Question about Transformers and pulsed AC and DC.

Discussion in 'General Electronics Discussion' started by Zip1234, Jun 13, 2013.

1. Zip1234

12
0
Jun 13, 2013
I have a question about Transformers and pulsed AC and DC.

If you run pulsed AC or pulsed DC(it is PW modulated) signal though a transmission transformer, will it saturate the transformer if the pulses are less than 20 nanoseconds in duration.

Is there a way to send pulsed AC through a transformer without saturating the magnet core?

2. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,496
2,837
Jan 21, 2010
Well, talking about pulsed AC doesn't make a lot of sense. It really becomes a complex AC signal.

Pulsed DC is somewhat simpler. It will pass through a transformer subject to a high pass filtering effect and with a change so that the output signal is symmetrical about ground (so a change in DC offset).

As to whether 20ns pulses will saturate the transformer, that depends on a number of factors. You also need to have a core that can cope with frequencies this high (since a fast rising edge has components of frequencies that extend a long way (theoretically to infinity, but practically maybe a couple of orders of magnitude)

3. Zip1234

12
0
Jun 13, 2013
The core is a standard powerline transformer core.

The signal band is 50-60Hz that is pulsed at 1 billion pulses per second.

Will that saturate the core?

4. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,496
2,837
Jan 21, 2010
A billion pulses per second?

What are you smoking?

The pulses would be well under 20ns -- in fact they would have to be less than 1ns, 500ps if the duty cycle were 50%.

And at that frequency the core isn't going to work like a core at all...

5. Zip1234

12
0
Jun 13, 2013
Yes. one billion pulses per second.
I don't smoke anything.
I am talking about a picosecond pulse modulated signal.
The frequency is still 50-60Hz but is pulsed at 1 billion pulses per second.

What will the core act like?

I am trying to send a modulated signal over a powerline wire without destroying anything in the process. That's all.

Last edited: Jun 17, 2013
6. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,496
2,837
Jan 21, 2010
The core will not react well to frequencies in the gigahertz range. The inductance of the core will be enormous, the inter-turn capacitance will be high. It is likely that the capacitive effect will dominate and the transformer will look like a short circuit and not like a transformer.

7. Zip1234

12
0
Jun 13, 2013
The frequency is only 60Hz.

The pulse rate is 1GHz.(pps)

8. davennModerator

13,833
1,950
Sep 5, 2009

As Steve said earlier Pulsed AC doesnt make sense

Last edited: Jun 17, 2013
9. davennModerator

13,833
1,950
Sep 5, 2009
If you have an actual project in mind ... ie. this isnt just some theoretical thing
it maybe time for you to tell us exactly what you are trying to do

so far you are trying to do 2 things which dont mix

Dave

10. Zip1234

12
0
Jun 13, 2013
I am trying to transmit a modulated signal over a powerline wire without frying anything. I read transformers filter higher freqencies.

Are there other ways to send the signal?

Powerline Telecommunications.

11. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,496
2,837
Jan 21, 2010
For what purpose?

There are many applications for this which already exist. The frequencies used vary from audio frequencies (where the signal needs to pass through transformers) to much higher frequencies where the intent is that it doesn't.

One problem with these higher frequencies is that they can generate massive RFI. Davenn might be able to comment more on that. Delivering internet via powerlines was an option that would have both worked AND destroyed HF radio.

12. Number

65
0
Jun 9, 2013
You're confusing 'pulses' & Hertz. Hertz is a pulse, or cycle (whatever you want to call it, periodicity, repetition, etc.). If you pulse at 1*10^9 times a second, your hertz, or cycle/periods/pulses would be one billion hertz. Pulses, as in a square, triangle or sawtooth for DC, is an easy illustration of this. AC itself does not pulse. It is just Alternating Current. Going between a positive & negative in a cycle/period. DC can be chopped to switch between high/low, an example is computing, where a pulse signals a gate or transistor/network to toggle on or off. Pulsed AC is the incorrect term here.

So to have one billion pulses @ 60Hz makes literally, no sense. It's like saying 1 divided by Hill. wtf does that mean?

What might help you is looking into sweeping signals, which can increase or decrease in amplitude or in frequency. Think of how sirens work, a small sound at first, followed by increasing noise/volume until it hits the peak and then decreases, causing the wailing sound you would hear.

DC itself is not AC, it is chopped up and has a very linear progression or recession in the cycle. Think of a triangle wave, it slowly, or quickly, goes up to a value, then immediately drops to another value.

DC also, generally speaking, does not alternate between positive & negative voltages, it is either on or off or somewhere inbetween. It does not have a nice parabolic shape to it. Now, you can make pulsed DC into a sinusoidal waveform, at which point, if I'm not mistaken, it stops being DC. An example of this would be an inverter which inverts the DC into AC. At which point the output is no longer called DC, or pulsed DC, it is now an alternating current between some positive & negative voltage/s.

I hope this adds some clarity. Check the examples I listed, I think it will give you a nice visual reference for the things you're trying to do.

