Inverters and Transformers

useful rply,my frnd

GA Del,

Likely the reactance of the transformer is what converts the square like voltage applied to the transformer to become rounded into a sign wave on the output. This is a classic demonstration that a transformer is a really a bi-directional device and what one considers an input can be actually an output and vice-versa.

GA del, short and clear...

Maybe this very simplified diagram will help you to explain how the voltage is reversed by the two transistors in an inverter (on at least on ones I have previously built as well as the high voltage circuits that I made in the 70's and 80's when I built my own electronic ignitions....). thus giving the two opposite half cycle output on the mains side.

This mechanical version would actually work if you had a 50 or 60 Hz finger....!!

If you want to try it out though as I have drawn it, do insert a low value resistor in each leg to stop a complete short circuit! 2 x 10 Ohm 1 watt would do it fine...

An oscilloscope would show the actual output voltage reached better as your fingers would be a mite slow!

Note for US countries a 2x12/110 volt transformer, for 230 volt countries a 2x12/230 volt transformer is require if you decide to actually build a proper circuit later..

An NE555 or similar can be used to provide a 50 or 60 Hz signal for transistor switching....

I forgot the diagram, sorry!

I hope the circuit includes something (eg diodes) to take the current as the magnetic field collapses at each switching of the 555, otherwise voltage spikes will likely wreck the darlingtons.

No,there is no mean to take current.there is only darlington and 555.can u pls elobarate tat thing,i cant get u.

Mostly Darlingtons have CLAMPING Diode across C & E to protect the transistor from back-EMF generated when transistor switches off.

To my mind, that is fully correct and you could say almost the "standard" with most MOSFETs in use today....Well put!!

Many forget that small but very important point (not that I am a MOSFET expert, I am not, but I do use them from time to time.....)

Regards.

As the PowerMosFETs are the major device in powerElectronics today...

Thanks

See Post #6 :

In stead of Switches think the Transistors [In your text Darlington in Push-Pull]

When the blue is conducting the Red one is OFF & vice-versa one after the other.

A GA foe #6

The 555 is probably set up as a multivibrator (a square-wave output oscillator).

It drives the darlington pairs on alternate polarities of the square wave, so that the two "13" windings of the transformer are connected from "+" to "ground" in alternate cycles. This results in power being transferred from the two windings to the 230V "primary", which is where the load would be connected. This results in a "crude" step up converter (or inverter), since the waveform at the load will not be a sine wave, but will be a square wave, modified by the response characteristics of the darlingtons and the transformer. This will probably also not be very efficient, as the transformer is usually not very good at responding to square waves (they contain a lot of higher harmonics which the transformer tends to distort and suppress). Also, it looks like the transformer is probably a steel core step down transformer, usually used for creating a power supply which runs off the 230V mains and steps that down to 13-0-13 volts AC at 50-60 Hertz. These are typically not designed for great efficiency, but for lower cost etc.

I slightly (only!) dissagree. You pointed out (correctly) that transformers don't like (allow?) square waves much. I agree with you.

I put it to you that the output of the transformer will be mainly sinusoidal due to the effects you yourself described....

I believe technically its called a "Modified" Sine Wave by such companies that make inverters.....

You are right. The harmonics are not completely suppressed by the transformers poor frequency response and inductive reactance, so the output can't be a sine wave, but the fundamental is by far the strongest wave, so the output is "sorta" like a sine wave.

My explanation is not well written; the main point I was trying to make is that this is an inexpensive way to use a transformer not meant for inverter use, as the step up component in an inexpensive, simple inverter. I actually had an inverter something like this years ago; it was a Heathkit I built to use in my van on long trips so my kids could watch VHS video tapes (this was long before such things were common in vehicles, as they are today). Most of my friends were impressed by my ingenuity! Of course, as a degreed engineer, I was embarrassed not to have just designed and built one myself, however, the Heathkit was so cheap and all the mechanical stuff like the heat sinks, enclosure, etc., are things which take too much time for my lifestyle, so I was glad to find the kit and use it.

