Good luck finding the information you are looking for. I have two 1KW Sun inverters. One failed on the AC side during a thunderstorm from a distant strike, I heard the breaker pop when the inverter shorted. I have good surge suppressors on both the DC and AC sides. The Sun unit that blew was on one side of a split phase system. The one on the other side survived. Nothing else was damaged. Opened it up and found two AC side IGBT's blown, vaporized traces, popped gate driver, and a few other components fried. Replaced em all and it no longer is shorted on the AC side, shows the correct line voltage, but gives an error message, which I don't recall. I suspect that when the gate driver went, line voltage swept through it and took out other stuff, including the output pin of micro driving it. High power circuits should be electrically isolated from the low power controlling side via opto isolators transformers, etc. I didn't see any of that in my unit. Not having a schematic for the unit, it's not worth my time to reverse engineer it to try to fix it, nor is it worth the money to replace it since it's an iffy design from what I can tell.
Yes, you are absolutely correct. AC shorts only rarely stay contained to the high voltage side. Damage will typically go into the low voltage electronics and fry critical components which make a repair mostly impossible. I now do have 4 boards resting at the SUN-GTIL Cemetery... ;)
The IGBTs actually shunt the DC power through the primary winding of the transformers. So they are actually on the DC side of the transformers. The AC (secondary) side of the transformers is connected in series via diodes to produce the 240VAC output power. The AC power is also connected to the inverter via a low pass filter comprised of a varistor, chokes and capacitors. This is standard practise and protects the device from transient spikes in the AC line. There is another circuit in grid tie inverters that is very sensitive to AC power disruptions and it is called the zero-cross detector. This circuit produces a pulse when the AC voltage sine wave crosses over the zero volt level. The MPU uses this signal to get the AC waveform produced by the inverter to match the AC waveform on the grid. So, if the zero-cross circuit produces a false zero-cross pulse and the firmware accepts it as a zero-cross signal then you get a big bang and the IGBTs will be the first part to blow. The way to stop this is to design the firmware so that it only looks for the ZC signal when it expects it to happen, which is exactly every 10 or 20 ms. My biggest fear is that the coders at Sun many not have done this which would make the device much more sensitive to any transient spikes and disruptions on the AC power line.
I stumbled onto this video looking for something else regarding the Sun1000GTIL. However, what caught my attention was the part about the differences in boards. I too recently had an inverter go up in smoke and ordered a replacement board from Aliexpress. The new board has those two relays at the AC side like you were showing, but the old board did not. After installing the new board, the inverter did work (sort of). I had to go into the settings screen and re-save the settings, because the main board was defaulting to some factory defaults despite what the display screen was showing. This tells me that the display board is somewhat independent of the inverter operation unless you make changes and/or re-save them. After I re-saved them, it is has been working fine. I also noticed that on your old board, there did not appear to be a place for that 3 pin ribbon cable that comes from that relay board, where my old one did have the header, just did not have the board installed. I do think you might a version issue of some sort, but I wonder if you just got a bad board. I assume you have tried re-saving the settings from your old display unit.
@@markl8921 hi. yes. I have played around with factory settings but nothing helped. I have as well asked if maybe there is a new display required too, and also considered if the mainboard had a fault and tried a different Control board. Every part i got had a "QC Pass" sticker, I discussed the inverter behavior with Jesudom, nobody had a clue why it did just what it did. I have no idea...
@@RolandW_DIYEnergyandMore Then I wonder if you got a bad board or the the wrong version (i.e. input/output voltage). I did have to select the correct one when I ordered mine. Good luck, hope you figure it out.
Roland, are you considering to replace the components on the older board with the replacement values that match the new board? Perhaps the older board can come back to life, with upgraded components (MOSFETs, Resistors).
unfortunately those modification required a completely redesigned circuit board. The new components don't fit into the old foodprints of old components as they are mostly larger. I am hoping that I can make a sort of running Frankenstein board from an old SUN2000 where I find good components from all the other older boards :))
My oldest SUN2000 is as well already 3-4 years old and still running. I think the older units seem to be better. Maybe they have used better components at that time. all of my younger units have failed within 1-2 years. strange...
The firmware would have had to be changed to operate the anti-islanding relay correctly. The relay should activate (close) a few seconds after the AC power comes on. I suggest checking the operation of the AC limit sensor. That sensor is trying to keep the export power at zero. The display should show you the AC grid voltage and the AC grid power. What happens when you clip the sensor onto the active line of a heater and then reverse the sensor? The AC grid power reading should show the wattage of the heater and it should go from positive to negative (or vise versa) when the sensor direction is changed. The old board probably only needs a replacement mosfet worth $2. I would fix that before buying a new inverter. Have you checked the mosfets with a multimeter?
