Does it suck? Chinese DIY Pure Sine Wave Inverter || Sinusoidal PWM (SPWM) Tutorial

[@alexbeardmore3588]

Nobody should underestimate the difficulty of inverters. They are hard to make properly, and mains voltage ones are frickin' dangerous too.

[@greatscottlab]

Well said.

[@tcurdt]

But it would be pretty great if there was a good open hardware design for a 1000-2000W 12V to 220V inverter around.

[@TCGProductions03]

@FQD2N Extremely high current is available, and there are nasty frequencies involved. Touch the wrong thing and you might have 40A at 6kHz flowing through you. I bet that would hurt like the dickens and you probably wouldn't be able to let go.

[@jaideep1337]

As someone who worked on a grid connected solar inverter. You're right. Its even harder when you have to design it according to set standards. Minimal Harmonics and leakage current etc.

[@TCGProductions03]

@FQD2N well what if you are electrically grounded?

[@Asu01]

The amount effort you put into this video is astonishing, it makes me guilty for not supporting you other than watching. Sourcing the parts on your own must be such a real pain in the ass to do, and the fact you said _"it doesn't blow up"_ multiple times clearly tell us you must be very happy with your hard work! And oh, the principles of sPWM looks real similar to class D amplifier, only with some feedback inputs. Now I remember my silly idea years ago, using audio amplifier with step up transformer to make a sine wave inverter because back then, sine wave inverter was so expensive, it's literally unaffordable but today's sine wave inverter are getting cheaper and more affordable, there's no point in making your own with the risk it poses. Seriously, just get yourself a proper one.

[@Electroblud]

Using an audio amplifier with step up transformer to make a sine wave inverter is something that can be done in a pinch and actually works fairly decently. Just don't expect a stable output. Oh and it also is very dangerous, obviously.

[@Asu01]

That's why I never do it :-)

[@jacktheninja]

@@Electroblud can be isolated.

[@KerbalLauncher]

@@Electroblud It's a pretty poor one, the proper way to make an inverter is to step-up the DC bus voltage before feeding it to your inverter, that way you only need a high frequency SMPS transformer instead of a mains transformer. Also, if you're looking for electronic components, Arrow.com offers free shipping worldwide on any order. They just don't have everything in stock like digikey does.

[@jamescooke3763]

I remember seeing a 3 phase 400Hz 220V power supply built using H&H audio amplifiers in a aircraft instrument factory during the 1980's. It was a very fine piece of equipment and gave a very impressive performance. The disadvantages were that its efficiency was abysmal and it was not very portable. It was mounted in its own 19" rack with heavy duty castors.

[@NarayanLoke]

I want ElectroBoom to do a video on a DIY Pure Sine Wave Inverter, so that I can see how badly this thing can blow up.

[@oshosanyamichael9589]

FULL BRIDGE RECTIFIER!!!

[@netrocker9990]

😀😀😀

[@shamnadabd7266]

Me too... Lol

[@th3osl333]

FUulBriDgeRectIFiyAR

[@midnightfoxxx8231]

That would be grand

[@chipheadnet]

I just became a patreon! This young man gives so much and I have learned something from every single video. He deserves all the support we can give him to keep this high quality content coming. I challenge the other 950k+ subscribers to become patreon and support this great content! Thank you GreatScott!

[@greatscottlab]

You're welcome and thanks for the support.

[@francoisleveille409]

Back in 2006, I built my own 150 watts pure sine wave inverter using the PWM method. Instead of using a micro controller, U sed a simple comparator which compared a 6 KHz triangle wave with a 60 Hz sine wave generated by a wave shaper (6 diodes and 2 resistors). I have to say I was amazed by the result.

[@georgievvladimir]

great idea.

[@jamesjohnson5700]

Hi What do I need to build 10k watt inverter?

[@francoisleveille409]

@@jamesjohnson5700 Depends on a lot of things. What input voltage, what output voltage, What type of what you want on output (sine or pseudo). A 10kw inverter may seem a big deal but you can already get 3kw inverters off the shelf at Walmart or a large hardware store.

[@jamesjohnson5700]

@@francoisleveille409 I wanna build my own with high & low voltage protection & can the input & output be the same I want sine

[@francoisleveille409]

@@jamesjohnson5700 You'll have to give more detailed specs if you want something from me. Also, I can't exactly give you a detailed schematics in a comment section.

