Inverter circuits problem

[StraightView]

Hi everyone, A while ago I have made these inverter circuits in the pictures below but they didn't work no matter what I did so I stopped working on them, but now I'm I returned to try made them work and I faced the same problem like in the first time which means that when I measure the transformer output there is voltage shown in the meter and even high voltage sometimes but it can't even power up a load of 12v like a 12v relay, also when I connect the 12v relay to the transformer output the voltage goes to 0 whiteout powering the relay!!, I'm kind of sure that the switching is working fine, I can hear the noises made by the transformer that indicates that the switching is working if I'm not wrong about this information.
I have tested many, many ways without success to solve the problem, I'm sure that the transformers winding is correct and I have tested many transformers, I tested many NPN power transistors, many mosfets, I changed the resistors values, I tried anything comes to my mind.
I have thought that this problem is because of low current at the output so I tried to change the coils values many, many times, I tried with increasing the turns numbers of the secondary many times, decreasing the primary turns many times, I changed the wire diameter of both primary and secondary many times, I have made the secondary wire diameter larger for increasing current, all these tries and the problem insisted.

I thank and appreciate anyone help to solve this problem, thank you everyone.

 

[betwixt]

Relays generally work on DC or low frequency AC. All but the multivibrator (3rd from bottom) circuits will oscillate at a frequency entirely decided by the inductance of the transformer and could be tens or thousands of KHz.

None of the circuits are particularly good or have any kind of frequency or amplitude stabilization so don't expect good or reliable results.

The transformer winding direction is vitally important. In the three lower schematics, the transformer primary is like one continuous winding in the same direction with a center tap. In the hand drawn ones if the winding feeding the base of the transistor is reversed it will not oscillate, its the direction of the winding that provides the needed phase inversion.

The 3rd from bottom has no dots to indicate connections. If you build it, be sure to connect all the junctions except the ones crossing over diagonally between transistor base pins.

Brian.

 

[KlausST]

Hi,

5Mbytes of data for a couple of lines. And a lot of text but almost no information we can work on.

Thus I can just give general advice from my experience.
When I started designin electronics I soon gave up on the trial and error method. It wwas just a waste of time, a lot of smoke, damged devices and mainly disappointment.

Thus i changed my way and studied reference ciruits, learned how they are meant to work. How, why and the calculations behind it.

Also with switching power applications the PCB layout is very important. You may have absolutoey the same schematic, use the same devices but still one PCB solution may work clean, reliably, generating low heat, whil the othere does nothing useful but generating heat .. then kill itself.

Almost any semiconductor manufacturer provides very detailed documents, tutorials, examples... some even videos.

Like Brian said the shown schematics are incomplete, don´t tell the math, don´t tell what´s important about the transformer. Don´t tell about PCB...

I can only recommend to read some of them. But for sure you are free to go your own. way.

Klaus

 

[BradtheRad]

I attempted to build my own power inverter using a transformer in reverse. It was a center-tap type similar to your last image. I scavenged the transformer from an appliance where it powered several circuits. I thought I'd get ample 120 VAC from it.
However all I could manage was to light a 7W bulb dimly.

Each time I increased power, it generated severe spikes which fried my mosfets. I had to give up. That was many years ago when I barely knew what I was doing.

Since then I realized power transformers are designed to be at the load end of the electric company. They see the primary side wound from lots of turns of thin wire. Its Henry value is combined with an ohmic value. It only lets through a certain amount of Amperes.
By running the transformer backwards, we're asking lots of turns of thin wire to conduct in series with our load. Electron flow is hampered.

 

[KlausST]

Hi,

I used power transformers in reverse.
They follow the ruels of physics in both directions. The surely work in both directions.
But hey are no ideal transformers, they are real transformers. And thus theirw windings and winding count is designed to adjust on reality. Thus a transformer designed for 230V-->12V (example) can not (perfectly) work as a transformer for 12V --> 230V.