On the replacement board the CT is only showing a random value which is somewhere 1/3 to 1/5th of the real value. So there is either a fault on the CT circuit or it is handled differently on this board, so it required a different firmware for it to show correctly too. The best old board that I have doesn't just need new IGBTs, but also the driving circuit and as well the internal 12V powersupply isn't working. I have ordered some parts to try and fix it. For this board I haven't given up yet...
@@RolandW_DIYEnergyandMore Could it be that the new board is meant for a different CT sensor? The display is showing 1/3 the real grid power. So, maybe if you put the ac wire 3 times through the CT sensor it will give the correct value? I suggest trying this on a heater with known watts and see what the display reads. I think I know why these Sun grid tie inverters are prone to failures and I do not believe the IGBTs are being directly blown by spikes on the AC line. The IGBTs are being blown by timing mishaps. The design fault is the lack of adequate filters between the noizy AC and DC circuits and the VCC supply. So, the spikes (and noise) go through to the VCC rail and this resets the mpu sometimes. This can leave the IGBTs turned on when they should be off. The current surge caused by the mismatched AC timing blows out the IGBTs no matter how big or strong they are. To fix the design fault, disconnect the 12V supply (to the MPU circuit) from the AC circuit. Then use a mini 12V power supply module from AliExpress to supply the 12V. Connect the power supply module's 240VAC input to the Inverter AC input after the low pass filter. I am not sure how VCC is connected to the inverter DC supply? But it would be better if that is supplied by a small buck converter with adequate voltage specs. This will provide good noise immunity to the shunting occuring in the DC circuit. The buck converter can be adjusted to match the 12V power supply output. This allows the MPU to remain active when the grid goes down. I am convinced that if you do this, you will never get another fault and the SUN grid tie inverter will go from being the worst inverter ever to the very best inverter ever. (They should give me a royalty payment for revealing this.)🙂
@cyber5515 thanks, you might be onto something. when I had a repair attempt on the old board last week, the new IGBTs where blown instantly when the MCU connected the AC to the grid. Later I found out that the internal PSU only provided 4V on the 13V output which later provides the 3,3V for the MCU via it's linear voltage regulator. This false voltage might have brought the MCU into an undefined state and driven both gates to HIGH. I am waiting for a new PSU regulator chip but already thought about supplying the 13V from an external power source. Let's see where this goes :)
Not feeling so bad about my Soyo with melted fuse now... Too bad the smaller manufacturers take too long to up their quality; they have to build to price i guess 😞
@@geekmystique hi. I was considering a Soyo in case I lost 2 inverters and would only be lest with one operational SUN. The low output is sort of a bummer 😔
As far I could see those type of inverters are the "All in One" type inverters which are Off-grid inverters with Grid bypass. I can only use Grid-tied, AC coupled inverters which are just pushing their power directly onto the Grid sinewave and using zero-export adjustment...
@@RolandW_DIYEnergyandMore Sorry Roland. I am on offgrid. As I see you have so many batteries, I thought you are on offgrid too. Anyway, I am using 2 sets of 24V 3kW small offgrid systems (1.6kWp and 0.8kWp panels). Beside setting and some minor issues, the systems are running fine for the past 2 years. I am getting about 250 kWh a month. Small investment small return. Currently using 2 sets of Y&H hybrid inverters. I also have a set of Anern and PowMr as spares. These 2 were bought with big discount.
@mtay65 yeah. I can have sometimes 12-15kW going over the phase wire at a time in our resort. Could never pass so much current over a bypass from a 4kW off grid All in One without burning it to ashes :) That's why in my system almost all is AC coupled.
@@RolandW_DIYEnergyandMore I see. However my setup is different from yours. I don't feed my gain into the main but directly into 2 units of air conditioners. I don't meddle with the grid.
Good luck finding the information you are looking for. I have two 1KW Sun inverters. One failed on the AC side during a thunderstorm from a distant strike, I heard the breaker pop when the inverter shorted. I have good surge suppressors on both the DC and AC sides. The Sun unit that blew was on one side of a split phase system. The one on the other side survived. Nothing else was damaged. Opened it up and found two AC side IGBT's blown, vaporized traces, popped gate driver, and a few other components fried. Replaced em all and it no longer is shorted on the AC side, shows the correct line voltage, but gives an error message, which I don't recall. I suspect that when the gate driver went, line voltage swept through it and took out other stuff, including the output pin of micro driving it. High power circuits should be electrically isolated from the low power controlling side via opto isolators transformers, etc. I didn't see any of that in my unit. Not having a schematic for the unit, it's not worth my time to reverse engineer it to try to fix it, nor is it worth the money to replace it since it's an iffy design from what I can tell.