[@rickpalechuk4411]

Well....thanks for taking one for the team, a lot of hoop jumping in this one. And the technical term of the day: "it doesn't blow up" Cheers

[@satibel]

that's why I always have a "blow up" filament lamp in series with the power supply of what I'm testing.

[@semicolontransistor]

Great video as always! Also, I would totally be happy to translate any Chinese manuals into English for you and your viewers. Keep up the good work!

[@flomojo2u]

You can import any PDF document into Google Docs and then translate it from Chinese to another language, the results are actually astonishing, very good. It can of course export to PDF as well. You can’t use Google’s online translator due to a size limit they imposed, but there is no such limit for Google Docs.

[@rikka0_059]

Im Chinese and I have built sinusoidal inverter before, not with this chip but with stm32 microcontroller. I can do the datasheet translation for you. Perhaps one day i should develop an open-source sinusoidal inverter and share it with everyone!

[@IustinianP]

Rikka0_0 we can gather ideas,and make an open source project,maybe..

[@王鑫-t3v]

You have a lot of Chinese fans and we can help you translate the manual. 其实我们阅读英文的datasheet的时候也很痛苦的。

[@tonnydavila2154]

That was a lot of work, there you have my thumb up... Some time ago I made my own inverter but with a half bridge, and it worked like a charm. Thank you Scott! A lot of knowledge in 12 min

[@greatscottlab]

Thanks for the feedback :-)

[@GuyonaMoose]

Most of the time I have no idea what your talking about but I enjoy your videos. Thanks for helping me to learn

[@Solar_Tech_Liam]

20W is actually quite good, the multi-thousand dollar pure sine wave off-grid inverters often have standby losses of 30-40W continuous. The main reason being transformer losses and running the control circuits. I would be very interested in this module for a 48V or higher input application, really just a question of the right MOSFETs and filter design at that point.

[@lyleg.9192]

I just read your comment after making a post something similar about inverter sizes and power drains

[@stanleyodo1249]

I designed my inverter system from scratch and generated spwm using PIC microcontroller. I had no load power of around 40W and had been researching if that is too much because I feel my MOSFET and transformer heats up more than required. I switched at 18kHz

[@mathiasbackof4993]

Our nice German engineer is back with another asweome video!

[@greatscottlab]

Oh stop it

[@mathiasbackof4993]

@@greatscottlab Im never gonna stop it : - )

[@Kevin-jz9bg]

​@@mathiasbackof4993 yes we don't stop cheering you until you stop working so hard to teach people electronics which is hopefully never because this channel is asweome :)

[@dimitriosvlamis8696]

Hi! Nice video as always! But i think you used the wrong pcb for the thing you're trying to do! This pcb was made for the version where you put 400V in the board from a different source. You used the same compoments for the version with the iron core. The IRF840 are high voltage, low current mosfets with a BIG RDSon so thats why you had so much idle consumption. Try using the IRF3205 like they said in the datasheet with 8mOhm RDSon and bigger current, even in parallel! Also the transformer should be 8V to 220V and not 12-220 :D

[@antonf.9278]

Please use Chinese if you want to communicate with him

[@dimitriosvlamis8696]

@@antonf.9278 对不起

[@antonf.9278]

Now I understand Thanks

[@markusantonov6373]

Are you Estonian?

[@dimitriosvlamis8696]

@@markusantonov6373 İf you're talking to me, no 😁

[@luongmaihunggia]

8:15 first time he gets angry on video.

[@chargedsupercap2270]

No, he got angry when he built a Bluetooth ceiling speaker. "I destroyed the receiver because I am such a *genius* ".

[@jm036]

@@chargedsupercap2270 He called things bullshit multiple times before aswell.

[@elek101]

@@jm036 But he never got this angry lol

[@kamunreser2492]

Ambilight part 3.

[@luongmaihunggia]

@@elek101 exactly.

[@JeremyDWilliamsOfficial]

Wouldn’t it be more accurate to use a controlled-speed spinning magnet to generate a pure sine wave reference voltage? I know from a mechanical sense it would be more complex, but it may be a valuable trade off to make a perfect wave. I’m suggesting a VERY small motor/magnet. Purely for the wave shape. Thoughts?