Example:
So let´s say it is a transformer with 14% winding loss (internet info of random 120VA transformer), spread about equally on primary and secondary. (at full lod)
And let´s coupling factor be 100% (unrealistic)
7% of 230V is 16.1V, so only 213.9V get to the "ideal" transfomer primary side.
7% on the 12V side means 0.84V. So the secondary winding need to generate 12.84V.
So the winding count n1:n2 equals 213.9V/12.84V.
This also eplains the higher non_load voltage: in this case it is: 230V * 12.84V / 213.9V = 13.81V
The output power is 120W, so the secondary current is 10A.
The secondary side winding resistance is 0.84V/10A = 84mOhms
Secondary side power loss is 8.4W
primary winding current is 10A * 12.84V / 213.9V = 0.60A.
Primary side winding resistance is 16.1V / 0.60A = 26.83 Ohms (you see this is a lot more than the secondary ones)
primary side power loss is 9.66W
Total input power is 120W + 8.4W + 9.66W = 138.1W

Now if you use the same transformer in reverse with 12V input.
The secondray voltage loss again is 0.84V, thus only 11.16V get tranformed.
to 11.16V * 213.9V / 12.84V = 185.9V
minus the loss of 16.1V gives just 169.8V. At full load.
Ooops, this is 35% less than expected!
To compensate for this you need to input 16.25V. But this won´t work due to core saturation.
Causing overly high current peaks --> destroying the semiconductors.

Klaus

 

[StraightView]

Relays generally work on DC or low frequency AC. All but the multivibrator (3rd from bottom) circuits will oscillate at a frequency entirely decided by the inductance of the transformer and could be tens or thousands of KHz.

None of the circuits are particularly good or have any kind of frequency or amplitude stabilization so don't expect good or reliable results.

The transformer winding direction is vitally important. In the three lower schematics, the transformer primary is like one continuous winding in the same direction with a center tap. In the hand drawn ones if the winding feeding the base of the transistor is reversed it will not oscillate, its the direction of the winding that provides the needed phase inversion.

The 3rd from bottom has no dots to indicate connections. If you build it, be sure to connect all the junctions except the ones crossing over diagonally between transistor base pins.

Brian.

Hi Brian nice talking to you, yes like you said the relay is DC but I have rectified the output with no result, also I have made all these circuits and I got the same problem. I'm careful about the winding directions, I have made the winding of two connected primary coils in the same direction, also in the hand drawing circuits I tested all ways means that I have tested the three pins of the transistor with all three pins of transformer primary, and like I have said there's noises made by transformer it shows that oscillation is working isn't?!.
The problem as I mentioned is that there's voltage shown in the meter sometimes even high voltage such as more than 100v but like I said it can't power up any device of 12v, it can power up LEDs.
Notice, the hand drawing circuits I have seen some videos which people making this circuit from the beginning and worked fine according to the videos, powering up much higher voltage devices such as a drill......., but when I made the circuit exactly as in those videos it didn't work!!, so unless if those videos are a scam??!!.
Thanks Brian.

--- Updated Aug 13, 2023 ---

Hi,

5Mbytes of data for a couple of lines. And a lot of text but almost no information we can work on.

Thus I can just give general advice from my experience.
When I started designin electronics I soon gave up on the trial and error method. It wwas just a waste of time, a lot of smoke, damged devices and mainly disappointment.

Thus i changed my way and studied reference ciruits, learned how they are meant to work. How, why and the calculations behind it.

Also with switching power applications the PCB layout is very important. You may have absolutoey the same schematic, use the same devices but still one PCB solution may work clean, reliably, generating low heat, whil the othere does nothing useful but generating heat .. then kill itself.

Almost any semiconductor manufacturer provides very detailed documents, tutorials, examples... some even videos.

Like Brian said the shown schematics are incomplete, don´t tell the math, don´t tell what´s important about the transformer. Don´t tell about PCB...

I can only recommend to read some of them. But for sure you are free to go your own. way.

Klaus

Hi Klaus, in my city there's no PCB

Hi,

5Mbytes of data for a couple of lines. And a lot of text but almost no information we can work on.

Thus I can just give general advice from my experience.
When I started designin electronics I soon gave up on the trial and error method. It wwas just a waste of time, a lot of smoke, damged devices and mainly disappointment.

