# Modified Sine Wave with Stable Voltage Output by NE555 (Bipolar)

#### Audioguru

Joined Dec 20, 2007
11,248
Hi Juan,
I haven't seen a PWM sine-wave inverter project. People can buy them for much less cost than they can be made.

Semiconductor manufacturers make very efficient class-D amplifier ICs.
They also make PWM controller ICs that have some parts of a class-D amplifier.
The TL494 PWM controller IC is very old. There are a few newer ones.

#### JUAN DELA CRUZ

Joined May 27, 2008
121
Constructing a Sine wave inverter is quite tricky.
I think I shall continue modifying my Modified Sine Wave Inverter instead.

07-01-2008, 12:56 AM
Audioguru

Senior Member

1) The first opamp's output is a square-wave, not a triangle-wave.
2) The second opamp is an integrator so its output is a triangle-wave, not a sine-wave.
I added another integrator in my previous M.S-wave ckt.

Do you think a diode in each output of NOR gates is desirable??...
....so that the Mosfet will conduct only in Positive Cycle.

Thanks

Last edited:

#### Wendy

Joined Mar 24, 2008
23,134
Keep adding parts and you will make a discrete TL494 at 100 times its price and 100 times its space.

Now add feedback for a regulated output voltage. The TL494 has it.
Yep, added a 4th op amp, this makes it a quad, and 4 more resistors at 3 cents each. High dollar indeed.

By my reckoning I have 3 chippies, 2 quad op amps and a 556, for the whole package, including the sine wave oscillator, which isn't part of the TL494. I've never dealt with the TL494, but the other stuff I have off the shelf. Gotta admit, 32 cents per chip is better than 2 dollars though, and it does appear to be an interesting little chippie.

I am a firm believer in buying finished products, when you can find them, as in the case of inverters. Do you know where to buy a sine wave inverter? Since I'm not in the market I've never looked, the only ones I've bought is for laptops, a non critical use.

Last edited:

#### Wendy

Joined Mar 24, 2008
23,134
Thanks Bill...
Are you referring to a low frequency transformer?

Can you provide a link of a Class D amps that goes through a Xformer?
The only reason a class D amp is even a little preferred is the driver transistors, it doesn't have much to do with anything else. The sine wave is riding inside the PWM, and the carrier frequency,which is the PWM, is pretty easy to strip out. The transformer itself may do a good job of this.

But lets back up a bit, it occurs to me we have gone way ahead of ourselves in a major sort of way.

I don't recall your specs, just the transformer will be from an old microwave oven. What is the frequency and RMS power the Philippines use? You probably know this, but in the USA it is 60Hz at 120VAC.

Why do you want a sine wave type inverter? They generate less noise, and have some other advantages, but overall there really isn't much difference for general purpose use.

What is the output power you are aiming for in wattage? The transformer will rule on this, as well as the voltage of the circuits that will be feeding it. Actually, the transformer will be the expensive part in this design overall, followed by the drivers themselves (MOSFets are assumed).

As far as a fixed sine wave generator is concerned, the 555 is not the best way to go. Something like a Wien Bridge Oscillator would work better for the base sine wave, I'm designing the 555 version so I can come up with a variable sine wave generator, something that can span 10 or so octaves (an octave is a frequency doubled from the one below it). To make something like this you'll need an oscilloscope, which brings me to the next question.

What kind of test equipment do you have access to? Phase shift sine wave oscillators are super simple, but they need a n oscilloscope to tweak them right. I'm still in the middle of my pc oscope project, but the basic package will work just fine as is.

The reason I'm bringing this up is it may be we are over designing this project, it may be there are simpler ways to do the same thing. I enjoy this kind of work, if you are too then by all means, full steam ahead. An instructor once taught me skull sweat is cheaper than hardware though.

One other thing, are you thinking of feeding a sine wave or triangle wave through a CMOS gate? If so this won't fly, analog and digital don't mix except for specific cases, which this isn't.

Last edited:

#### JUAN DELA CRUZ

Joined May 27, 2008
121
Thank Bill...

I don't recall your specs, just the transformer will be from an old microwave oven. What is the frequency and RMS power the Philippines use? You probably know this, but in the USA it is 60Hz at 120VAC.
...60Hz at 220Vac

Why do you want a sine wave type inverter? They generate less noise, and have some other advantages, but overall there really isn't much difference for general purpose use.
Sine wave type inverter is better than Modified sine right? It can power almost all types of appliances unlike the Modified sine that have trouble in powering motors w/ variable speed. And powering a sensitive load such as PC w/ a Sine wave is essential. Hence , I decided to make a Sine wave inverter instead of a modified one. But my knowledge regarding that type of inverter is limited. That why I'm seeking for some aid.