Yes, you are absolutely correct. AC shorts only rarely stay contained to the high voltage side. Damage will typically go into the low voltage electronics and fry critical components which make a repair mostly impossible. I now do have 4 boards resting at the SUN-GTIL Cemetery... ;)
The IGBTs actually shunt the DC power through the primary winding of the transformers. So they are actually on the DC side of the transformers. The AC (secondary) side of the transformers is connected in series via diodes to produce the 240VAC output power. The AC power is also connected to the inverter via a low pass filter comprised of a varistor, chokes and capacitors. This is standard practise and protects the device from transient spikes in the AC line.
There is another circuit in grid tie inverters that is very sensitive to AC power disruptions and it is called the zero-cross detector. This circuit produces a pulse when the AC voltage sine wave crosses over the zero volt level. The MPU uses this signal to get the AC waveform produced by the inverter to match the AC waveform on the grid. So, if the zero-cross circuit produces a false zero-cross pulse and the firmware accepts it as a zero-cross signal then you get a big bang and the IGBTs will be the first part to blow.
The way to stop this is to design the firmware so that it only looks for the ZC signal when it expects it to happen, which is exactly every 10 or 20 ms. My biggest fear is that the coders at Sun many not have done this which would make the device much more sensitive to any transient spikes and disruptions on the AC power line.
I stumbled onto this video looking for something else regarding the Sun1000GTIL. However, what caught my attention was the part about the differences in boards. I too recently had an inverter go up in smoke and ordered a replacement board from Aliexpress. The new board has those two relays at the AC side like you were showing, but the old board did not. After installing the new board, the inverter did work (sort of). I had to go into the settings screen and re-save the settings, because the main board was defaulting to some factory defaults despite what the display screen was showing. This tells me that the display board is somewhat independent of the inverter operation unless you make changes and/or re-save them. After I re-saved them, it is has been working fine.
I also noticed that on your old board, there did not appear to be a place for that 3 pin ribbon cable that comes from that relay board, where my old one did have the header, just did not have the board installed. I do think you might a version issue of some sort, but I wonder if you just got a bad board. I assume you have tried re-saving the settings from your old display unit.
@@markl8921 hi. yes. I have played around with factory settings but nothing helped. I have as well asked if maybe there is a new display required too, and also considered if the mainboard had a fault and tried a different Control board. Every part i got had a "QC Pass" sticker, I discussed the inverter behavior with Jesudom, nobody had a clue why it did just what it did. I have no idea...
@@RolandW_DIYEnergyandMore Then I wonder if you got a bad board or the the wrong version (i.e. input/output voltage). I did have to select the correct one when I ordered mine. Good luck, hope you figure it out.
Roland, are you considering to replace the components on the older board with the replacement values that match the new board? Perhaps the older board can come back to life, with upgraded components (MOSFETs, Resistors).
unfortunately those modification required a completely redesigned circuit board. The new components don't fit into the old foodprints of old components as they are mostly larger.
I am hoping that I can make a sort of running Frankenstein board from an old SUN2000 where I find good components from all the other older boards :))
Forget what I said. Rest of video clears this up.
Said it before roland, i see a deye in the future ;)
My 1000's with the lower voltage are still running now (x fingers) and it has been 3 years they just need to be rebooted every once in a while.
My oldest SUN2000 is as well already 3-4 years old and still running. I think the older units seem to be better. Maybe they have used better components at that time. all of my younger units have failed within 1-2 years. strange...
Please try deye inverters their warranty response is good
Yes, there will be the time when I have to change, but that will be a complete system replacement including the battery bank...
Hallo Roland,
Schon mal an Victron MP2 gedacht.
Sind mittlerweile recht günstig und können
auch laden…
Grus aus Deutschland
Uwe
@@uweheckmann5111 hallo Uwe. Der Victron ist AC gekoppelt und hat Null-Export? Ist aber wahrscheinlich Einphasig, oder? Schau ich mir mal an, danke!
