Signals

[AxelD]

Hi,

I have a few silly questions regarding signals and waveforms:

1. when we talk about frequency, does the waveform make a difference. i.e. will filters and frequency dependent resistors (using capacitors/inductors) work for all waveforms (pulses included).

2. Assume i have an emmiter follower with an output of 0.7v as a result of having to bias the transistor. If i have a tiny input signal that gets amplified to 1mv, do i see now see an output of 0.71v?

Thanks appreciate the assistance.

 

[RadioRon]

1. when we talk about frequency, does the waveform make a difference. i.e. will filters and frequency dependent resistors (using capacitors/inductors) work for all waveforms (pulses included).

Your question is quite broad so we may need to break it down to smaller questions. The waveform and the frequency are very closely interlinked. As the waveform repeats, so the frequency is set. But if the waveform that is repeated is a rectangle or a triangle or some weird wobbly thing the frequency remains the same as long as the waveform repeats in the same period.

The most basic waveform for our purposes is the sine wave. This is exactly the waveform that you get if you were to view a single point on a rotating circle by viewing the circle on edge. It is also the waveform you get when a person dangles on a long and lossless bungee cord. I mean to say that it is the most basic waveform. Other waveforms, including square waves, triangle waves and all others can be broken down into a sum of many sine waves, believe it or not. It is typical for any waveform to be constructed of a sine wave at the repeating frequency of the waveform plus a sine wave at double that frequency, plus a sinewave at triple that frequency and so on. The amplitudes of these "harmonic" sinewaves determine what the shape of our waveform is.

Since any weird or arbitrary waveform other than a sine wave can be said to be the sum of many sine waves at "fundamental and harmonic frequencies" the effect of a frequency dependent reactance, like that from an inductor, will vary depending on the amplitudes of those harmonics. So components like inductors for example, or filters for that matter, will have different degrees of effect on waveforms of different shapes.

2. Assume i have an emmiter follower with an output of 0.7v as a result of having to bias the transistor. If i have a tiny input signal that gets amplified to 1mv, do i see now see an output of 0.71v?

The bias of a transistor is a DC thing. That 0.7 volts is the DC voltage at the emitter. When you put a signal through of, say, 1 mV, (which I assume is 1mV RMS AC, which is 2.8mV peak to peak AC), then on the emitter you should see a 2.8mV peak to peak AC swing centered at 0.7 volts. In other words, the voltage should swing (its an AC signal) from 0.6986 Volts to 0.7014 volts.

 

[AxelD]

Thank you.

May i ask how the output would behave if the signal was a pulse (1mv - 0 - 1mv - 0 - ...)

As far as your explanation goes and to verify my understanding, a simple pulse (1mv - 0 - 1mv - 0 - ...) could then be filtered using a filters?

 

[RadioRon]

Thank you.

May i ask how the output would behave if the signal was a pulse (1mv - 0 - 1mv - 0 - ...)

As far as your explanation goes and to verify my understanding, a simple pulse (1mv - 0 - 1mv - 0 - ...) could then be filtered using a filters?

You may, but you will need to be more specific as I do not know what this "output" is that you refer to , and the implied "input" and "what is in between the input and the output" as well. The universal language for electronics engineers is the schematic diagram and I recommend that you continue your questions using this language. ( or a block diagram if you think that may help).

A simple pulse, which I would assume to be a rectangular shaped waveform with a very short duty cycle (that is, it is at 1mV for a very short time compared to its time at 0 mV), can be filtered. For example, if you put the pulse through a low pass filter, you will slow down the transition of the edges and soften the shape to be less rectangular and more smooth. But it does depend on the nature of the filter too.

 

[crutschow]

As RadioRon stated, an arbitrary waveform can be described as the sum of many different harmonic frequency sinewaves. If you put this waveform through a filter it will remove or suppress the frequencies (spectrum) outside the filter bandwidth, leaving a signal with only the sinewave spectrum within the filter bandwidth. The degree to which the waveform is altered thus depends upon how much of the original spectrum is removed by the filter.

This can be calculated by using Laplace Transforms, but generally for most engineering purposes, it's sufficient to just simulate the waveform and filter output.

 

[AxelD]

Thanks guys. Ill post a pic later to provide more clarity on some of the questions.

 

[AxelD]

Hi,

Ive attached a picture of what i think occurs with the pulse. Am i on the right track?

The voltage levels in stages 2-4 are approximations. In stage 4, can someone confirm that if the frequency of the pulse is correct, you will see a reasonable copy of the wave at stage 3 and the filter will have no impact? If on the other hand the frequency is not as expected, there will be no pulse at stage 4?

Thanks again.

 

[RadioRon]

Not sure what you mean by "the frequency ...is correct".

In general, we would not expect to see the waveshape as you have shown it at stage 3, since this is before the filter. The waveshape after the filter depends on what is in the filter and you have not described that. You imply that the filter is a low pass configuration made up of a resistor in series followed by a shunt capacitor. This implication comes from the shape of the waveform that you show at stage 4. The filter can be more elaborate than that, and if so, the waveshape at stage 4 would look a little bit different, perhaps a bit smoother.

However, in general, if the frequency is below the corner frequency of the low pass filter, then your waveshape at stage 4 is roughly correct, but at stage 3 it would look little changed from stage 2. If the frequency goes above the corner frequency of the filter, then the amplitude at stage 4 goes down, to zero eventually as the frequency goes much higher than the filter corner.

 

[crutschow]

For a step input or fast risetime pulse the output risetime (0 to 63%) of a simple RC low-pass filter is simply (R × C) seconds. It's basically independent of the pulse frequency unless the pulse width and/or pulse spacing are less that about (4 x R x C) seconds. Below that, the filter will start attenuating the amplitude of the pulse as well.

 

[AxelD]

Thanks that provides more clarity.