[@polymetric2614]

Well the problem with the wave isn't because of the low current signal. It's because of the nonlinear characteristics of the inductor, the transformer, and the load. So, for example, in an exceptionally simple case, a rectifier draws more current when it gets to the peak of the sine wave, which brings the voltage down (mainly due to resistance upstream in the power path, like the inductor and the mosfets) only at the peak and gets us that sine with the flat bit at the top. I actually really like your thinking there, though. Largely because I thought of a similar thing a while ago. There do exist rotary converters, which is basically just either a DC motor spinning a single-phase generator, or a single phase motor spinning a 3-phase generator. The problem with these is that they're big, heavy, loud, but most importantly, they constantly consume current even at no load. They're good for some uses, like if you need 3 phase power when you only have single phase, and you only run it when you're running a big tool or something so you don't care about its inefficiency because it's not running for very long. But inverters are much better for general purpose use, which is why they get used so much. I believe they're also more efficient even at full load. You get a lot of eddy current losses in the big steel plates you need for a mains frequency transformer/motor. Inverters step up voltage at much higher frequencies (10s-100s of kHz), which means a much smaller transformer -> less steel -> less eddy current losses -> more efficiency. The trick to making a good sine wave with the inverter is all in the filtering. Picking the right combination of inductors and capacitors to balance the cutoff & resonance, good high frequency filtering while not suppressing the fundamental frequency that you want to keep, an also keeping component cost and size in mind. High current, high inductance, low frequency inductors get real big real fast. And since they use thick copper wires, they get expensive real fast too. And you can't use the advantage of the high frequency switching needing a smaller core like we did with the transformer because mains needs to be low frequency. This, by the way, is why cheaper inverters have more THD. They use cheaper components that don't filter as well (and probably spent less R&D money as well) so you get more harmonics. I'd say likely the most expensive parts of any inverter are the filter components.

[@jlucasound]

Hopefully your video will influence all Chinese electronics manufacturers to provide properly written, English manuals. It will benefit them as well as us and I DO love electronic modules that come from China! So many to choose from and so inexpensive. And they usually work!

[@bkboggy]

Looks like there's a big value in learning Chinese, considering they're the biggest source of electronic components these days.

[@vaio232]

great info Scott, thanks for shedding light on spwm. you could always buy a commercial that does everything well for us but having building on your own, gives us a way to customize the thing anyway we want

[@bertondroid]

I use Google translate with the camera of my phone to translate chinese.. which works quite well! ;)

[@exogator]

You can use Chrome to do it on the webpage too

[@Gameplayer55055]

And you will get very strange language with it. Because chinese hieroglyphics are overcomplicated

[@ravien6142]

questoin: type the translate languege

[@adityag.5372]

Just use the "translate document" feature of google translate. Makes life much easier

[@JeremyDWilliamsOfficial]

Chinese doesn’t translate well to English. It’s fundamentally different. You can make a translation just fine for the words, but not necessarily for the meaning behind them. That’s always an issue. 🤷‍♂️

[@DAFUQ486]

You completely unmounted my doubts about SPWMs and Inverters in less than 5 minutes. All Hail Great Scott!

[@IrishSkruffles]

Hey Scott, the circuit shown on the EGS002 document is for converting *400V DC* to AC, so there may be some issues with your method of using the filtered hbridge output into a transformer.. Did you get any rough efficiency calculations when you were powering the lightbulb?

[@peterdkay]

Most inverters step low voltage up to 400V then use full MOSFET bridge to create 240VAC from that 400V. This improves the efficiency and size as transformers work at 20kHz to generate 400V and conversion to 240VAC only needs simple filtering. Example: I have a 3000W Chinese sine wave inverter which has a no load power of around 10W

[@WereReallyRelayCamping]

When doing this pwm to an H bridge you do not need to feed pwm to both sides of the bridge, you can feed pwm to one side and just turn on the other. Most ebike speed controllers work that way just for a bit of info

[@Mrengineer1]

Awsome Video Dear!

[@swalehmuhammad2428]

Good good

[@abdullahmuhammad8296]

Azeem bayi ya kafi cheta banda haa, app ki tarah :)

[@MasterofOrion]

Wtf

[@MasterofOrion]

Weird

[@fawadansari89]

I have used EGS002 in my inverters and it works very fine. I have been using it in line conditioners also

[@laptrinhplc3031]

ahmed fawad can i question for you ? I am building inverter 12vdc-220vac using egp1000w and gs002 make by Chine. This is the same video above. But output vac only 120vac, i am using tranformer 7.5vac to 220vac and irf840. Please, can you help me? Share for me your project. Thanks so much.