What is the output power you are aiming for in wattage?
...approx. 500W-1000W

The transformer will rule on this, as well as the voltage of the circuits that will be feeding it. Actually, the transformer will be the expensive part in this design overall, followed by the drivers themselves (MOSFets are assumed).
You mean we shall be using a high frequency type Xformer?
...not a low f one like the Xformer from an old microwave oven?

To make something like this you'll need an oscilloscope
...I don't have my own Oscilloscope

The reason I'm bringing this up is it may be we are over designing this project, it may be there are simpler ways to do the same thing. I enjoy this kind of work, if you are too then by all means, full steam ahead. An instructor once taught me skull sweat is cheaper than hardware though.
I agree, tinkering w/ this stuff is quite tricky yet enjoyable.
I'm with you Bill....full steam ahead.

Last edited:

#### Wendy

Joined Mar 24, 2008
23,134
For what it's worth, I'm still working on the PC Oscope project. Basically I'm at the hole drilling pexiglass cutting stage, you can see my drawings on my blog. I may be wasting my time, it is entirely possible the PC is not capable of any kind of accuracy, but hope springs eternal. Either way I'll post results when I start getting some, and I'll put up experimental results on the Projects Collection, things like display pictures and frequency responses.

The reason I mention this is an Oscope is pretty handy looking at sine wave oscillators, making sure the wave form is not clipping and setting up the peak to peak amplitudes. I have currently loaned my old (ancient actually) HP 120 Oscope out, so can't verify a lot of my so called better ideas.

How about frequency counters? This function is included on many modern DVMs, though it is only 4 digits. Better [URL="http://www.bgmicro.com/index.asp?PageAction=VIEWPROD&ProdID=10386]units[/url] with more digits and functions are available for reasonable prices. Reason I ask is I'm trying to get a feel for what you can do with your home lab. I was a student on a shoe string budget once, nowdays I have a little more money to waste on equipment, this being a form of entertainment for me. Cheaper than going to the movies.

AG is right about one thing, it is entirely possible (probable?) you can spend more money on something like this than if you went out and bought one off the shelf. It is also possible you will save money, or that at some point you will smoke the transformer and give up the project. All are reasonable outcomes for this. I remember a project well over 20 years ago I got in a hurry (and was bone tired), connected the AC transformer backwards, and smoked the entire project (100% of the semiconductors, the transformer, the capacitors, and the fuse). The case was still usable, but I threw it all away. All part of the fun.

You've noticed I like drawing. I really need to sit down and learn to use a SPICE program, it is MUCH cheaper and faster than actually building electronics, and will let you know if you are wasting your time. If you haven't started trying to learn one I recommend it.

High frequency transformers are a different animal than power transformers. Without seeing the context of where you got yours (other than a microwave oven) it is hard to say whether this will work. The important thing is to know the turns ratio. It might be worth throwing together a breadboard using the parts you will be using for the finished inverter to check it out, makeing sure it will put out the voltages you think it will. Something like a simple 555 with a driver output feeding a square wave to this transformer and measureing the output voltage. It'll be a lot harder to make changes to compensate later if it puts out more (or less) AC voltage than you think it will, and it doesn't need to carry a load to do the basic check out. If it turns out to be dead you've saved a lot of time there too.

Are you planning on feeding this inverter with 12 Volts, from a car battery or equivalent?

#### RmACK

Joined Nov 23, 2007
54
Hey everyone,

Just read through this long thread and thought I'd make a few points:

-Yes a sine wave inverter will cost way more to build than buy off the shelf, but like most electronics, if I don't need it so can't justify spending the money on it, I'll spend 5 times as much building it! -There's the hobby element you see.

-An easy way to make a VERY clean sinewave distortion <1% is to use something like the XR-2206 IC from exar. I have used these, they're not fun to set up but do make very clean sine waves

-Computers needing sine waves, um well actually, they rectify the mains and smooth it with capacitors to produce DC to step down to a lower voltage with an isolated buck converter. So in theory they shouldn't care about a MSW. Also 50/60Hz transformers have a high magnetising inductance which tends to smooth the MSW quite a bit.