Thus i changed my way and studied reference ciruits, learned how they are meant to work. How, why and the calculations behind it.

Also with switching power applications the PCB layout is very important. You may have absolutoey the same schematic, use the same devices but still one PCB solution may work clean, reliably, generating low heat, whil the othere does nothing useful but generating heat .. then kill itself.

Almost any semiconductor manufacturer provides very detailed documents, tutorials, examples... some even videos.

Like Brian said the shown schematics are incomplete, don´t tell the math, don´t tell what´s important about the transformer. Don´t tell about PCB...

I can only recommend to read some of them. But for sure you are free to go your own. way.

Klaus

Hi Klaus,
In my city there are no PCD for sale, brother, and not because my city is a backward village devoid of technology, but rather because the vast majority of people are not interested in some fields and tools, such as electronics, you will not find people interested in building circuits, I personally know many, many people, from family to many close ones, many, many friends, but I do not and did not know anyone in my life who builds or is interested in building electronic circuits not for the sake of a hobby or even for commercial reasons, in my city people only go and buy electronic devices when they need them, there are stores who fix electronic devices but even those stores they don't need PCB because when you fix electronic devices you just remove the damaged components with another new ones.

I like building electronics it's my hobby, so I'm trying to work with what we have or give up, but I can't give up because I like it like I said.

--- Updated Aug 13, 2023 ---

I attempted to build my own power inverter using a transformer in reverse. It was a center-tap type similar to your last image. I scavenged the transformer from an appliance where it powered several circuits. I thought I'd get ample 120 VAC from it.
However all I could manage was to light a 7W bulb dimly.

Each time I increased power, it generated severe spikes which fried my mosfets. I had to give up. That was many years ago when I barely knew what I was doing.

Since then I realized power transformers are designed to be at the load end of the electric company. They see the primary side wound from lots of turns of thin wire. Its Henry value is combined with an ohmic value. It only lets through a certain amount of Amperes.
By running the transformer backwards, we're asking lots of turns of thin wire to conduct in series with our load. Electron flow is hampered.

View attachment 184370

Hi, thanks for your reply.
What confuses me is that there is many videos in YouTube which people makes these circuits from the beginning to the end and it seems work fine, powering up even devices of 220v from 12vdc power source!!! So I'm not sure if those videos are just a scam or I'm missing something.

 

[BradtheRad]

What confuses me is that there is many videos in YouTube which people makes these circuits from the beginning to the end and it seems work fine, powering up even devices of 220v from 12vdc power source!!! So I'm not sure if those videos are just a scam or I'm missing something.

I built low-power oscillators where a transformer provides the up-and-down triggers for oscillations. (I've seen this topology called blocking oscillators.) The circuits are in the instruction manual of old-time Radio Shack electronic lab kits. I discovered component values are finicky, otherwise nothing happens.

The mystery is in the construction of inductors and transformers. It's not obvious what's wrong if the circuit does nothing.

Supply voltage is also finicky. I applied variable voltages from my power supply. Often there was a point when I turned up voltage where oscillations stopped. It's successful only within a certain range of voltage.

It's not obvious what's wrong when the circuit does nothing. Do I need to change a resistor value? The transformer ratio? Primary inductance?

When an experimenter puts up a Youtube video, he probably had to troubleshoot many problems, until his project worked. He only needs to film the steps that worked.That's when it looks easy.

If you ask me I think those guys have lots of new batteries to put in handheld projects. I suspect their power inverter projects require that they wind many transformers, to learn the right way to make one so it carries sufficient power.

 

[Easy peasy]

In the 1st ckts above - there is no reference to phasing - this could be your big issue

 

[StraightView]

I built low-power oscillators where a transformer provides the up-and-down triggers for oscillations. (I've seen this topology called blocking oscillators.) The circuits are in the instruction manual of old-time Radio Shack electronic lab kits. I discovered component values are finicky, otherwise nothing happens.

The mystery is in the construction of inductors and transformers. It's not obvious what's wrong if the circuit does nothing.

Supply voltage is also finicky. I applied variable voltages from my power supply. Often there was a point when I turned up voltage where oscillations stopped. It's successful only within a certain range of voltage.