[@jonaslithen7240]

One suggestion to decrease idle consumption, add some inductance in series with transformer primary, other wise the current through the capacitor will be too high. I have done some experimenting with these exact boards :) I fully agree that they...work, but are not very "ready for use".

[@sylkelster]

Your knowledge is incredible, and your soldering technique, well....

[@thomgt4]

The most effective one I've made only created the sine after bumping the voltage to 320V with a DC-DC boost converter. This negates the whole problem of having stupid lossy transformers and difficult feedback loops. It has it's own disadvantages, like high frequency, high voltage switching and a high voltage DC supply (more dangerous than ac) but it was hella efficient and didn't care about load attached

[@josepeixoto3384]

that is the way to do it,my 2 inverters of2500/5000 watts are like that.

[@e.h.lipton73]

See alot of comments, and no where do I see a suggestion for heat sink and cooling of FETs most inverters if not all will shut down when overheating and have an led that indicates low voltage supply from source or to much draw heating up circuit, thee led will flash before shutting down completely. You need to finish this up,,, try heatsinking the FETs and fan. We know resistance changes as temperature goes up and so do the electrical engineers.

[@loukashareangas4420]

The chip on the EGS002 board has a variable frequency operation mode as well as a fixed V/f option to accomodate the reduction of inductive reactance as frequency increases. I think that this can be used to create a single phase VFD (230V 50Hz in) that should be able to be used with single phase induction motors to enable speed control. This is on my backburner at the moment, so anyone feel free to try it out!

[@olandorobertson2510]

Will do. I deals with vfd controllers too. I thought that chip could be used to control motors. Good point.

[@deepaknanda1113]

​@@olandorobertson2510 any progress ??

[@rakeshswami11]

Wish if we had content like this in our college era.

[@annaplojharova1400]

The "ASIC" could well be a microcontroller with the firmware in ROM. Definitely the ATMEGA in Arduino will do the job as well. In fact it does nothing else than the table generation you tried yourself, but instead of directly feeding to the counter, these numbers get multiplied by an amplitude variable. This variable comes from sampling the output voltage at the moment when it is supposed to be at its peak, comparing it with the desired peak value and then adjusting the amplitude variable till the measured voltage matches the target. The ADC is scheduled so at most times it measures the current, once the 20ms cycle it replaces one sample by voltage measurement and few times per second by the thermistor measurement. What makes me to think it is actually a generic microcontroller with custom firmware and not a true ASIC is how much of auxiliary analog circuit it needs around (not talking about the MOSFET predrivers, but really about signal processing within 5V rails), which could way easier be integrated onto the ASIC chip, so saved the BOM cost if it were truly designed as an ASIC. But all the circuit around really suggest there was no dedicated IC design involved, only programming of an existing micro. Which makes perfect sense, as designing an IC is VERY expensive and you have to sell millions of them to get that money back on savings compare to using some standard MCU. Don't think there are that much inverters sold to justify a fully custom IC for that...

[@himanshubhaskar6631]

Your Reverse engineering electronics skills are Insane.

[@besserwisser-702community4]

8:23 FULL BRIDGE...

[@crazyksp8344]

#FULLBRIDGERECTIFIER

[@TCGProductions03]

*presses echo pedal* "FULL BRIDGE RECTIFIER"

[@koharaisevo3666]

Expecting him says Brückengleichrichter :D

[@chargedsupercap2270]

*RECTIFIER*

[@0ADVISOR0]

Rectumfrier!!!

[@bansukm20]

One of the best tech/electronics channels in YT, 👍😍, as usual

[@timm3802]

Hi Scott Thanks for a great video....... I'm using the EGS002 and it's working fine..... IF you build your own board. But the most expensive part is the toroid (transformer). A bigger toroid=more power. Btw, try to use as high voltage as possible. Then you decrease the amps, because the AMPS is the backside of low voltage. No (normal) wire can hold 100A continuously at 12volt, but just by increase the voltage to 24volt, the amps will decrease by 2x. And at 48volts you have "only" 25amps running in the cable to the battery. Thats why the Chinese circuit shows 400V DC. Because they can keep the cost down on wires and transformers aso.

[@ГеоргиКовачев-г4к]

Tim, can you share a schematic for a inverter?

[@timm3802]

@@ГеоргиКовачев-г4к Yes no problem. But i don't know where to share it...any suggestions ?