Also, if you wanted a 1 billion Hertz signal, I do not think that it would harm the cable. I have heard of high current or high voltage wires used for certain applications. However I've never heard of a high frequency cable (although I'm sure it exist) so my hunch is that you won't do any harm. However, take it for face value and do a little digging into the subject. My recommendation would be to call a local engineering office and see what they say. Not only is talking to someone face to face better, because you get immediate feedback on your questions/comments. I've contacted my local city engineering department a few times when I had a question or two that I needed addressed in person, and it turned out very good.

Well in any case, I hope you get the answer you're looking for and your project goes smoothly.

13. Zip1234

12
0
Jun 13, 2013
Let's say we use DC current that switches from positive and negative voltage. Will a unpulsed DC current saturate a standard power transformer?

What if the duty cycle is low enough?

What happens if the transformer(and the magnetic field it has) oscillates at a precise frequency. Will it stop annuating microwave frequencies?

Last edited: Jun 17, 2013
14. Number

65
0
Jun 9, 2013
Switching from positive to negative with DC would just be AC. As it switches from a positive voltage to a negative voltage. Direct Current by definition will not switch from positive to negative, as that would just be AC. If you, however, chop it up and have a nice frequency, 60Hz or so as an example, you can use the transformer. However if I am not mistaken if you use a square/triangle/sawtooth wave through a transformer your creating a boost converter.

Steve might be able to expand on this a little bit more, but for intents and purposes I believe you're just boosting the DC via a transformer.

I'm not fully understanding what it is that you're trying to accomplish. What is all the need for converting DC to AC or using a transformer anyways? Is the signal the only thing you're trying to manipulate or is there another application that you're trying out as well?

It seems as though you're doing a lot of things that aren't necessary for your desired outcomes.If all you want is AC, by all means that is easy to do and easy to build. If you're trying to get DC to AC that's doable as well, in the form of an inverter. The use of the transformer I am still not grasping from your explanations of sending a high frequency signal through a wire. In which case you might want to look into transmitters and see if you can find anything that is suitable to your needs/wants.

Hope this helps, it is an interesting problem you present. I hope we can figure this out.

15. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,496
2,837
Jan 21, 2010
As mentioned above, that's AC. Pulsed DC would typically be 0V or whatever the DC voltage is. In your case you are also talking about reversing the voltage.

It might. Saturation of a transformer is determined by the peak flux density being reached in the core. This is determined by the size of the core, the number of turns, and the current. In a given inductor it is thus determined by current. At very high frequencies (dv/dt) things change a bit because the core is imperfect -- which is a reason to change core materials to suit the frequency.

A continuously changing voltage (e.g. a sine wave) will always see the reactance of the inductor an this will limit current even if the DC resistance is zero.

A waveform that contains periods where the rate of change is zero (e.g. a square wave) can (if the state is maintained too long) result in the current being limited by the resistance of the winding. This is typically a small value and can lead to very high currents, and thus very easily into core saturation.

A waveform that has very fast rising and/or falling edges (e.g. square waves again) can result in low current due to the high frequency components of the signal that create the edges. This can result in slow changes in current and large voltage spikes.

It is more the on time that is the issue. For a given frequency, reducing the duty cycle will control the current as long as the core does not reach saturation.

Transformers don't oscillate.

The question is therefore meaningless.

Inductors have an inductance (duh), and a resistance (not surprising), but also capacitance between the windings. Thus a given inductor will act as if you have a pure inductor with a series resistance all in parallel with a capacitor.

At DC, the resistive effect is dominant. As the frequency rises the inductive effect becomes dominant. As the frequency rises still further, the capacitive effect begins to dominate. A typical inductor thus behaves like a resistor, and inductor, or a capacitor -- depending on the frequency.

So an inductor designed for 50/60Hz will act like an inductor (or a transformer, etc) at 50Hz, but more like a resistor at perhaps 5Hz, and completely like a resistor at 0.1Hz. At 5000Hz the capacitive effect might be noticed, and the transformer might also be degraded by the core losing energy. At 50kHz, a huge amount of energy would be lost in the core and the capacitive effects would probably be very noticeable. At 500kHz the transformer probably begins to look like a capacitor.

Remember that you're not talking 500kHz, or 1MHz, but 1000 times higher again -- 1GHz. At these frequencies, even the wires connecting the transformer will look like capacitors. The transformer probably appears to be a dead short (not what you would expect from the inductance!)

16. Rleo6965

585
9
Jan 22, 2012
How about observing the actual pulse signal to your iron core transformer using hi frequency oscilloscope.

17. Zip1234

12
0
Jun 13, 2013
Thank you all for your help. I have learned a lot.

What if you made a transformer vibrate(physically vibrate)? Will that have any effect on the magnetic field inside it?

18. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,496
2,837
Jan 21, 2010
Practically speaking, no.

It will do something because there is some magnetic leakage and this will interact with the earth's magnetic field. However the effect will be extremely tiny and may not be measurable.

19. Proschuno

87
1
Aug 1, 2011
Zip1234, are you saying that there is a 1 Ghz squarewave riding on the 60 Hz sinusoidal? similar to how an SMPS pulses regular 50-60hz AC on and off at like 40-100Khz? is that what you're trying to say?

20. Zip1234

12
0
Jun 13, 2013
Yes. That it it.

So will a vibrating(ultrasound tuned at 1ghz through the wire) transformer be able to attract a 1Ghz RF DC signal via resonance?