It's not obvious what's wrong when the circuit does nothing. Do I need to change a resistor value? The transformer ratio? Primary inductance?

When an experimenter puts up a Youtube video, he probably had to troubleshoot many problems, until his project worked. He only needs to film the steps that worked.That's when it looks easy.

If you ask me I think those guys have lots of new batteries to put in handheld projects. I suspect their power inverter projects require that they wind many transformers, to learn the right way to make one so it carries sufficient power.

(If you ask me I think those guys have lots of new batteries to put in handheld projects. I suspect their power inverter projects require that they wind many transformers, to learn the right way to make one so it carries sufficient power.)

This make sense.
Thanks.

--- Updated Aug 14, 2023 ---

In the 1st ckts above - there is no reference to phasing - this could be your big issue

The phasing I have made is the same in the last picture where there is dots, the dots in my transformer is the beginning of the winding.

 

[betwixt]

Adding to Brad's words, many videos on You Tube are complete fakes. They may look real but often have hidden wiring or use light bulb loads that look to be mains powered but have flashlight bulbs inside them.

Incidentally, if your oscillations are higher than about 1KHz a normal rectifier diode will not work so measuring the recovered DC will give a wrong reading. You should drop the frequency if using an iron cored transformer or use a fast recovery diode for higher frequencies.

Brian.

 

[Easy peasy]

Only one picture shows actual dots ( phasing ) adjacent to the windings - I greatly fear your implementation misses this important point - there are no phasing marks ( dots ) on the very first images you show in post #1, if you don't get the phasing correct - the thing will never work.

 

[StraightView]

Adding to Brad's words, many videos on You Tube are complete fakes. They may look real but often have hidden wiring or use light bulb loads that look to be mains powered but have flashlight bulbs inside them.

Incidentally, if your oscillations are higher than about 1KHz a normal rectifier diode will not work so measuring the recovered DC will give a wrong reading. You should drop the frequency if using an iron cored transformer or use a fast recovery diode for higher frequencies.

Brian.

yes I'm using iron core transformer and for the rectifier I'm using those big rectifiers with 4 pins.
I have suspected that those videos could be fake, but we know that there's inverter circuits in real life in many devices so Brian I would like to ask you about which type are those inverter circuits that has been used in many devices?! Also if we made the transistors's/mosfets's driven by a phase shift oscillator or tuned oscillator can solve this problem?!.

--- Updated Aug 14, 2023 ---

Only one picture shows actual dots ( phasing ) adjacent to the windings - I greatly fear your implementation misses this important point - there are no phasing marks ( dots ) on the very first images you show in post #1, if you don't get the phasing correct - the thing will never work.

Yes I agree, I'm aware about the phasing, the video below of one of transformers I have made.
The primary windings are those with 3 wires obviously, the wire in the right is the beginning of the first coil and it ends in the middle, the second coil beggins in the middle and ends in the lift wire, and both coils are in the same direction, I have tested all wires which means I have tested the power source positive with the three wires and the same thing about the negative and the transistor/mosfet base/gate, I tested all mays which means that I'm sure that is not the phasing that cause this problem.

 

[betwixt]

The bigger bridge rectifiers like you describe are no good above about 1KHz, in fact they appear as a load on the transformer but produce low DC output.

For the sake of a few components it is far safer to make an oscillator at a known frequency (50Hz or 60Hz) then use its output to drive a transformer. It avoids all the frequency problems and also offers the possibility of adding dead time to the waveform driving the transformer. Dead time in this context is a short delay between switching one driver transistor off and the other one on. It avoids a potential problem that happens if both driver transistors are partially conducting at the same time, one in the turning off phase and the other turning on. If they both conduct at the same time they short out the primary and absorb the power the other transistor is pushing to the transformer.

Brian.

 

[StraightView]

The bigger bridge rectifiers like you describe are no good above about 1KHz, in fact they appear as a load on the transformer but produce low DC output.

