
Top