[@ryanunirabaga]

@@timm3802 can I have a copy of your designed inverter? I want to build one with only small power rating like 200-300W for lights backup only uses and phone/laptop chargers.

[@ГеоргиКовачев-г4к]

Tim M may be on my email? [e-mail]

[@timm3802]

@@ГеоргиКовачев-г4к no problem, I will mail it to you.

[@og9806]

It was your mistake. It has peak to peak voltages of 230V, but for 230V RMS, you need an 8.5V to 230V transformer.

[@jaakkolehto1487]

The problem with your inverter is that if you use 12V to 230V transformer, the peak voltages it can produce are only 230Vpk, while mains peak voltage is 325V. And if we divide 230 by 1.414 (square root of 2), it is around 162V as your oscilloscope showed. So if you wanted 230Vrms, you would have to use around 17V

[@sullivanzheng9586]

I can read Chinese and as I can see from the manual, it requires very complicated tweaking. You were EXTREMELY lucky to have worked it out without manual just by trying, risking and erring, and still got away without being electrocuted or have your lab burnt to ground :::)))) Your takeaway is just to the point....never try anything without a manual. I do cross fingers for those who are also trying this.... high power inverters are impossible for DIYers without 1-2 year EE college training.

[@洛天依-n1x]

I used this driver board, EG8010. I completed my graduation project. However, it seems that my version is different from yours. My version requires a DC of about 400V. After the H-bridge, the output is accurate AC220V.

[@jc-zh9kl]

GreatScott gets pissed! the man the myth the legend lol great vid!

[@baldbadger7644]

I can translate the manual for you if you will make a video about SSTC one day

[@ajayajith1745]

Hey man nice video, but I thing at the beginning of the video u said Spwm is used for obtaining pure sinusoidal wave but I it's only used for controlling the inverter's output ,sine wave is obtained by using filters correct me if I am worng

[@ewbaite]

Maybe I can help you translate it? Am I the only Chinese subscriber?

[@billcrm003]

no，我也是

[@altairfoo1920]

我也是

[@rhiantaylor3446]

That would really help, here is the chinese datasheet. datasheetspdf.com/pdf-file/968003/EGmicro/EGP1000W/1 Using Google, it suggest the board can either be used with a high voltage input or with a low voltage and a transformer. In the latter case, I thought you need a small second transformer on the output side to provide an isolated low voltage voltage to the control board but I couldnt see that in the video......

[@OmarElmasry1]

Please do sir

[@stephensu4371]

no

[@ivangutowski]

Thank you for the level of details you went into, this solidly confirmed that it's better for me to work for a few days and buy and inverter rather than making one :)

[@Davide21570]

2:43 What a perfectly drawn sine O_O

[@PHM_Tech]

Yes wow

[@thomasb4422]

Regarding the waveform at 10:30: It looks like your LED lamp uses a full bridge rectifier with a smoothing capacitor. The voltage at the transformer drops very hard while it's loading the capacitor, so the output just seems to have a high internal resistance. I guess that's because you used a transformer with 230V for the input, but just hooked it up "backwards".

[@robdavis3220]

Great video. I think it's worth pursuing this project further , if your comfortable working with mains voltages :0) These things are particularly dangerous as there is very little protection for you on the output. No earth leakage is going to trip if you accidentally tough the output L and N. That can ruin your day... You can move your LC filter to the output of the transformer. That way is does not have to handle the high currents you will see on the primary. The large transformer inductance will also naturally act as a filter to the high frequency pwm waveform. The H-bridge circuit can also only generate a "PEAK" voltage equal to the supply voltage , so a 220/12V transformer will only be able to deliver just over 150Vrms. I think 1000W is a bit ambitious for a 12V system. Do-able , but your primary currents are going to be very high. You would be better off at 24 or 48V.

[@786mab]

dear! you can use "office lens" to translate the manual. simply takes the picture and convert it into English language and save in desired document format.

[@strssko]

It is pretty good translation. You just need to do some shots of vodka and then you will understand everything.

[@NicoDsSBCs]

Great video. I can't understand how you can make a video every week. This must have been so much work. Thank you. Greetings. NicoD

[@n3r0z3r0]

You did everything wrong! 1. Make High frequency (27-100KHz) step up inverter with pulse transformer DC 310V for 220v output 2. Use this Sin - wave full bridge to output perfect sin wave with regulated voltage. No bulky transformers needed.

