Cheapest Inverter Premium Parts = Surprising Results!
video description
Date: 2025-06-18
Related videos
Comments and reviews: 20
colinsmith6340
The first bank of transistors you saw, the one split between 3 heatsinks are the H bridge driving the step up foward converter. They are the ones that handle all the current and will actually get the hottest. The IGBTs are also a H bridge, but since they only pass up to 10 amps, their I2R losses are much lower. The first bank is the main high frequency voltage step up to the 400 odd volts that the IGBT bridge then chops up for the AC waveform.
Basically, these things contain 2 class D audio amplifiers. They run at a fixed frequency of course. The first bank chops up the 50-60 volts into 15-20odd khz DC that goes through the high frequency transformer. The high voltage on the other side gets rectified (you will find a set of 4 or multiple of 4 high speed diodes) to charge the main res caps. Some higher end inverters might use synchronous rectifiers, but its not really that necessary at the lower currents of the high side system.
That high voltage rail is then used to feed the second H bridge which uses high frequency sine wave synthesis to turn into a sine wave. Thats what the gigantic inductor and capacitor bank are for, to filter out the high frequency noise and spit out the low frequency AC waveform.
Think of it as a class D subwoofer amplifier that only ever creates 50(60) hz.
reply
The first bank of transistors you saw, the one split between 3 heatsinks are the H bridge driving the step up foward converter. They are the ones that handle all the current and will actually get the hottest. The IGBTs are also a H bridge, but since they only pass up to 10 amps, their I2R losses are much lower. The first bank is the main high frequency voltage step up to the 400 odd volts that the IGBT bridge then chops up for the AC waveform.
Basically, these things contain 2 class D audio amplifiers. They run at a fixed frequency of course. The first bank chops up the 50-60 volts into 15-20odd khz DC that goes through the high frequency transformer. The high voltage on the other side gets rectified (you will find a set of 4 or multiple of 4 high speed diodes) to charge the main res caps. Some higher end inverters might use synchronous rectifiers, but its not really that necessary at the lower currents of the high side system.
That high voltage rail is then used to feed the second H bridge which uses high frequency sine wave synthesis to turn into a sine wave. Thats what the gigantic inductor and capacitor bank are for, to filter out the high frequency noise and spit out the low frequency AC waveform.
Think of it as a class D subwoofer amplifier that only ever creates 50(60) hz.
reply
greatscott
Great mod but those little things are pricey. Not many would see a return on investment swapping them out or even buying a unit with them. But prices on them should come down over time. I would recommend on solar if you are reliant on it I would have a spare inverter regardless of brand. When it goes down you are down till you get a replacement. I really like the minisplits coming out that have direct solar input but they also act as a hybrid inverter, as they can divert power to a battery or the grid etc. HVAC is the biggest power uses so it makes sense to do it that way.
My only issue with being offgrid is that powerfactor is now a big thing you need to consider and nothing is going to make a perfect sine wave without a rotating magnetic field. I would love to see the world crank up the hertz but that will never happen. Fortunatly our US grid is slightly more efficient at 60 vs most of the worlds 50hz. Going to aircraft 400hz will never happen.
reply
Great mod but those little things are pricey. Not many would see a return on investment swapping them out or even buying a unit with them. But prices on them should come down over time. I would recommend on solar if you are reliant on it I would have a spare inverter regardless of brand. When it goes down you are down till you get a replacement. I really like the minisplits coming out that have direct solar input but they also act as a hybrid inverter, as they can divert power to a battery or the grid etc. HVAC is the biggest power uses so it makes sense to do it that way.
My only issue with being offgrid is that powerfactor is now a big thing you need to consider and nothing is going to make a perfect sine wave without a rotating magnetic field. I would love to see the world crank up the hertz but that will never happen. Fortunatly our US grid is slightly more efficient at 60 vs most of the worlds 50hz. Going to aircraft 400hz will never happen.
reply
daskasspatzle2396
Well, you made the inverter somewhat more efficient For me the show stopper with this kind of chinese inverters was the idel power draw. Having four of them in a three phase system, it adds up. The ones i had should have consumed 50 Watts max according to specification, which is quite high already. But measured around and even over 80 Watts, so over 300 for all four of them times 24 hours per day would have made my solar system pointless. Disregarding, that one of the four inverters had a broken mainboard, i think the cheap inverter will in the long run still not make you happy, if for example a electrolytic capacitor fails, or even the fans, which they seem to use a lot.
