Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Smoothing a Square wave

Status
Not open for further replies.

syee10

Member level 3
Joined
Apr 2, 2010
Messages
63
Helped
1
Reputation
2
Reaction score
0
Trophy points
1,286
Location
Malaysia
Activity points
1,758
I have a square waveform as in the 1st figure which is not a smooth waveform. It has a high peak that i dun want in my real circuit. I want it to be a smooth square waveform as in the 2nd figure. So how am i going to convert the square wave with the peak to a smooth square waveform? Please assist me..
 

Attachments

  • Square wave 2.png
    Square wave 2.png
    4.4 KB · Views: 107
  • Square wave 1.png
    Square wave 1.png
    3.6 KB · Views: 90

how are you creating this waveform?
 

I have a square waveform as in the 1st figure which is not a smooth waveform. It has a high peak that i dun want in my real circuit. I want it to be a smooth square waveform as in the 2nd figure. So how am i going to convert the square wave with the peak to a smooth square waveform? Please assist me..

Try going thru. a switch. Like, input ur wave to the base of a BJT and output at collector, to see if any improvement... Ur BJT must work as a switch. In this way, ur output wave will be reversed, which should be not an issue.
 

If you know the maximum voltage of your squarewave, you could put a zener diode with a slightly higher breakdown voltage on the line to shunt the excess power to ground, minimizing the peak excursion (clamping).
**broken link removed**

If the squarewave amplitude = 5V, and Vpeak = 12V, then use a 5.1V or 5.6V zener diode. You'll still have a little peak, but it will be clamped to around 5.1/5.6 volts.

If you want to truely flatten the signal, you could run the clamped output signal into a comparator to "square it up", but you lose all of the original amplitude... that's bad, if your squarewave amplitude carrys any information.

Or, you could use a zener slightly below your squarewave amplitude, say 4.7V, if you can survive dissipating the power, and reducing the squarewave's amplitude.
 

Are you sure that that is the generated signal ? Is the oscilloscope probe calibrated ?
 

Put a comparator with referense on the middle (or double not gate)... Use open collector devices in order to set your maximum voltage.
 

Ringing on the rising edge of a square wave is often caused by sending the signal down a coaxial cable, but not terminating that cable in a resistor that is the same value as the coax line's characteristic impedance. That is the best way to stop the ringing/overshoot.

A poorer way, but effective at lower speed signals is some sort of clamp diode arrangement. If the waveform is overshooting the +5 volt rail, you could clamp the signal thru a schottky diode to that rail. If it is undershooting below ground you can clamp with a 2nd diode to ground.

Another way is to simply lowpass filter the signal, with series L/shunt C, or series R/shunt C structure.

A way I have used to clamp rining high speed bus lines that had non-standard loading circuits is to have a simple buffer stage physically located right at the load. In other words. have a digital gate at the end of the line just before the load, and buffer the signal that way. Because of the very short distance between the buffer gate output and the load device, you do not get much time delay ringing, and the digital data trasnfer has no ringing/errors. This is especially useful on the Load enable line on various busses, as the data is transfered before the rising clock edges so its ringing can die down before the load enable command is initiated.
 

Hi,
Connect a resistor to the positive terminal of your square wave and the other terminal of that resistor connect it to the N junction of a zener diode and the P junction of the zener diode connect it to the other terminal of your square wave.
The voltage of the zener diode must be less than that of the unclean square wave voltage by at least 10%.
the resistor is used for limiting the consumed current from your square wave.
 

Can I connect a 5.1v zener diode with a resistor with our line voltage (230v AC)? What is the typical maximum allowable voltage of a zener and how can I find it?
 

Can I connect a 5.1v zener diode with a resistor with our line voltage (230v AC)? What is the typical maximum allowable voltage of a zener and how can I find it?

Recall that a zener diode is simply a diode with a well-defined reverse breakdown voltage. In an AC circuit, the voltage across the diode will bias it in both directions (forward biased on one half of the sinewave, reverse biased on the other half of the sinewave). So, half the time the diode would be trying to clamp around +5.1V, the other half, it'd be trying to clamp around -0.8V. So, from a theoretical standpoint, no, that would not work. From a practical standpoint it's rather dangerous (hooking circuits up to your mains w/o protection like a fuse, or full understanding of what you're doing and shocking yourself or starting a fire).

If you are simply trying to make a 5V regulator, you first need to rectify your incoming AC voltage, so that its only a positive voltage, relative to ground. The most direct way of power rectification and regulation is to buy a "wall wart" power transformer, like this. They are relatively inexpensive and come in many voltage/current ratings. Additionally, you can buy "universal" power adapters that you can select from several output voltages.
 

Actually I planned to use this square wave as an input of zero crossing detector of microcontroller pin. So, to get better slope of square wave I wanted to use mains AC by cutting the upper part of the wave.
 

Sayee10,,
This is an overshoot of the square wave created by you.There are four ways of processing to contain this .
a)Ceramic capacitor at the input point to ground .
b)Source resistance damping to RC low pass filter .
c)Capacitance integral feedback from output to input to limit bandwidth .
d)Knowing the Vhi of the square wave and use a zener in series with a forward rectifier to limit the forward to Vz +Vcutin and reverse to very high value of Vrev. of diode .
 

Hi My friends!
The snubber network can help here . the poles of your circuit are stable , but , not enough , thus the snubber network , can more shift them to the left of s plane.
Best Wishes
Goldsmith

---------- Post added at 01:08 ---------- Previous post was at 01:07 ----------

By the way you can use two not gate after clipping with zener
 

Hope you would have tried out different dampers and request a feedback ..
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top