[@jonaslithen7240]

That is the cheap way...that also blows up easily. Seriously, even the expensive ones, like Sinergex keep blowing up. I have repaired many. A full bridge + 50Hz transformer can make a robust inverter with nice output. One thing though, the 1000W board has a input current monitoring circuit that causes a rather high voltage drop, this is not practical with e.g. 12V input.

[@InternalWorm]

To increase efficiency you should take another point of view. For first boost 12V DC to around 320V DC and then modulate it to get the proper sine wave. This is how it's done in many modern inverters. Anyway, greetings from Poland. Keep doing great job.

[@javierperez_21]

Make a video about jfets or class a amplifiers!!! Thanks for the awesome video!

[@greatscottlab]

It is on my to do list

[@drw0if]

Ooh, good idea! It would be awesome an electronic basics on class a, b, ab amplifiers!

[@javierperez_21]

@@greatscottlab yeeeesss!!! Thanks! I really like your videos!

[@Mr31Chris03]

Im using it too but i use 12v for the controller and for primary i use 12v to 400v dc-dc converter that allows me to use the 230v ac output without the need for a transformer but in this case you need to chose other FETS or IGBT's that can handle the 400v DC Link under load the output is stil a nice sinewave

[@sefalibhakat143]

please on SMPS

[@MarcoFranceschini1971]

Great as usual Scott...regards from Italy.

[@cxmmax4265]

can u show to do hud glasses ?

[@greatscottlab]

Not sure if I am that interested in such a subject.

[@qwertyui90qwertyui90]

WHat about skipping the chinese controller board and all that stuff and just making your own inverter 100% from your own design > ?

[@razhterize]

8:15 Funniest thing in your channel in past 2 years lol

[@hitbyligtning9661]

GREAT EFFORT in your analysis !

The 50hz transformers should be banned. Smps 500khz-1mhz 12v to 300v dc, pwm Into h-bridge, filter, feedback and you no longer lose 50% of the energy in winding resistance and FeSi core eddies.

[@ariefnoorrahman3492]

I am power electronic engineer... Well, let me tell you the efficiency of 50Hz xfmr is >98% for distribution and transmission system While, typical smps operate at around 50-100kHz, of course there is some operating at 500k-1000kHz...but you will pay 3-4x the typical sms Fyi, I was working on wide array of projects: 500kHz converter, solid state transformer, grid converter

[@adamsucksatyt]

This is actually insanely simple. It just works like an AM radio It modulates the 50 Hz sine wave in a different higher frequency So the carrier wave gets modulated and cleaned up with a low - pass filter which leaves you with the perfect sine wave Which then gets transformed up to 230V and voila you got an pure sine wave inverter

[@pooorman-diy1104]

I'll go simpler way ....DC motor ..coupled to simple AC generator ....voila ... pure sine wave inverter ...lol

[@clarkso65]

That is the best way no fuzz:))

[@addydiesel6627]

And call it = noisy hardware pswi 😂

[@abhaydeepkamboj9951]

Yeah it will work but with considerable energy loss , the following method show in video have negligible energy loss

[@masrurridho]

i like this chanel becase every video he always explain with drawing manually.

[@alexmustang8177]

*I will develop a Doremon anywhere door and steal your oscilloscope*

[@beedslolkuntus2070]

Alex Mustang 😂😂😂😂☹️☹️ The good old doraemon reminds me a lot, do you know a site where we can see the anime

[@sasanka41]

​ @GreatScott! There are three filter electrolytic capacitors (C1, C2, C3) and two decoupling capacitors (C14, C15) on the on-board power supply bus. The withstand voltage of the capacitor should be selected according to the power supply voltage. If working in high frequency mode and the power supply bus voltage is (330V~400V), select 450V withstand voltage. If working in industrial frequency mode and the power supply bus voltage is 48V, the capacitor withstand voltage should be 63V. The electrolytic capacitor capacity is selected according to the actual system power. Generally, the high frequency mode should be greater than 47uF, and the power frequency mode should be greater than 470uF. The decoupling capacitor is generally 0.1uF CBB capacitor, and the withstand voltage must be greater than the power supply bus voltage. If the withstand voltage of the capacitor is less than the bus voltage of the power supply, the capacitor may explode!