-You will want a much higher switching frequency than 44kHz for a class D amp. Please show me a filter that will limit the amount of 44 Khz switching noise going through my speaker (full power when idle) but not decrease my 20 kHz audio by more than 3dB! I have got my class D amp up to 333 kHz but I cannot keep my triangle wave symmetrical & linear so have to drop it to 250 kHz. This sounds pretty good. I have a working prototype but it's not finished. Will be put on my website www.rory.co.nz for all to enjoy when done (hopefully by the end of the year) and I intend on making a sine inverter and later UPS with it as a building block by the end of next year.

-BTW, my class d draws 1.2W from 14V when idle & puts out about 14W into 4 ohms. I aim to cut the idle to <0.5W The fets don't get warm and have no heatsink Keep up the good work guys!

#### Wendy

Joined Mar 24, 2008
23,134
Howdy RmACK,

Start another thread for your D amp and I'd love to discuss it. Don't forget schematics and other information.

I've another thread that is languishing a bit, where I'm trying to come up with a decent function generator that doesn't require specialty chips (and uses minimum components).

For the vast majority of appliances inverter waveforms don't matter. Any device that converts the AC into relatively low power DC (which describes computers) just doesn't care about waveform, just RMS. Where it does matter is any gadget that uses the waveform in some fashion, such as a SCR based dimmer circuit.

Last edited:

#### RmACK

Joined Nov 23, 2007
54
Hi Bill,
I've created this thread with my schematics and some info. Hopefully I can get some photos soon. I'll check out your sine forum. As mentioned, I have designed a universal bridge that will first be cranked up in switch frequency for class d and then lowered & have higher current fets substitued for use in a ups. I may generate a 400V bus & use another bridge with high voltage fets for the sine pwm.

Rory

#### JUAN DELA CRUZ

Joined May 27, 2008
121
For what it's worth, I'm still working on the PC Oscope project. Basically I'm at the hole drilling pexiglass cutting stage, you can see my drawings on my blog. I may be wasting my time, it is entirely possible the PC is not capable of any kind of accuracy, but hope springs eternal. Either way I'll post results when I start getting some, and I'll put up experimental results on the Projects Collection, things like display pictures and frequency responses.

The reason I mention this is an Oscope is pretty handy looking at sine wave oscillators, making sure the wave form is not clipping and setting up the peak to peak amplitudes. I have currently loaned my old (ancient actually) HP 120 Oscope out, so can't verify a lot of my so called better ideas.

How about frequency counters? This function is included on many modern DVMs, though it is only 4 digits. Better units with more digits and functions are available for reasonable prices. Reason I ask is I'm trying to get a feel for what you can do with your home lab. I was a student on a shoe string budget once, nowdays I have a little more money to waste on equipment, this being a form of entertainment for me. Cheaper than going to the movies.

AG is right about one thing, it is entirely possible (probable?) you can spend more money on something like this than if you went out and bought one off the shelf. It is also possible you will save money, or that at some point you will smoke the transformer and give up the project. All are reasonable outcomes for this. I remember a project well over 20 years ago I got in a hurry (and was bone tired), connected the AC transformer backwards, and smoked the entire project (100% of the semiconductors, the transformer, the capacitors, and the fuse). The case was still usable, but I threw it all away. All part of the fun.

You've noticed I like drawing. I really need to sit down and learn to use a SPICE program, it is MUCH cheaper and faster than actually building electronics, and will let you know if you are wasting your time. If you haven't started trying to learn one I recommend it.

High frequency transformers are a different animal than power transformers. Without seeing the context of where you got yours (other than a microwave oven) it is hard to say whether this will work. The important thing is to know the turns ratio. It might be worth throwing together a breadboard using the parts you will be using for the finished inverter to check it out, makeing sure it will put out the voltages you think it will. Something like a simple 555 with a driver output feeding a square wave to this transformer and measureing the output voltage. It'll be a lot harder to make changes to compensate later if it puts out more (or less) AC voltage than you think it will, and it doesn't need to carry a load to do the basic check out. If it turns out to be dead you've saved a lot of time there too.

Are you planning on feeding this inverter with 12 Volts, from a car battery or equivalent?
Yes... if possible..