For the sake of a few components it is far safer to make an oscillator at a known frequency (50Hz or 60Hz) then use its output to drive a transformer. It avoids all the frequency problems and also offers the possibility of adding dead time to the waveform driving the transformer. Dead time in this context is a short delay between switching one driver transistor off and the other one on. It avoids a potential problem that happens if both driver transistors are partially conducting at the same time, one in the turning off phase and the other turning on. If they both conduct at the same time they short out the primary and absorb the power the other transistor is pushing to the transformer.

Brian.

So, I was wrong about those rectifiers I mentioned, I have fast recovery diodes I will used them.

I believe that what you said it can solve the problem such as it could be that the transistors are conducting in the same time, and about the oscillation frequency to drive the transformer, in fact I have read before that these inverter frequency to drive the transformer is 50/60Hz but I can't fully remember, also I believe that the type of the power source plays a huge role for these circuits too, by the way, I was using smps circuit as power source.
but it's waiting me a lot of hard work because I have to figure out how to build an oscillator with known frequency because I don't have a oscilloscope, I always relayed on testing and guessing. I think that astable multivebrator won't work to drive the transformer because it's already in the third picture which I have build it many times and didn't work.
I have two circuits I have made a while ago, a phase shift oscillator and tuned collector oscillator, they work fine but I have no idea of their frequency, so I will try to work relaying on testing and guessing and hope to figure out something, I will let you know guys if there was any positive result.
Thank you all for your helps.

 

[StraightView]

The bigger bridge rectifiers like you describe are no good above about 1KHz, in fact they appear as a load on the transformer but produce low DC output.

For the sake of a few components it is far safer to make an oscillator at a known frequency (50Hz or 60Hz) then use its output to drive a transformer. It avoids all the frequency problems and also offers the possibility of adding dead time to the waveform driving the transformer. Dead time in this context is a short delay between switching one driver transistor off and the other one on. It avoids a potential problem that happens if both driver transistors are partially conducting at the same time, one in the turning off phase and the other turning on. If they both conduct at the same time they short out the primary and absorb the power the other transistor is pushing to the transformer.

Brian.

Hi Brian, finally, this is the first time I got real power from this inverter the (video/picture below), like I have said, a while ago I have tried many, many times to make these circuits work, I tried everything but the problem was that the meter shows voltage even high voltage in the output but it can't power up even a 12v load which means I didn't get even the input voltage.

but now finally I got some real power, I was about to make an oscillator circuit to drive the transistors but I thought to try to test the circuit using fast recovery diodes like you said and also to change the power source, I was using smps circuit so I changed it with 220v/9v transformer and finally the circuit worked, the meter shows 53vdc after I rectified the output, I tested it with two relays in the same time, one is 12vdc and the other is 24vdc and both are working fine, but I believe that either it still not working properly because according to step up transformer's values it should generate high voltage/current or I need to modify more the transformer values or the circuit itself?!, I'm not sure.

thank you very much Brian you were very helpful and I like you brother, and thanks to everybody tried to help.
I will keep working on it because I need more power, because I'm planning to use this inverter to run a battery Drill, if you have any advice about it, the Drill is 14v but there's no mention of its ampers, but according to my testing it needs much power.

 

[BradtheRad]

It looks as though you found the right track by trying a more reliable setup.

My own attempt at making a power inverter had all the things to make it easy even though it wasn't successful

* I used a car battery.
* I took square waves from my waveform generator with frequency counter.
* I could see what was happening on my Tektronix oscilloscope.

It's appealing to combine oscillator in the step-up transformer. But that requires adjusting everything just right. Even in simulation it requires adjusting everything just right.

There are inexpensive projects which duplicate instruments commonly used in electronics. Here's a frequency counter made from two 4017 IC's. LED's form a tens column (left) and a ones column (right). Apply incoming pulses to the clock pin (bottom right). Every second reset all IC's with a brief signal. Although the lit led's move upward rapidly, a quick eye can recognize which ones are lit just before each reset signal. My simulation has a waveform of 30 Hz. Thus the leds indicate 30 just as reset is about to occur.

If you wish to count frequencies higher than 99, add a hundreds column by adding another 4017. Etc.