[@Frankyyodi]

See? without decent manual you still done it i with right manual still exploded

[@MrDenisJoshua]

A greate project can be a microinverter DIY for the photovoltaic panels :-) Thanks a lot for the video

[@thebestofall007]

Maybe a Chinese viewer who knows English well can chime in here and translate that manual for you.

[@Raven-fu1zz]

I know the video is over a year old, but i would love to see you design an inverter based off of the Egs002, i love the stuff you design

[@heregundir8292]

A French watchin an english video, which use frequently chinese.... Languageception -_-

[@me3333]

Considering he is a native German I would say that's super languageception

[@heregundir8292]

@@4.0.4 Cakes ?? Why ? xD Bread, ok, frog legs, ok, but cakes.... Why :)

[@heregundir8292]

@@4.0.4 i agree with the kebab thing. But that's a bad part of french gastronomy... 😉

[@leozendo3500]

Very cool project. This is what I was interested a long time ago. You can ask your Chinese viewer to translate the manual, should be a very simple task.

[@francescovalle8678]

This is why inverter are expensive...

[@stevenA44]

Chinese products with a proper English manual? LOL Yea right!! Really like your videos. So much good info in them.

[@m4gmu5hell]

*slightly salty* .___.

[@araigumakiruno]

8 Hours ago!?!?!?

[@MrEmrys24]

It must be from the dried tears shed trying to understand the Chinese data sheet

[@wi_zeus6798]

@@araigumakiruno Patreon

[@davey2k12]

FFS bro you got some knowledge with electronics and talks to ......you a legend ....keep it up

[@legominimovieproductions]

Now i feel the need to design a good working, safe, and reliable and cheap inverter XD

[@thatoneguy99100]

This is really interesting to see your approach, I was just working on my own version of this last month but I got busy and shelved it for a while. Maybe now I'll try to finish it.

[@MikaelIsaksson]

Took me a while to figure out that putten-shamaeter is potentiometer in your videos, otherwise, totally excellent videos!

[@jeremyallard5449]

You are patient beyond belief

[@GGBGameplays]

I'm Electric Engineering student interest on inverters. I've tried some high frequency push-pull projects using TL494, some MOSFETs and old ATX PSU's transformers, but I couldn't handle much power in output without frying the MOSFETs. I don't have an oscilloscope, so it's hard to me to understand what in fact is happening in to the wave, bad wave form could be the reason to the MOSFETs' to have short circuit between drain and source. I used some IRFZ44N wich are for 49A, but they barely could handle some watts in my projects. I was trying to have a DC voltage as output and then try to transform into sine wave with an H-Bridge later, but I didn't have decent results in the DC/DC inverter.

[@alelectricll]

Oscilloscopes are very handy indeed. Anyways I assume you rewire the transformer so you have two identical coils to push pull? Any significant imbalance can be disastrous. A possible problem could be incorrect frequency drive to the transformer if the transformer wasn't rewired and the operation frequency is unknown. Now for the MOSFETS, are there any snubbers for them? I admitting I haven't use the tl494 just the sg3525. Also what is a few watts? 20W?

[@GGBGameplays]

​@@alelectricll well, I could light a bulb of around 25-30W, also could power a laptop connected directly to the power supply (no battery) but with agressive voltage oscillations. The circuit was pretty simple, I just used some BJT's before the MOSFET's and connected each directly to the ATX trafo, and powered the center tap with the battery voltage. As I used the IC's internal's transistor collector to driver the MOSFET I needed an external BJT to invert the wave format. I tried also to drive from the emitter, but it didn't seem to work well (the MOSFET didn't turn off). But driving with the collector I didn't use resistor between the gate and source, the MOSFET was discharged through the BJT BC547B, and this maybe was a problem. I don't have much knowledge about electronics because it's not the focus of my college course, I had to learn some things by myself.

[@alelectricll]

@@GGBGameplays I see, sg3525 wouldn't necessarily need external BJTs that's why I used them mostly. But yes, I see a problem. Connect the internal transistor collectors to VCC; the emitters would drive the MOSFETs. Now connect a resistor about 1k to each emitter and ground. You'll now need 2 in4148 switching diodes and 2 BC557 transistors. Connect each emitter to the base of a BC557 and the collector of each BC557 to ground. Now you are going to connect a diode to each BC557, with anode on the base and cathode on the emitter. With a suitable gate resistor, connect the emitter of the BC557 to the gate of the MOSFETs. The diodes will turn the MOSFETs on and the BC557 will turn them off. There should also be a 10k resistor connected to the gate and source of the MOSFETs. This will prevent false turn on of the MOSFETs. This is one way on how to use the TL494 to drive MOSFETs. A concern for me is if the transformer is driven at correct frequency also.