Now i have four Multiplus II, which should have 18 Watts idle consumption, measured 21 Watts. Might be just in specification with some measurement accuracy taken into account.
Maybe in some years we have good AND affordable inverters. SIC will definitely help in this regard.
reply
Well, you made the inverter somewhat more efficient For me the show stopper with this kind of chinese inverters was the idel power draw. Having four of them in a three phase system, it adds up. The ones i had should have consumed 50 Watts max according to specification, which is quite high already. But measured around and even over 80 Watts, so over 300 for all four of them times 24 hours per day would have made my solar system pointless. Disregarding, that one of the four inverters had a broken mainboard, i think the cheap inverter will in the long run still not make you happy, if for example a electrolytic capacitor fails, or even the fans, which they seem to use a lot.
Now i have four Multiplus II, which should have 18 Watts idle consumption, measured 21 Watts. Might be just in specification with some measurement accuracy taken into account.
Maybe in some years we have good AND affordable inverters. SIC will definitely help in this regard.
reply
ShiraanVII
Hey!
Got 2 chinese units (well, almost everything is made in china, excellent et mediocre stuff, and one failed, during idle time, not heavy use (400 watts) and another one, rated for 4500W failed while making 1200 watts (was running 100% solar, batteries charging, and both of when, even with proper connection, were NOT GROUNDED; danger!
I went with CyberPower (called Nitram too, just a re-branding, 6000 watts in grid mode (solar batteries utility) and in off grid mode: 5000 watts (solar batteries, or emergency mode 3000watts: no solar, no grid, batteries only.
Pretty happy, living at Reunion Island, during the day, normal use, I run 100% on solarbattery if cloudy.
Power outtage during the day is no problem, but during night, I got only 1KW of battery use (won't go under 70%, gel batteries, can't discharge too much. Next purchase: lithium, 4 KW, to begin with.
We'll see!
reply
Hey!
Got 2 chinese units (well, almost everything is made in china, excellent et mediocre stuff, and one failed, during idle time, not heavy use (400 watts) and another one, rated for 4500W failed while making 1200 watts (was running 100% solar, batteries charging, and both of when, even with proper connection, were NOT GROUNDED; danger!
I went with CyberPower (called Nitram too, just a re-branding, 6000 watts in grid mode (solar batteries utility) and in off grid mode: 5000 watts (solar batteries, or emergency mode 3000watts: no solar, no grid, batteries only.
Pretty happy, living at Reunion Island, during the day, normal use, I run 100% on solarbattery if cloudy.
Power outtage during the day is no problem, but during night, I got only 1KW of battery use (won't go under 70%, gel batteries, can't discharge too much. Next purchase: lithium, 4 KW, to begin with.
We'll see!
reply
CATA20034
First of all, one is transform-less inverter, and it will have usually above 98% eff (boost inverter, the other one has an insulation transformer, to boost the 48 to 400V and is cascaded by the inverter and by this you will have a max of 94-96% eff. Regarding the replacement of IGBT with SIC, you swapped 10 pieces, usually are only 4 or 6 in the inverter stage, so don't know what would be the purpose of the others replaced. The on-off time measured over the SIC, is quite bad. You should get lower than 60s with kelvin connected SIC. So you have bad results due to 3 leg SIC, 4 or 7 leads are the optimized for fast SW. You may improve your results with low gate resistor and proper negative drive but the 3 terminal SIC will only offer slight improvement. Let me say fast kelvin IGBT will provide the same transition as a 3 leg SIC. So you will be limited by the package and not by the silicon die.
reply
First of all, one is transform-less inverter, and it will have usually above 98% eff (boost inverter, the other one has an insulation transformer, to boost the 48 to 400V and is cascaded by the inverter and by this you will have a max of 94-96% eff. Regarding the replacement of IGBT with SIC, you swapped 10 pieces, usually are only 4 or 6 in the inverter stage, so don't know what would be the purpose of the others replaced. The on-off time measured over the SIC, is quite bad. You should get lower than 60s with kelvin connected SIC. So you have bad results due to 3 leg SIC, 4 or 7 leads are the optimized for fast SW. You may improve your results with low gate resistor and proper negative drive but the 3 terminal SIC will only offer slight improvement. Let me say fast kelvin IGBT will provide the same transition as a 3 leg SIC. So you will be limited by the package and not by the silicon die.
reply
shyne2643
Important note: don't connect your oscilloscope's probe grounding clip to mains level neutral or even worse, live wires!
