High Frequency Component in Square Wave questions

Hi,

I learnt that square waves for examples produced by power interver has high frequency components. What's the typical value of the frequency supposed the source is 110 volts,

60 hertz. Is there no way to remove the high frequency components? How does it affect the load?

emma

Reply to
emma
Loading thread data ...

formatting link

formatting link

--
Jim Pennino

Remove .spam.sux to reply.
Reply to
jimp

Yes. That's what makes them square waves. Remove the high frequency components, and you have sine waves.

Theoretically, multiple frequencies, extending to infinity.

Yes, but not if you want to continue to have a square wave.

--
St. John
Reply to
St. John Smythe

Bandpass filter.

--
Uncle Al
http://www.mazepath.com/uncleal/
 Click to see the full signature
Reply to
Uncle Al

Of course. They have to.

There is no such thing as "typical value", it depends how close to suare you want them to be.

Of course. You can filter away all but the fundamental. In which case you're back to a sine wave.

That depends on the load.

Mati Meron | "When you argue with a fool, snipped-for-privacy@cars.uchicago.edu | chances are he is doing just the same"

Reply to
mmeron

That's why radios don't work well on with square wave A.C.

The more expensive invertors let you switch from square wave to sine wave when you want to run something like a radio.

Power efficiency comes at a cost.

formatting link

Reply to
Solarman

Low Pass will do it. But you will need more than a 110 volt square wave to start with if you expect to get a 110 volt sine.

TM

Reply to
Tm

Guys,

I tried building a power inveter using two transistors as push-pull.. meaning if one is one, the other is off, this is connected to

12-0-12 transformer to produce 110 volts.

So as I understand it, when the transistor is in saturation, the current is supposed to form the square wave amplitude. It is not possible to raise it instantaneously that's why there is an increasing effect. Now how does the high frequency component got generated?? Also when it is in the maximum amplitude, how does it generate the high frequency component in the square waves?

What would it take to build a square wave power inverter that totally eliminate the high frequencies riding in the square wave??

Thanks.

emma

Reply to
emma

Try playing around with this Java applet:

formatting link

If your browser doesn't support Java, read these explanations:

formatting link
formatting link

Jerry

Reply to
Jerry

A repeal of some fundamental physical laws. As has already been explained to you, when you remove the high frequencies, you no longer have a square wave.

--
St. John
Reply to
St. John Smythe

On a sunny day (28 Jun 2005 18:37:11 -0700) it happened "emma" wrote in :

Any square wave can be thought of as an infinite amount of sine waves. In practice you will have some harmonics, f, 2f, 3f, 4f etc. Look up Fourier analysis.

Some inverters make a near square wave, some a modulated pulse width that is filtered into an approximate sine wave, some just combine some smaller square waves to make something that looks a bit like a sine. Some make really good sine waves.

120, 180, 240, 300, etc, with varying amplitude.

Yes, LC filter, but you will lose some power.

If you for some reason need a pure sine wave, get a converter that does that.

Many applications / appliances do not need a real good sine wave, as these have their own power processing / power supplies - those could handle square waves even. In fact in some cases square wave could be beneficial.

Reply to
Jan Panteltje

So the high frequency can be removed. I thought they can't. How complicated is the LC filter design to remove 100% of the high frequency component in the square wave?

In pure line Alternating Current (sine wave) from power companies. Is there any high frequency component in the sine wave or is it pure 100% 60 Hz AC??

emma

Reply to
emma

On a sunny day (29 Jun 2005 03:57:12 -0700) it happened "emma" wrote in :

There are 2 ways, one is a low pass filter, the other a resonant system. Low pass is the simplest, it filters out everything above 60Hz in this case. It could be as simple as an inductor (L) in series with a capacitor to ground. But there are a lot of catches.... For this reason it will almost always be simpler to buy a converter that already outputs a -reasonable- sine wave, simpler and cheaper then designing and building one (given some initial likely failures).

First you will have to decide how 'pure' your sine wave should be. How much power from harmonics should be allowed?

Perhaps it would be better if you specified what you need the output for (what sort of equipment).

It is a pure sine wave, but because of various loads it will carry some distortion. For example thyristor controlled devices can cause pulses, some power companies send extra higher frequency signals over it for control, some even for internet these days. A TV with a rectifier will cause flat tops on the sine wave. So do not expect a very clean wave! Neither is the voltage always exact.

Reply to
Jan Panteltje

The output is feed to a resistive coil. I'm analyzing the magnetic field produced by plotting it in 3D.

Suppose I have pure AC from the power lines. How would I add high frequency component to it? Suppose I want to add 300 Hz to ride on the sine wave, any idea how I can do that??

Another. Suppose I feed a function generator to a power amplifier. Would the power amplifier introduce distortions to the signal produced in the function generators? What's the worse distortion that the power amp can introduce to it?

Thanks.

emma

Reply to
emma

On a sunny day (29 Jun 2005 04:55:25 -0700) it happened "emma" wrote in :

Now 'resistive' and 'coil' are 2 things that contradict a bit. Any coil will behave as an inductor (and have some resistance). So you can view it as a R in series with an L. (so Z = R + jwL, were 'w' stands for omega or 2 x pi x frequency). From this you can see the 'impedance' depends on the frequency applied.