-Yes a sine wave inverter will cost way more to build than buy off the shelf, but like most electronics, if I don't need it so can't justify spending the money on it, I'll spend 5 times as much building it! -There's the hobby element you see.
.....but I realized that tinkering w/ a 'sine-wave inverter' will really cost a lot than buying a new one. I thought there are simpler & cheaper ways to do the same thing w/ my low f Xformer. Unfortunately, I don't have enough fortune (unlike Lucky RmACK) that can be spend to support my hobby. I wish I could but you know.....
- "Living in First World Country like yours...is really different."

I think it is better for me to deal w/ a 'less tricky & cheaper' yet effective kind of inverter...A Simple MSW Inv.

Yesterday 04:38 PMRmACK
-Computers needing sine waves, um well actually, they rectify the mains and smooth it with capacitors to produce DC to step down to a lower voltage with an isolated buck converter. So in theory they shouldn't care about a MSW. Also 50/60Hz transformers have a high magnetising inductance which tends to smooth the MSW quite a bit
Yesterday 07:06 PMBill_Marsden
For the vast majority of appliances inverter waveforms don't matter. Any device that converts the AC into relatively low power DC (which describes computers) just doesn't care about waveform, just RMS. Where it does matter is any gadget that uses the waveform in some fashion, such as a SCR based dimmer circuit.
Once and for all..
....a Sine-wave Inverter is better than MSW Inverter...but indeed a SW Inv. needs a lot of fortune & know-how (I didn't have) to be build by a DIY enthusiast.

Goodluck Pals...

#### Attachments

• 5.1 KB Views: 68
Last edited:

#### Wendy

Joined Mar 24, 2008
23,134
Electronics isn't expensive, but R&D is. Working with tried designs, where you know where you're going, you can set a definite budget. If you aren't sure where you'll end up the costs add up. Looking around the internet I came across two projects from the same person that started with your basic component, here and here. There is nothing wrong with ambitious projects, the subject of inverters keep coming up at AAC over and over, so my feeling is we (this site) need to come up with a selection so that we can point out designs when they're requested.

If you're getting discouraged I suspect you need to move your goals back, you seem like you're interested enough, but it is common for folks new to the hobby to bite off more than they can chew at first. Again, nothing wrong with that, but it is better to work your way up to projects, so you have a better idea what you're getting into.

Way back when I was starting out I made an electric fence circuit. It worked, but it was weak, until I discovered that car batteries were the ticket, they provided the surge of current the circuit needed as opposed to big caps, which didn't. At that time my Dad, who had just retired from the air force, was trying to live in the country, complete with a small herd of 5 cattle. One of them was ornery, and liked to jump fences. Now in the real world any determined bovine can destroy a fence, which is what this one did regularly. Having just finished this project I volunteered it to seal up a section she had just ruined until we could get around to fixing it. She kept trying that single strand of wire, backing off, and looking at it disbelievingly. We got tired of watching her shoulders twitch every time she pushed her way through and were walking to work on the garden when we heard a loud "OMPH!". We ran back in time to see her sitting on her haunches like a dog, which a cow simply doesn't do, and shaking her head. I suspect she wrapped her tongue around the wire between pulses and got a good one. It burned up my fence charger, but it was worth all the effort, she never tried that fence again, and went on to new sections to destroy, eventually we ate her.

My point is even failures can be spectacular successes, and the learning is the thing. At that time I was broke, completely. An old color TV that was scrapped by my Dad (it had a round display) provided years of parts for my future experiments. Sound familiar? My Dad's favorite expression whenever his TV fuzzed out (country life meant weak TV reception back then) was "Turn it off". More often than not it worked, I turned off whatever oscillator I was experimenting with (using a cardboard toilet paper tube many cases) and he got his TV back.

If you're getting discouraged don't give up, just scale back a little. For me one of the fustrating things that can happen is to see a schematic/project on the internet and loose it. It happens, it's one of the reasons I treasure this site, but even then it still happens. I came across a finished project someone pointed to that looked a lot like this...

This design has never been tested, but the other one had. I think the big difference is there were base resistors. Working designs always trump experimental ideas, always. But the other schematic is gone forever, until I see it again (and you can bet it will be bookmarked). I spent several hours looking for it for this thread a couple of days ago.

You'll note a lot of the old hands will point out parts that will do exactly what is required, and their "only" 39 cents. They are cheap, but most companies require a minimum order, so there goes the cost savings (I remember that one well). This is why I try to use off the shelf parts that can be had either at the local stores or in my parts bin (later prefered).