It can damage the scope, cause injuries, and will definitely make the differential breaker trip as the grounding clip of oscilloscopes is connected to earth, as well as all other visible metallic parts. This only worked here because it is a fully batery operated inverted which doesn't have any earthing, which is also bad practice but it's only for the demo.
Second important note: SiC MOSFET gate driving is very different from IGBT and Si MOSFET, it isn't a drop in replacement = more efficiency, like said in the video the driving stage isn't designed for it. But more importantly: some driving stages might even DESTROY SiC MOSFETs as those CANNOT endure negative gate voltages, whereas classical Si MOSFETs and IGBTs don't mind.
reply
Important note: don't connect your oscilloscope's probe grounding clip to mains level neutral or even worse, live wires!
It can damage the scope, cause injuries, and will definitely make the differential breaker trip as the grounding clip of oscilloscopes is connected to earth, as well as all other visible metallic parts. This only worked here because it is a fully batery operated inverted which doesn't have any earthing, which is also bad practice but it's only for the demo.
Second important note: SiC MOSFET gate driving is very different from IGBT and Si MOSFET, it isn't a drop in replacement = more efficiency, like said in the video the driving stage isn't designed for it. But more importantly: some driving stages might even DESTROY SiC MOSFETs as those CANNOT endure negative gate voltages, whereas classical Si MOSFETs and IGBTs don't mind.
reply
UniCrafter
I went from IGBT to SiC on a buck converter I made, the efficiency improvement honestly astounded me. I went from delta 125C at a load of 1000w to a delta of 25C. Also the SiC are more robust in shoot-through events in my experience. Something to note though: IGBTs can be happy with gate voltage of 15v, but SiC prefer 18v. Also some efficiency improvement comes from running higher frequency, requiring fewer turns on the inductor, and thus lower series resistance. I've had issues with load sharing with SiC, would be interesting to see how this does.
The buck converter has inductance of 19uH, switching at 160kHz, IGBT was Infineon K100EEH7, SiC is 65R010M2. Managing to get over 3kW out of it.
reply
I went from IGBT to SiC on a buck converter I made, the efficiency improvement honestly astounded me. I went from delta 125C at a load of 1000w to a delta of 25C. Also the SiC are more robust in shoot-through events in my experience. Something to note though: IGBTs can be happy with gate voltage of 15v, but SiC prefer 18v. Also some efficiency improvement comes from running higher frequency, requiring fewer turns on the inductor, and thus lower series resistance. I've had issues with load sharing with SiC, would be interesting to see how this does.
The buck converter has inductance of 19uH, switching at 160kHz, IGBT was Infineon K100EEH7, SiC is 65R010M2. Managing to get over 3kW out of it.
reply
RyanRoss-d6z
At 5: 32 those are the battery side mosfets. If you follow the traces, they are connected to a center tap on the main transformer, and flip flop each opposite leg to get the square AC to go through the transformer coil, which its what creates the highV AC on the left hand side of the board. The highV AC is then rectified and smoothed by those giant caps, and are the power source for the IGBTs on the highV side. The solar FETs are on the daughter board. I think you'd have a great increase in efficiency by replacing the battery side FETs, but they are difficult to take out and low RDSon FETs of 150V and the same current limit are expensive
reply
At 5: 32 those are the battery side mosfets. If you follow the traces, they are connected to a center tap on the main transformer, and flip flop each opposite leg to get the square AC to go through the transformer coil, which its what creates the highV AC on the left hand side of the board. The highV AC is then rectified and smoothed by those giant caps, and are the power source for the IGBTs on the highV side. The solar FETs are on the daughter board. I think you'd have a great increase in efficiency by replacing the battery side FETs, but they are difficult to take out and low RDSon FETs of 150V and the same current limit are expensive
reply
davebutler1264
I bought a cheap hybrid inverter on ebay. It is labeled Daxtromn AGH-6. 2KW PRO. But there are many similar units available.