The distortion of the output waveform which made it not a true sine wave didn't seem to bother the devices plugged into it. The output transistors were Germanium power transistors (lower voltage drops in the junctions compared to silicon transistors and thus more suited for low voltage operation, resulting in better efficiency with little added cost), which were a bit hard to find when they failed on one trip due to overloading the inverter; it had little output protection. Overall, though, it worked fine, and was many times less expensive than anything else I could find. Now, I have a couple of inverters for my car which cost no more than $30 and can put out a couple hundred watts of fairly clean 110V 60 Hz "modified sine wave" power with input and output protection, fuses everywhere, very attractive and effective extruded aluminum heat sinks as part of the enclosure, audible alert for problem conditions, etc. I am still amazed by the efficiencies attained by the electronics industry over the past several decades!

One other thing: the "modified sine wave" does result in a bit more heating of the transformer and output transistors, though I have never bothered to figure out why. I did a test way back when, by using a pure sine wave to drive the output stages, and measured the output power, temperatures, etc. I then ran the same tests with the oscillator in the inverter itself. The output was about the same, perhaps a bit less, but the temperatures were higher (don't remember by how much). Could be a lot of reasons for this, if it is important to know, but I haven't needed to do that. All the switchers I have worked on used toroidal transformers, FET drivers, fast diodes, etc., and actually TRIED to use square waves as much as possible, for efficiency, rather than sine waves. The FETs run much cooler and thus don't have to be as high powered if they can switch rapidly between "on" and "off" states, such as when driven by a square wave, while they dissipate quite a bit of power, generate more heat, and are less efficient, if a sine wave is used; I am sure you probably know why or, someone here will explain why (my reply is getting too long and boring to go into that!).

Larry

I was embarrassed not to have just designed and built one myself.

Ah, there speaks a true engineer...nice post
Del

Excellent "riposte" if I may say so......well written too.

The extra heating in the transistors when driving them with a sinusoidal input you have explained well.

The heating in the transformer with a sine wave is at its minimum, but when driven rapidly with square waves, the heating is at its highest for any particular load.

I guess its just the "forcing" by the square wave causing a greater heating effect in the laminations of the core. I am sure that the usage of a toroidal transformer will both reduce the heating effect (not completely though....) and slightly increase the efficiency overall, also reducing electro smog near to the inverter as well as well as any possible mechanical noise that some more conventional transformers can make.....

I personally feel that when confronted with a problem (build, buy or design and build in this case), then you found the most cost/time effective solution. I do not feel that you need to reproach yourself for being efficient at saving both costs and time!!!

The heating effects with square waves or so called "modified sine" waves attributable to the higher frequencies in the waveform. This produces more losses in the core laminations.

If you put a square wave in you will get a square out with relatively little distortion. You get some "ringing" at the leading edges but otherwise pretty much the same shape.

Modified sine usually refers to a waveform that is switched hard on or off but to a voltage approximating the Vpeak of a sinewave and being off for a period of time before going negative. The pulse width is adjusted to keep the RMS voltage correct. This simple trick results in a surprisingly low THD. I cannot remember the actual figure but 5% comes to mind, it may even be less.

Modern sine wave inverters are also class D designs, the mosfets are either on or off. They achieve their sinusoidal waveform by HF (typically40kHz) pwm. Their efficiency is only marginally below the low frequency units which is more than made up for by the better efficiency at the load

Regards

Chas

"If you put a square wave in you will get a square out with relatively little distortion. You get some "ringing" at the leading edges but otherwise pretty much the same shape."

I must not be understanding your statement. Are you saying that the transformer will pass a square wave from winding to winding with little distortion? Does that mean a transformer, such as is used in older audio amps, for instance, has little distortion, at a wide range of frequencies? If so, I am confused, as I have always experienced quite a bit of difficulty getting transformers to pass square waves, triangle waves, and others with a wide band of harmonics needed to form the wave shape, due to THD, among other things. This is an old post but I am interested in what you meant here. BTW, I am an engineer (BSEE) so you can talk technical up to that level; at PhD levels, though, I lack some depth in higher math!