If you want to back up, or try the above design with 1KΩ resistors coupling the transistor buffers, this is how I would go. It would let you know if the transformer will work, if nothing else (a major considertion). Think how you would feel if after building it you found the transformer was the reason the microwave didn't work.

I keep talking about the Oscope project. I found cheap boxes with rotory switches for $3, and it is taking a lot more work using them than if I had bought new. But it it works it will be good test gear for a work bench. If you haven't looked up the basic idea from the AAC book and the download link it is worth it, inexpensive and extremely useful. Just a suggestion, either start with a scaled back inverter (you can always improve it later) or something basic, like a sine wave oscillator. I say this because your last post seemed very discouraged. Last edited: Thread Starter #### JUAN DELA CRUZ Joined May 27, 2008 121 Electronics isn't expensive, but R&D is. Working with tried designs, where you know where you're going, you can set a definite budget. If you aren't sure where you'll end up the costs add up. Looking around the internet I came across two projects from the same person that started with your basic component, here and here. There is nothing wrong with ambitious projects, the subject of inverters keep coming up at AAC over and over, so my feeling is we (this site) need to come up with a selection so that we can point out designs when they're requested. If you're getting discouraged I suspect you need to move your goals back, you seem like you're interested enough, but it is common for folks new to the hobby to bite off more than they can chew at first. Again, nothing wrong with that, but it is better to work your way up to projects, so you have a better idea what you're getting into. Way back when I was starting out I made an electric fence circuit. It worked, but it was weak, until I discovered that car batteries were the ticket, they provided the surge of current the circuit needed as opposed to big caps, which didn't. At that time my Dad, who had just retired from the air force, was trying to live in the country, complete with a small herd of 5 cattle. One of them was ornery, and liked to jump fences. Now in the real world any determined bovine can destroy a fence, which is what this one did regularly. Having just finished this project I volunteered it to seal up a section she had just ruined until we could get around to fixing it. She kept trying that single strand of wire, backing off, and looking at it disbelievingly. We got tired of watching her shoulders twitch every time she pushed her way through and were walking to work on the garden when we heard a loud "OMPH!". We ran back in time to see her sitting on her haunches like a dog, which a cow simply doesn't do, and shaking her head. I suspect she wrapped her tongue around the wire between pulses and got a good one. It burned up my fence charger, but it was worth all the effort, she never tried that fence again, and went on to new sections to destroy, eventually we ate her. My point is even failures can be spectacular successes, and the learning is the thing. At that time I was broke, completely. An old color TV that was scrapped by my Dad (it had a round display) provided years of parts for my future experiments. Sound familiar? My Dad's favorite expression whenever his TV fuzzed out (country life meant weak TV reception back then) was "Turn it off". More often than not it worked, I turned off whatever oscillator I was experimenting with (using a cardboard toilet paper tube many cases) and he got his TV back. If you're getting discouraged don't give up, just scale back a little. For me one of the fustrating things that can happen is to see a schematic/project on the internet and loose it. It happens, it's one of the reasons I treasure this site, but even then it still happens. I came across a finished project someone pointed to that looked a lot like this... This design has never been tested, but the other one had. I think the big difference is there were base resistors. Working designs always trump experimental ideas, always. But the other schematic is gone forever, until I see it again (and you can bet it will be bookmarked). I spent several hours looking for it for this thread a couple of days ago. You'll note a lot of the old hands will point out parts that will do exactly what is required, and their "only" 39 cents. They are cheap, but most companies require a minimum order, so there goes the cost savings (I remember that one well). This is why I try to use off the shelf parts that can be had either at the local stores or in my parts bin (later prefered). If you want to back up, or try the above design with 1KΩ resistors coupling the transistor buffers, this is how I would go. It would let you know if the transformer will work, if nothing else (a major considertion). Think how you would feel if after building it you found the transformer was the reason the microwave didn't work. I keep talking about the Oscope project. I found cheap boxes with rotory switches for$3, and it is taking a lot more work using them than if I had bought new. But it it works it will be good test gear for a work bench. If you haven't looked up the basic idea from the AAC book and the download link it is worth it, inexpensive and extremely useful.

Just a suggestion, either start with a scaled back inverter (you can always improve it later) or something basic, like a sine wave oscillator. I say this because your last post seemed very discouraged.
Ok...now I understand.

"Correction does much, but encouragement does more."

Thank you very much Mr. Bill Marsden for everything.
...I really appreciated it.

P.S.