It is dangerous!
It is extremely noisy with fans like jet engines coning on as soon as the sun comes out. The power it makes must be way off normal because my induction hob refuses to switch on.
But worst of all, it redirects grid/utility power inyo the earth. It dried out my earthing rod and started causing shocks from the kitchen sink.
The supplier insisted this is normal!
Just disconnect the earth wires was his solution.
Horrible device!
Avoid cheap hybrid inverters if you can!
reply
I bought a cheap hybrid inverter on ebay. It is labeled Daxtromn AGH-6. 2KW PRO. But there are many similar units available.
It is dangerous!
It is extremely noisy with fans like jet engines coning on as soon as the sun comes out. The power it makes must be way off normal because my induction hob refuses to switch on.
But worst of all, it redirects grid/utility power inyo the earth. It dried out my earthing rod and started causing shocks from the kitchen sink.
The supplier insisted this is normal!
Just disconnect the earth wires was his solution.
Horrible device!
Avoid cheap hybrid inverters if you can!
reply
tiagotiagot
Hm, I hadn't considered that; I've been thinking for a while about looking into getting some balcony (more like window-hung) solar and a battery setup for my apartment to have an whole-house UPS kinda thing (no feeding back to the grid, just charging on the cheaper hours, and if possible adding up power if the current demands exceed what just the UPS can provide); but if the battery and inverter are gonna be heating up the place that would complicate things a lot; the shitty building walls already make my AC struggle to approach the target temperature in hotter days.
reply
Hm, I hadn't considered that; I've been thinking for a while about looking into getting some balcony (more like window-hung) solar and a battery setup for my apartment to have an whole-house UPS kinda thing (no feeding back to the grid, just charging on the cheaper hours, and if possible adding up power if the current demands exceed what just the UPS can provide); but if the battery and inverter are gonna be heating up the place that would complicate things a lot; the shitty building walls already make my AC struggle to approach the target temperature in hotter days.
reply
zodiacfml
while I still don't a solar PV system of my own, at this point I think it is time for the cutting edge to start phasing out inverters or AC current for household. I mean, it is feasible for household outlets to have 110-240V DC power. Appliances would simply need DC converters to lower the voltages required. Household appliances with motors controlled by inverters are effectively DC.
DC-AC and AC-DC conversions is signficant loss of efficiency and cost of equipment.
reply
while I still don't a solar PV system of my own, at this point I think it is time for the cutting edge to start phasing out inverters or AC current for household. I mean, it is feasible for household outlets to have 110-240V DC power. Appliances would simply need DC converters to lower the voltages required. Household appliances with motors controlled by inverters are effectively DC.
DC-AC and AC-DC conversions is signficant loss of efficiency and cost of equipment.
reply
greatscott
Hey GreatScott,
What an awesome video! It was super exciting to be working with you on this project - We're super excited to be working together on showcasing our CoolSiC MOSFETs! These components are designed to enhance efficiency, reliability, and sustainability in various applications, and you showcased this in the coolest way possible: )
Thanks for your creativity and passion for innovation. Together, we're building a better future!
reply
Hey GreatScott,
What an awesome video! It was super exciting to be working with you on this project - We're super excited to be working together on showcasing our CoolSiC MOSFETs! These components are designed to enhance efficiency, reliability, and sustainability in various applications, and you showcased this in the coolest way possible: )
Thanks for your creativity and passion for innovation. Together, we're building a better future!
reply
jeromecolico1960
I love the way you touch about the difference between the switching and conduction loss which is an important parameters in looking out for the efficiency, I would love to see more about an optimization series such as the improvement of the driver.
I was curious though how did you ensure that the gate voltage outputted by the driver is enough,
Could you also share the schematic for the evaluation of the switching loss
reply
I love the way you touch about the difference between the switching and conduction loss which is an important parameters in looking out for the efficiency, I would love to see more about an optimization series such as the improvement of the driver.
I was curious though how did you ensure that the gate voltage outputted by the driver is enough,
Could you also share the schematic for the evaluation of the switching loss
reply
Tarodenaro
You uh. buying loads of expensive Mosfet that makes it more economical to buy better inverter, this is the kind of DIY mentality that plague Audiophile thinking by upgrading $40 ChiFi with $100 OpAmp they could get optimal results with a terrible design and awful power supply, and they used it to play LoFi girl music.