--- Updated Aug 17, 2023 ---

By clicking this link,

https://tinyurl.com/2yq2qn48

1) It opens website falstad.com/circuit
2) Loads my schematic in the simulator
3) Runs it on your computer

 

[betwixt]

Well done for getting it work so far.
You still really need a fixed oscillator to make sure the transformer can work efficiently. I would suggest using a 555 timer with a variable resistor to set the frequency. 555s should be able to provide enough base current to drive one of the transistors, the other will then follow. Ideally you drive both transistors, it will give better symmetry to the waveform.

Consider that if you try to run the output into something frequency sensitive, like an other transformer or a motor, any deviation from their designed frequency will result in less power or in the case of a motor, it might run at the wrong speed.

Brian.

 

[StraightView]

It looks as though you found the right track by trying a more reliable setup.

My own attempt at making a power inverter had all the things to make it easy even though it wasn't successful

* I used a car battery.
* I took square waves from my waveform generator with frequency counter.
* I could see what was happening on my Tektronix oscilloscope.

It's appealing to combine oscillator in the step-up transformer. But that requires adjusting everything just right. Even in simulation it requires adjusting everything just right.

There are inexpensive projects which duplicate instruments commonly used in electronics. Here's a frequency counter made from two 4017 IC's. LED's form a tens column (left) and a ones column (right). Apply incoming pulses to the clock pin (bottom right). Every second reset all IC's with a brief signal. Although the lit led's move upward rapidly, a quick eye can recognize which ones are lit just before each reset signal. My simulation has a waveform of 30 Hz. Thus the leds indicate 30 just as reset is about to occur.

If you wish to count frequencies higher than 99, add a hundreds column by adding another 4017. Etc.

--- Updated Aug 17, 2023 ---

View attachment 184439

By clicking this link,

https://tinyurl.com/2yq2qn48

1) It opens website falstad.com/circuit
2) Loads my schematic in the simulator
3) Runs it on your computer

thank you. You mean that I can upload only schematic of oscillation circuits and the simulator circuit tell me what is the frequency?, another question, does this website work with a smart phone?.
You know, not knowing the frequency of an oscillator circuit is a big problem.

--- Updated Aug 17, 2023 ---

Well done for getting it work so far.
You still really need a fixed oscillator to make sure the transformer can work efficiently. I would suggest using a 555 timer with a variable resistor to set the frequency. 555s should be able to provide enough base current to drive one of the transistors, the other will then follow. Ideally you drive both transistors, it will give better symmetry to the waveform.

Consider that if you try to run the output into something frequency sensitive, like an other transformer or a motor, any deviation from their designed frequency will result in less power or in the case of a motor, it might run at the wrong speed.

Brian.

Yes, it seems that the circuit need more work, but I don't think that they sell 555 timer here in my city, unless if I use some old similar IC taken from old devices's boards I have, I have many of them I believe that some of them are similar to 555 timer such those in companies made smps, maybe.
Also, I don't know why I don't like using ships!, even knowing that they will help a lot and gives a lot if benefits, if I used them I feel like I lost the appetite of making circuits, but I will do more work and if the problems insisted I will try find an IC from those old I have similar to 555 timer and use it or use a crystal one I have several of them if they are compatible for these circuits?.

 

[BradtheRad]

thank you. You mean that I can upload only schematic of oscillation circuits and the simulator circuit tell me what is the frequency?, another question, does this website work with a smart phone?.

Yes, I saw it run on my iPhone. Everything is tiny of course. My fingertips are awkward as I try to tap where I want.

Command scope traces by tapping on a component and choosing New Scope trace. Scope traces travel across the screen. Frequencies, duty cycles, appear after a few cycles. By tapping on a scope trace it opens options for you to choose.

 

[StraightView]

Yes, I saw it run on my iPhone. Everything is tiny of course. My fingertips are awkward as I try to tap where I want.

Command scope traces by tapping on a component and choosing New Scope trace. Scope traces travel across the screen. Frequencies, duty cycles, appear after a few cycles. By tapping on a scope trace it opens options for you to choose.

OK, thanks, I will try it later when I work on oscillator.