SunGold Power SP6548 Flickering lights after new install

[Jazzmonger]

I had exactly the same thing happen with my LV6048 inverter. All of my LED shop lights flickered periodically, and for no apparent reason. Then I installed 2 LV6548s and the problem went away. I isolated and captured the problem on an oscilloscope trace - it's periodic noise on the output of the inverter sine wave that is the culprit.

normal trace of 120v inverter output, no flicker:


Periodic noise (no cause that I could find) induced by the inverter at periodic intervals:



THIS is what causes the flicker. Get rid of the noise on the 120v line and the flicker will go away as well.

As a kid I was the renowned Bay Area expert in eliminating engine noise in massive car stereo systems in the 1980's. Looking at the traces on my scope, this is literally no different. It's line noise. Before getting the LV6548s I ordered a 20amp line filter that should eliminate this problem. They make a 50amp version as well but I was going to try it first on the light circuit just to be sure it solved the problem, then filter the entire 50a output. Note this would require 2 50a filters, one for each 120v leg. Or just use the 20a filter on the light circuits.

this is what I ordered:
line filter


Now, I have it, but never tested it bc I ended up installing the LV6548s for my shop and house and they don't suffer from this issue, but based on the scope traces it should fix the flickering problem. For $16, it certainly might be worth a try?

anyone else tried this???

Jeff


.

 

[nofear72784]

It is certainly worth a shot. Getting 4ga wire on that would be interesting lol. The issue is this is not periodic noise or induced by other items on the circuit, it is constant. We should not have to noise filter 6.5kw of output to use an inverter with lights or a UPS.
I have used those filters in the past, they work well just be sure to ground them. Other big issue is they are not code compliant in my area.
Should have a decent idea today if my "fix" actually works. Fingers crossed.

Something like this would be easier to implement if you were filtering just a lighting circuit but deff not going to pass an inspection with it installed. inline filter 6A

 

[efficientPV]

You need a digital scope that will display at least 5 seconds of waveform. Then at the location of an anomaly it can be expanded to see what is happening. Flickering will be caused the AC waveform being modulated (varying amplitude). A little waveform distortion will not cause the flickering on a non dimmed LED.

 

[Quattrohead]

General waveform distortion no, but those spikes could cause a power supply transistor to conduct more and cause a spike in the regulated output.
I'm thinking supply chain and labour issues are causing corners to be cut in order to deliver goods.

 

[nofear72784]

General waveform distortion no, but those spikes could cause a power supply transistor to conduct more and cause a spike in the regulated output.
I'm thinking supply chain and labour issues are causing corners to be cut in order to deliver goods.

I tried using an Xstar V4 18650 charger off the inverter a few days back but the transformer was letting out a whine and the display was flickering. Switched it back to bypass, no issue. Just going to leave it there until we get an answer from Sunpower. Most other ac/dc chargers I tried worked just fine though. Charged a few Ego batteries off it no problem. Feels like a waste having the 48v pack charged and pv available but not using it. At least the lawnmower and trimmer will be running off sun juice.

 

[Jazzmonger]

The issue is this is not periodic noise or induced by other items on the circuit,

Periodic=repeats= repeats constantly. I.e. sometimes multiple times a second.

you can see the noise in the captured waveform which contains the noise and it coincides exactly with flickering. I expanded the captured waveforms for clarity.

 

[nofear72784]

Periodic=repeats= repeats constantly. I.e. sometimes multiple times a second.

you can see the noise in the captured waveform which contains the noise and it coincides exactly with flickering. I expanded the captured waveforms for clarity.

Smacks forehead. Should have had more coffee this morning.

 

[Jazzmonger]

You need a digital scope that will display at least 5 seconds of waveform. Then at the location of an anomaly it can be expanded to see what is happening. Flickering will be caused the AC waveform being modulated (varying amplitude). A little waveform distortion will not cause the flickering on a non dimmed LED.