I never really understand why people did this, but it's good for the content i guess.
reply
You uh. buying loads of expensive Mosfet that makes it more economical to buy better inverter, this is the kind of DIY mentality that plague Audiophile thinking by upgrading $40 ChiFi with $100 OpAmp they could get optimal results with a terrible design and awful power supply, and they used it to play LoFi girl music.
I never really understand why people did this, but it's good for the content i guess.
reply
hamzaterzi8801
11: 27 When you increase the gate voltage, the efficiency will probably increase even more. Because the Rds resistance decreases. But is there enough voltage for the SiC Mosfet gate Maybe you can increase the voltage with a separate booster regulator. If you replace the other circuit elements (diodes etc) in the circuit with the most efficient versions, I think the efficiency will increase to 98%.
reply
11: 27 When you increase the gate voltage, the efficiency will probably increase even more. Because the Rds resistance decreases. But is there enough voltage for the SiC Mosfet gate Maybe you can increase the voltage with a separate booster regulator. If you replace the other circuit elements (diodes etc) in the circuit with the most efficient versions, I think the efficiency will increase to 98%.
reply
jamieharland9080
Good video but very complicated however saying that for the parts your time and labour would it not have been more efficient to buy a Vitctron or top quality product as you have no warranty opening a product! But I personally believe that the time would be spent better on a good quality product if you could make that any better I think my one is 96% efficient! But hay makes a good video
reply
Good video but very complicated however saying that for the parts your time and labour would it not have been more efficient to buy a Vitctron or top quality product as you have no warranty opening a product! But I personally believe that the time would be spent better on a good quality product if you could make that any better I think my one is 96% efficient! But hay makes a good video
reply
mohsenSL
good video, we need More videos that study and compare solar energy products
you skipped the best part of the video, which is the most exciting part: improving the timing of the drive circuit for the best performance of the skipped the best part of the video, which is the most exciting part: improving the timing of the drive circuit for the best performance of the sic mosfet.
reply
good video, we need More videos that study and compare solar energy products
you skipped the best part of the video, which is the most exciting part: improving the timing of the drive circuit for the best performance of the skipped the best part of the video, which is the most exciting part: improving the timing of the drive circuit for the best performance of the sic mosfet.
reply
SkippiiKai
5: 52 Your drawing of the IGBTs include diodes like the body diodes of MOSFETs. IGBTs do not have body diodes. You can buy IGBT components where the manufacturer includes a schottky diode on the same die so it works like a MOSFET, but it's not part of the IGBT and should not be in the schematic unless you are specifically referring to one of the combined packages.
reply
5: 52 Your drawing of the IGBTs include diodes like the body diodes of MOSFETs. IGBTs do not have body diodes. You can buy IGBT components where the manufacturer includes a schottky diode on the same die so it works like a MOSFET, but it's not part of the IGBT and should not be in the schematic unless you are specifically referring to one of the combined packages.
reply
martinleska4292
Learn us how to test and check if this hybrid inverters backfeeding grid constantly or only for short period when huge load from output is removed with or without battery from solar panels. If you have legal solar installation with allowed backfeeding to grid, another extra inverter for controlled period of time doesnt hurt anything.
reply
Learn us how to test and check if this hybrid inverters backfeeding grid constantly or only for short period when huge load from output is removed with or without battery from solar panels. If you have legal solar installation with allowed backfeeding to grid, another extra inverter for controlled period of time doesnt hurt anything.
reply
africanlife5922
Really a Felicity Solar Inverter I am using 12 pcs of the 5048 in parallel. My biggest Problem with the Chinese Inverters which i am using thins nearly 10 years now is the high base power they need. the 5048 for example needs around 50 Watts in idle. So maybe in one of the next videos you can try to bring that down: -)
reply
Really a Felicity Solar Inverter I am using 12 pcs of the 5048 in parallel. My biggest Problem with the Chinese Inverters which i am using thins nearly 10 years now is the high base power they need. the 5048 for example needs around 50 Watts in idle. So maybe in one of the next videos you can try to bring that down: -)
reply
Add a review, comment
Other channel videos