You have to decouple the 300 Hz from the mains, and also block the 60 Hz from the driving amp. At these low frequencies that requires large inductors and capacitors, not very practical.

--- L---

60Hz ---- --------------------------- load --- C--- | 300 Hz stop | parr resonant | ----- | | L C | | 60 Hz stop parallel resonant ----- | 300Hz

This is the best system, and you could then take 60 Hz from the mains, attenuate, and simply add 300 Hz from a signal generator, via a resistor network:

60Hz ---- R------------ amplifier ---------- load | 300Hz ----R------ | R | //// ground

We have seen that the 'impedance' of the coil you use is very low at low frequencies (practically near its resistance in Ohms), while at higher frequencies it will have a higher impedance.

The amplifier should be able to drive the low impedance, say if the coil is 4 Ohm, and the amplifier can drive 4 ohm speakers this should be possible. You will have to keep the output amplitude below clipping to keep a sine wave. At the 300Hz, what the impedance is, depends on the inductance of the coil, that is set by the number of turns and if there is some metal core. How much current you can drive into the coil at frequency f is set by the impedance and the peak to peak undistorted output from the amplifier you use. Here is a power amp I use for experiments, it is quite rigid and protected against overload (within reason).

formatting link

Reply to
Jan Panteltje

When I say resistive coil. It's not one load. I used an electric bulb in the output.. then the coil is put in series to it. The coil has reading of 1.7 ohms.

Do you know of a program where I can input the waveforms (i.e. sine or square and its micropulsations) and it can output or plot the magnetic field of the waveforms in 3D? I have to do it manually and need something for comparison.

That means I can't use any 100 ohm lamp in the output of the speakers. I can't find any AC resistor 110 volts at electronic stores. The most they have is a 1 watt resistor for used in circuit. What's the other names for AC resistors? Are their power amplifers that don't have resistance limit in their output?

I won't build the amplifier. I have wasted so much time constructing pcb, circuit, soldering for a function generator and variable power inverter and it's only up to 9 khz. I want something as large as 10 megahertz. Can commercial function gen create very clean sine wave? What's your function gen?

About the power amp. I want to buy a commercially available one. Are older transistor based solid state circuit better or worse compared to modern ic based power amp in preventing distortions in the waveform sent from the function gen input?

What do you call power amps where it is not used for audio only but also for function gen amplification??

I need to get very clean source so I can prevent any contaminations of frequencies in the output.

Many thanks.

emma

Reply to
emma

Is this supposed to mean soething?

As has been said by several others, put a filter on the output and you get a sine.

--
Jim Pennino

Remove .spam.sux to reply.
Reply to
jimp

On a sunny day (29 Jun 2005 07:32:09 -0700) it happened "emma" wrote in :

I am still a bit puzzled to what you are trying to do. Here are my questions: If you say 'magnetic field' what field are you referring to, the one from the coil? So then you have a probe, and measure everywhere near and in the coil the magnetic field, and plot that? I am sure there are examples of solenoid (single layer coils) with field lines on the net, search via google: solenoid field lines I find many many hits, here

formatting link
The field will be symmetrical if you coil is, so 3D = 2D

But you could use a simple trick, add a 3 Ohm resistor is series with the coil, and use an amplifier that can drive 4 Ohm. This way the amp will not be overloaded.

amplifier --- 3 Ohm -- 1 Ohm coil -- ground

Resistors behave the same for AC and DC (except from very high frequencies, many MHz, where they act like an inductor depending upon their construction).

I designed my own. Now I use an FPGA (programmable gate array) with a fast DA converter. You calculate the waveform for say 1024 points, upload to the (very fast) FPGA internal RAM, and then program the FPGA so it outputs at the right speed. Or just program the required pulse sequence (in Verilog). You are limited by the clock speed divided by the number of samples you want.

64 samples per period at 10MHz output would require a 640 MHz clock! Mine does not go that fast (and and neither the DA). And 64 samples is not exactly a 'pure' sine wave. If you just want the sine waves, use any LC signal RF generator. If you want only frequencies in the audio range, use the PC sound card, and a good sound editor. In the past I have made waveforms by writing a simple BASIC program, calculate for example some values, and write it to a raw file (binary sound file). Then most sound editors will be able to make a wave file.

If all else fails get a good RF generator from ebay? While you are at it get an oscilloscope, and all this is difficult to do without being able to see the waveforms (even if just to check nothing clips). A 10MHz scope should be dirt cheap on ebay.

If you say 'I want 10MHz' then you have a problem. Please also state the amount of power you want (at that frequency). The other thing is how will you detect the magnetic field at 10MHz? You can detect LF fields very good with an old playback head, but not in the MHz range.

I would call a 10MHz power amp a power amplifier with 10MHz bandwidth....

One way of doing that is using an amplifier with as narrow a bandwidth as possible, for example if you want 10kHz, a 10MHz wide amp will only add unwanted stuff!

I am still not 100% clear on what you want to do, how will you measure the magnetic field?

Reply to
Jan Panteltje

..and why experiment at 110V when say 9V would be a lot safer?

Reply to
CWatters

You can get an estimate of the high frequency component from the rise and fall times of the "square" wave. A low pass LC filter will reduce the high frequencies. Expensive.

[Old Man]

Reply to
Old Man

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.