Well, when you capture a waveform that causes led flicker, please post it for all to see. My capture does not show amplitude increase, yet when the noise on the waveform appears, the led lights all flicker. So, either the leds are flickering and inducing the noise (not likely) or the inverter's waveform is distorting and causing the flicker. I have about a dozen 4' lights in my shop, 4 different brands/wattage and they all flicker at the same time.

My conclusion that it's the inverter and noise on the waveform causing the led lights to flicker is simple. I get flicker when using the Lv6048 and capture noise on the waveform and I don't when using lv6548's with a clean waveform on the exact same led lights. I also isolated the light circuit so the led lights were the only load.

Of course, Your mileage will vary.... and until someone actually tries using a noise filter like I bought to see if it works, the jury remains out. Perhaps I'll do that over the winter when there's nothing else to do.

 

[nofear72784]

Hard to see due to the phone camera shutter speed but it is there

It's easier to see when I did not have the correct range scope. Dances around like a wet noodle.

 

[Jazzmonger]

We should not have to noise filter 6.5kw of output to use an inverter with lights or a UPS.

I totally agree, but we should not be outsourcing so much of this stuff to China....but we do and we get questionable products and components as a result. So, we get cheap, sometimes malfunctioning products billed as quality. It's the price we currently pay.

 

[nofear72784]

I totally agree, but we should not be outsourcing so much of this stuff to China....but we do and we get questionable products and components as a result. So, we get cheap, sometimes malfunctioning products billed as quality. It's the price we currently pay.

Some serious truth in that. Lost my last job to outsourcing and downsizing.

 

[Jazzmonger]

Hard to see due to the phone camera shutter speed but it is there

Hard to see due to the phone camera shutter speed but it is there

It's easier to see when I did not have the correct range scope. Dances around like a wet noodle.

What kind of inverter is that? It is NOT producing a pure sine wave.

 

[nofear72784]

What kind of inverter is that? It is NOT producing a pure sine wave.

The flat spot is clipping due to the lower range of that scope. That is why I found a scope with a larger range. You can see the shifting in the wave better on the scope with the blue backlight.

 

[Jazzmonger]

Aha. Now I understand! So you are confirming my trace?

 

[goldserve]

Not sure how I would characterize that scope because the settings are all whacked. You need a proper trigger point (ideally the zero crossing) in order to characterize the jitter. The frequency you are reading with the scope is all over the place and doesn’t seem normal.

In the really old school scope, the waveform isn’t perfectly smooth but that’s the same for simulated sine wave UPS. Can you increase the persistence to get a picture of the long trend of the sine wave.

 

[nofear72784]

Aha. Now I understand! So you are confirming my trace?

yes, seems similar to what you saw

 

[nofear72784]

Not sure how I would characterize that scope because the settings are all whacked. You need a proper trigger point (ideally the zero crossing) in order to characterize the jitter. The frequency you are reading with the scope is all over the place and doesn’t seem normal.

It is 100% not normal. Leds flicker and it tripps all the UPS's in the house, 4 in total. 1 basic unit running the ONT, 1 TripLite running network switches, 1 Cyberpower on a server and another Cyberpower on the main desktop PC. We have a Panamax on some equipment and that takes care of the issues and never shows a fault. Remove the Panamax and the TV is not happy and the amplifiers have a significant noise floor that is not there otherwise.

 

[goldserve]

It is 100% not normal. Leds flicker and it tripps all the UPS's in the house, 4 in total.

I’m not doubting something is not normal but from what I see in those YouTube videos, it’s not showing the real problem because there are some basic instrumentation issues. If the frequency jumps around that much from cycle to cycle, that’s a lot of distortion. The problem wouldn’t be from bad sine wave or amplitude excursion, etc.

 

[nofear72784]

Therein lies the million dollar question... 3 people here that have the unit have this issue along with an owner of an MPP LV6548 with YCCfans. Only commonality is the DC fans. All the review MPP units, EG4's etc all have different fans and no one is seeing these issues.

 

[goldserve]

With proper instrumentation, you could try to disconnect the fan and see if you see any difference in your lights and the scope shots. It would have to be a VERY poor design for some 12v dc noise to make it all the way back through the stages of regulation and ultimately the boost converter to sine wave output.