SunGold Power SP6548 Flickering lights after new install

[Lilikoi_Heavy_Industries]

Shutter speed on the phone got me on this one but you can still the wave good enough. Output from the inverter with PV and Grid input disconnected:

It might be good to add your entire troubleshooting debacle onto your youtube videos in some way so they have some context of all that you've been through. The video title should have the inverter makers' name in it as well as the model number so that other people who are considering this inverter would find it as a resource.

I understand the quick upload is extremely helpful for sharing data, thanks for coming through with some scope readings!

 

[nofear72784]

What's stopping you from opening a dispute on paypal? "Wait a few days" is just checking to see if you have so much money that you don't mind losing it that much.

I don't want to publicly bash a company without at least giving them a shot to do the right thing. I am at the end of my rope though. We took out a loan to purchase everything and went quite a bit over due to the unforeseen issues with the city inspector. So, there is no money and yes we very much mind loosing it.

 

[Quattrohead]

Your mains input looks horrible. Can you scope it without the inverter on or connected at all.

 

[Quattrohead]

Here is my mains 120v, not exactly pretty not no where near as bad as yours.


And as a matter of interest whilst I had the scope up, here is the output of a Growatt 6k LF unit, beautiful waveform with no load.

 

[Quattrohead]

My 3 ton AC compressor just kicked on, the Growatt growls briefly and then purrs like a kitten. I am actually shocked how lovely the waveform is.

 

[nofear72784]

Your mains input looks horrible. Can you scope it without the inverter on or connected at all

My 3 ton AC compressor just kicked on, the Growatt growls briefly and then purrs like a kitten. I am actually shocked how lovely the waveform is.

Well now this is interesting!! I left the PV on and it's bringing in 3kw.

 

[nofear72784]

Well now this is interesting!! I left the PV on and it's bringing in 3kw.

Lights still flicker, dimmers do not work and waveform is still like overcooked spaghetti but by damn the wave is MUCH improved

 

[nofear72784]

And I'll be damned my grid input is 100x better.

 

[nofear72784]

I really want to open up the inverter. There is something going on between the MPPT and the inverter

 

[nofear72784]

I did find the inverters specifications for those that were asking. I see 10% THDV for non-linear loads and 5% for linear.

 

[Zwy]

And I'll be damned my grid input is 100x better.

Start adding loads on both inverter and grid power. With and without PV.

Definitely not pretty on your first videos. Was this with LED lights on?

 

[nofear72784]

Yes they were on. I am beyond intrigued to see inside this box. How is DC charging voltage from the mppt input which has the battery disconnected when I was testing (so the output was 100% from solar) so much better!? Granted the lights still flickered but that is at least a halfway decent sinusoidal wave. After reading the 10% harmonic distortion reading into a non linear load I don't think that wave would get any cleaner under that load. Now, how to get that waveform under no pv input.

 

[shopman]

A suggestion(maybe you already tried this so my apologies if you did). Disconnect everything from the inverter aside from the battery. No grounds, AC out, Ac in or PV array. Just the inverter on its own. Replace the screw that your removed from the inverter so its back to stock configuration. With just the batteries connected, turn the inverter on and see what the output waveform looks like on your scope.

 

[nofear72784]

I did try that but did not post it, same output as earlier today. Even with the more reasonable wave the lights are still flickering.

 

[Zwy]

With a non neutral switching transfer switch such as the Reliance commonly used, this may affect voltage regulation on output. The configuration you currently have where the input and output neutrals are connected (thru the common pass thru of the transfer switch) has not been confirmed to be supported by the manufacturer of the inverter. MPP LV6548 and other models are not, I'm quite certain the Sungold isn't either.

This may have damaged the unit and what you are seeing is the result. If I were you, I'd definitely not mention how this was wired if you are seeking warranty. It really is the fault of the manufacturer and the poor documentation provided. If they specifically stated in their documentation not to tie output and input neutrals together, then the problem would have been a result of you not reading the documentation. I look at this as they did not provide proper documentation if this is the cause.

 

[SpongeboB Sinewave]

On your post #99, "Output from the inverter with PV and Grid input disconnected:", I see some added voltage on the waveform happening aperiodically. They may actually have been periodic but you would have to slow the sweep down a bunch in order to see how often that would be.
The more recent waveforms look good BUT your lights are still flickering. As long as your lights are OK, then I think you should try video capturing the waveform again so we can see how it varies.