The firmware would have had to be changed to operate the anti-islanding relay correctly. The relay should activate (close) a few seconds after the AC power comes on. I suggest checking the operation of the AC limit sensor. That sensor is trying to keep the export power at zero. The display should show you the AC grid voltage and the AC grid power. What happens when you clip the sensor onto the active line of a heater and then reverse the sensor? The AC grid power reading should show the wattage of the heater and it should go from positive to negative (or vise versa) when the sensor direction is changed. The old board probably only needs a replacement mosfet worth $2. I would fix that before buying a new inverter. Have you checked the mosfets with a multimeter?
On the replacement board the CT is only showing a random value which is somewhere 1/3 to 1/5th of the real value. So there is either a fault on the CT circuit or it is handled differently on this board, so it required a different firmware for it to show correctly too.
The best old board that I have doesn't just need new IGBTs, but also the driving circuit and as well the internal 12V powersupply isn't working. I have ordered some parts to try and fix it. For this board I haven't given up yet...
@@RolandW_DIYEnergyandMore Could it be that the new board is meant for a different CT sensor? The display is showing 1/3 the real grid power. So, maybe if you put the ac wire 3 times through the CT sensor it will give the correct value? I suggest trying this on a heater with known watts and see what the display reads.
I think I know why these Sun grid tie inverters are prone to failures and I do not believe the IGBTs are being directly blown by spikes on the AC line. The IGBTs are being blown by timing mishaps. The design fault is the lack of adequate filters between the noizy AC and DC circuits and the VCC supply. So, the spikes (and noise) go through to the VCC rail and this resets the mpu sometimes. This can leave the IGBTs turned on when they should be off. The current surge caused by the mismatched AC timing blows out the IGBTs no matter how big or strong they are.
To fix the design fault, disconnect the 12V supply (to the MPU circuit) from the AC circuit. Then use a mini 12V power supply module from AliExpress to supply the 12V. Connect the power supply module's 240VAC input to the Inverter AC input after the low pass filter. I am not sure how VCC is connected to the inverter DC supply? But it would be better if that is supplied by a small buck converter with adequate voltage specs. This will provide good noise immunity to the shunting occuring in the DC circuit. The buck converter can be adjusted to match the 12V power supply output. This allows the MPU to remain active when the grid goes down.
I am convinced that if you do this, you will never get another fault and the SUN grid tie inverter will go from being the worst inverter ever to the very best inverter ever. (They should give me a royalty payment for revealing this.)🙂
@cyber5515 thanks, you might be onto something. when I had a repair attempt on the old board last week, the new IGBTs where blown instantly when the MCU connected the AC to the grid. Later I found out that the internal PSU only provided 4V on the 13V output which later provides the 3,3V for the MCU via it's linear voltage regulator. This false voltage might have brought the MCU into an undefined state and driven both gates to HIGH. I am waiting for a new PSU regulator chip but already thought about supplying the 13V from an external power source. Let's see where this goes :)
I think going with anything that isnt a loq frequency inverter is a mistake
Not feeling so bad about my Soyo with melted fuse now... Too bad the smaller manufacturers take too long to up their quality; they have to build to price i guess 😞
i guess Soyo is not an option for you? Quality might be marginally better but they only go up to 800-900 watt from batteries..
@@geekmystique hi. I was considering a Soyo in case I lost 2 inverters and would only be lest with one operational SUN. The low output is sort of a bummer 😔
48V 4.2kW hybrid inverters are so cheap nowadays....buy 2, 1 for the spare.
@@mtay65 hi. Do you have a particular model in mind? It must be AC coupled with export limiter... thanks!
As far I could see those type of inverters are the "All in One" type inverters which are Off-grid inverters with Grid bypass. I can only use Grid-tied, AC coupled inverters which are just pushing their power directly onto the Grid sinewave and using zero-export adjustment...
@@RolandW_DIYEnergyandMore Sorry Roland. I am on offgrid. As I see you have so many batteries, I thought you are on offgrid too. Anyway, I am using 2 sets of 24V 3kW small offgrid systems (1.6kWp and 0.8kWp panels). Beside setting and some minor issues, the systems are running fine for the past 2 years. I am getting about 250 kWh a month. Small investment small return. Currently using 2 sets of Y&H hybrid inverters. I also have a set of Anern and PowMr as spares. These 2 were bought with big discount.
@mtay65 yeah. I can have sometimes 12-15kW going over the phase wire at a time in our resort. Could never pass so much current over a bypass from a 4kW off grid All in One without burning it to ashes :)
That's why in my system almost all is AC coupled.
@@RolandW_DIYEnergyandMore I see. However my setup is different from yours. I don't feed my gain into the main but directly into 2 units of air conditioners. I don't meddle with the grid.