[@GGBGameplays]

@@alelectricll when I tried to drive using emitter some weeks ago it didn't work as intended. I followed some existing projects on YT, just changed some components. That's the schematic: www.clubedohardware.com.br/uploads/monthly_2020_05/277673637_circuitoinversor.png.69a289a21aa9a53c351eef39ee2a7390.png In the simulation it works as expected, but in real life one of the MOSFET gets hot really fast, also there's waste of power (I measured high current flowing from the battery with nothing more than some diodes and one capacitor for filtering DC voltage). Sometimes the MOSFET ends up with short-circuit between gate and source also drain and source. I tried two power BTJ transistors (13007 found in some cheap ATX PSU) instead MOSFETs, the same happened. Regarding to your instructions, in my tests with the IC only, using a resistor connected between emitter and ground, and collector to VCC, the measured voltage on the emitter is very close to the supply voltage (no oscillation). When emitter is open circuit, the measured voltage is around 40% of VCC (which is the duty cycle that I set). So I guess it wouldn't work, but I've seen some similar schematics using the emitter directly to the base (no resistor for grounding), this way should work better. However I'll try to put my hands on some SG3525 soon, I might have less headaches driving the MOSFET directly from the IC.

[@alelectricll]

@@GGBGameplays sadly I was denied accessing the schematic. Is there another way? Hmm one side heats up faster than the other... Maybe a wrong pin on the transformer too? Check just to be sure. Interesting... pull down resistors shouldn't have any effect on the IC switching. Yes you can use it without pull down resistors but it isn't better... your switching speeds would suffer a bit. The sg3525 or even tl598, you only need gate resistors with the MOSFETs. I would like to hear any updates if you can.

[@popcoron11]

マニュアルが如何に必要かと言う貴重なビデオでした。

[@amnesie6615]

Tolles Video, wie immer!

[@dabChang]

Give you a thumb for your amazing skill and Chinese subtitle, thanks for translating (even thou I'm a layman of electronics)

[@fortun8diamond]

Yo Scott. I AM MAKING an inverter with mains output soon. so yeah, driven by Arduino to generate SPWM. I will include a feedback system. It's actually simple. All you need is a full bridge inverter that generates SPWM and analog components that smooth it out into a sine wave. Finally, a transformer to boost it. I will leave out the transformer until I get a good sine wave, though.

[@b4fball]

Great video. I think there is no compromises on inverters. I wouldn't risk and just pay more for good one.

[@jmanbrosef689]

You can also buy the pcb with the parts in kit form

[@georgesiamiotis464]

Hey Scott, bravo for saving lives. (at least mine)

[@user-qp4gs4ky8h]

Thanks for your video! BTW, as a Chinese viewer, I could help you translate Chinese data sheets to English or even German.

[@andrewandrosow4797]

If there is sine wave - you can try to decrease current consuming (without a load) - tune an output capacitor value to compensate reactive power of the 50/60Hz transformer.

[@MSaqib-bm6np]

I have a chinese friend who helped me lot on this chinese stuff, go make a friend but make sure he understands english well :)

[@DaHaiZhu]

11:09 - what is that? some kind of USB voltage and amp meter? name and where to get one?? Thanks!

[@willpreston6881]

I really like the animated drawings. It adds a lot to the learning experience, and I think you should do them more.

[@dipubiswas8520]

I am using this modules properly after some Engineering modifications. You need to modify the transformer. The turn ratio will be 6.5 to 240 ac instead of 12/220v conventional transformer for 12V battery. And for more noload efficiency you need to add 40~100 uH power inductor in series of primary coil. After all it is a good spwm modules with cheap cost.

[@newq]

I'm not sure why, but I really like this guys handwriting. Might be because mine sucks so much.

[@vengug7303]

Great video Sir .... Thanks for sharing .... I need to small help .... The EGS002 Have any settings for HIGH FREQ (380VDC to 220VAC) .... The EGS002 Used Low frequency inverter also 12VDC to 220VAC ..... How to H Bridge Switching will work Please ....

[@rolandmeiner4970]

Juhuu, another great scott

[@ChrisEpler]

@9:02 Did that solder start liquifying on the output?