Just showing a few cycles of a nice waveform as a still photo may not show up on the scope. We can get by the shutter speed issue so that's OK.

 

[nofear72784]

With a non neutral switching transfer switch such as the Reliance commonly used, this may affect voltage regulation on output. The configuration you currently have where the input and output neutrals are connected (thru the common pass thru of the transfer switch) has not been confirmed to be supported by the manufacturer of the inverter. MPP LV6548 and other models are not, I'm quite certain the Sungold isn't either.

This may have damaged the unit and what you are seeing is the result. If I were you, I'd definitely not mention how this was wired if you are seeking warranty. It really is the fault of the manufacturer and the poor documentation provided. If they specifically stated in their documentation not to tie output and input neutrals together, then the problem would have been a result of you not reading the documentation. I look at this as they did not provide proper documentation if this is the cause.

I would hope that is not the case as the transfer switch is code compliant in my area with an inverter. Proper documentation is certainly needed.

 

[Zwy]

I would hope that is not the case as the transfer switch is code compliant in my area with an inverter. Proper documentation is certainly needed.

A 3 pole neutral switching would be code compliant and actually preferred over a 2 pole non neutral switching in most cases.

From the various discussions in threads here and correspondence with suppliers, manufacturers and engineers, the consensus was having a common neutral like your example was not supported .

A common neutral in the system wired as you have it creates a neutral ground loop in bypass mode. Current will be shared by the neutral circuit in the inverter and also thru the neutral from the transfer switch to the N-G bond at the disconnect or main panel. Whether this creates another possible ground loop path with the input and output grounds tied together thru the inverter depending on circuit resistance and distance to/from N-G bond, I do not know but is possible. This is why a 3 pole neutral switching transfer switch on output with an EGC only run to inverter input is desired.

I did run across this thread which is quite interesting. https://diysolarforum.com/threads/s...sing-taiwan-based-all-in-one-inverters.41802/

The poster believed the 48V AC charger created the problem as it might be PWM based. This might actually be the case if charging is set to be split between solar and AC input. It could also be a result of using a common neutral with grid bypass wired in. Just thought I'd throw that in considering you saw a cleaner waveform with PV connected.

 

[nofear72784]

What's odd is that we have this exact same configuration (grid connected reliance transfer switch with pass though neutral) running in our cabin and have 0 issues. That is a small 3kw Growatt though that only runs at about 5% capacity max. It was already installed when we purchased it, I never took a close look at the wiring as it works just fine. I have no problem changing the wiring if that is what is required but without documentation stating so I am reluctant to kick another hornets nest with this Building Inspector.
I'm sure we will get to the bottom of this issue it's just a matter of time and patience. Does anyone know if SunGold is part of Signature Solar now? All my emails are responded to at night as if China based but last night I was given a Texas number to call "if urgent".

 

[nofear72784]

A 3 pole neutral switching would be code compliant and actually preferred over a 2 pole non neutral switching in most cases.

From the various discussions in threads here and correspondence with suppliers, manufacturers and engineers, the consensus was having a common neutral like your example was not supported .

A common neutral in the system wired as you have it creates a neutral ground loop in bypass mode. Current will be shared by the neutral circuit in the inverter and also thru the neutral from the transfer switch to the N-G bond at the disconnect or main panel. Whether this creates another possible ground loop path with the input and output grounds tied together thru the inverter depending on circuit resistance and distance to/from N-G bond, I do not know but is possible. This is why a 3 pole neutral switching transfer switch on output with an EGC only run to inverter input is desired.

I did run across this thread which is quite interesting. https://diysolarforum.com/threads/s...sing-taiwan-based-all-in-one-inverters.41802/

The poster believed the 48V AC charger created the problem as it might be PWM based. This might actually be the case if charging is set to be split between solar and AC input. It could also be a result of using a common neutral with grid bypass wired in. Just thought I'd throw that in considering you saw a cleaner waveform with PV connected.

Unfortunately I have it set to solar charging only. But, funny you brought that up. I did notice last night that even when set to solar only the BMS was reporting more amps than the panels were producing. I threw my amp clamp on the input and saw 2 amps coming in on the mains connection. Now is it using that to supply power to the main functions, is it leaking into the battery through a bad circuit? Many questions. But, even with mains completely disconnected it still showed a poor waveform.
TBH I am fairly exhausted with poking and prodding this box. I don't have a problem performing a few tests to help the manufacturer figure out a problem but with 0 feedback for a week I feel like this is an exercise in futility. My wife has about had it with the time I am spending on this. Hoping for a resolution shortly.

 

[nofear72784]

On your post #99, "Output from the inverter with PV and Grid input disconnected:", I see some added voltage on the waveform happening aperiodically. They may actually have been periodic but you would have to slow the sweep down a bunch in order to see how often that would be.
The more recent waveforms look good BUT your lights are still flickering. As long as your lights are OK, then I think you should try video capturing the waveform again so we can see how it varies.

Just showing a few cycles of a nice waveform as a still photo may not show up on the scope. We can get by the shutter speed issue so that's OK.

Here is the inverter with PV as the only connection. It's a much better waveform but still quite dirty.

 

[nofear72784]

I saw Will's review of the new EG4. He noticed a circuit breaker on the input and some changes in the inverter caps and mosfets as well as a different firmware. This SunGold unit also has the breaker on the input and the firmware is different than the LV and the EG as it shows revision 69. It seems like this unit is different than the LV6548 and different than the EG4 as the PV input max voltage is the same as the LV6548. This can be an issue if anyone tries to mix and match these seemingly similar units. Food for thought for anyone looking at these 3 units.

 

[Zwy]

What's odd is that we have this exact same configuration (grid connected reliance transfer switch with pass though neutral) running in our cabin and have 0 issues. That is a small 3kw Growatt though that only runs at about 5% capacity max.

The GW 3Kw may or may not have the ability to have N-G bond depending on when it was manufactured as there were some changes but model number was not changed.

It's another case of poor documentation and proper labeling. Yes, I have one in my truck camper. Never ran a 120V LED off it yet though.

It was already installed when we purchased it, I never took a close look at the wiring as it works just fine. I have no problem changing the wiring if that is what is required but without documentation stating so I am reluctant to kick another hornets nest with this Building Inspector.
I'm sure we will get to the bottom of this issue it's just a matter of time and patience. Does anyone know if SunGold is part of Signature Solar now? All my emails are responded to at night as if China based but last night I was given a Texas number to call "if urgent".

 

[Zwy]

Unfortunately I have it set to solar charging only. But, funny you brought that up. I did notice last night that even when set to solar only the BMS was reporting more amps than the panels were producing. I threw my amp clamp on the input and saw 2 amps coming in on the mains connection. Now is it using that to supply power to the main functions, is it leaking into the battery through a bad circuit? Many questions. But, even with mains completely disconnected it still showed a poor waveform.

In that linked thread, he saw the same thing. It seems models or systems wired with a common neutral are prone to this. The LV5048 would be a common neutral as it was the predecessor to the LV6048.

TBH I am fairly exhausted with poking and prodding this box. I don't have a problem performing a few tests to help the manufacturer figure out a problem but with 0 feedback for a week I feel like this is an exercise in futility. My wife has about had it with the time I am spending on this. Hoping for a resolution shortly.

Have you ever tried just running a string of LED's on the inverter output without any AC input wired without the transfer switch?

That is what I would try, disconnect AC input and output to the transfer switch, wire up a circuit with some LED's to the inverter and see if the LED's still flicker. If the lights flicker, then it is a problem with the inverter. If the lights don't, then the problem is related to how this is wired.

 

[nofear72784]

I could. I did put the scope on the output with nothing connected but DC battery in and the wave was shifting in the same manner so I don't think it will change anything. I have some edison bases but I am past the point of wanting to tinker any longer. They need to send me some proper documentation instead of wasting all of our times doing R&D for them. It has been 1 week, not a single useful response from SunGold short of "the engineering has to build a mock setup to test you will hear back in 2 day max" then "we are in training". I genuinely appreciate all the input, lot's of good information was shared.
If I hear anything back from them I will be sure to update this thread. If I do not and get stuck with this unit I will update this thread to make others aware all the same.