Creating a Pulse Width Modulator Circuit

[swty todd]

We have been asked to make any circuit of our choice using op-amps on a breadboard or a PCB. We have found this circuit of PULSE WIDTH MODULATOR on the net and we have decided to go for it.But our teacher wants us to explain the designing of the circuit too namely why have we used the resiistors of that particular value only ,using equations and so on.

I am uploading the circuit as an attachment

please help.

 

[MATLABdude]

STEP 1: 'Print' your document to PDF using something like Adobe Acrobat, or the free PDF Creator:
http://sourceforge.net/projects/pdfcreator/

STEP 2: Upload the .pdf document somewhere other than Physics Forums so that someone can see it prior to approval (which can take a while), and then post the link. Say, your personal webspace.

STEP 3: ?

STEP 4: Learn

 

[berkeman]

Step 47: Mentor approves attachment at quarter-to-midnight.

 

[berkeman]

We have been asked to make any circuit of our choice using op-amps on a breadboard or a PCB. We have found this circuit of PULSE WIDTH MODULATOR on the net and we have decided to go for it.But our teacher wants us to explain the designing of the circuit too namely why have we used the resiistors of that particular value only ,using equations and so on.

I am uploading the circuit as an attachment

please help.

You know the rules, swty. We don't do your work for you. You tell us how the circuit works, and we can offer hints and corrections here and there.

BTW, the circuit schematic makes no sense as presented. It is not a PWM circuit as shown. What is meant by +/-5V input?

If you want to do a PWM circuit for your project, that would be great. But you need to explain why a triangle wave generator is important for a basic opamp based PWM circuit first, and then explain what is missing from the schematic that you pasted here, and then fix it, and then...

 

[MATLABdude]

Thought I recognized the style... It'd be a whole lot easier (and more proper, from an academic standpoint--learn to http://en.wikipedia.org/wiki/Citation" !) to just point us to the source you originally got the circuit from. For instance:

Pulse Width Modulator, from "Op Amp Circuit Collection", National Semiconductor, 2002, p. 9. Available online at:
http://www.national.com/an/AN/AN-31.pdf

Still, what are your thoughts on how this circuit works? And that circuit is awfully complex. Perhaps you should start with the Free Running Multivibrator on p. 8?

 

[berkeman]

Thought I recognized the style... It'd be a whole lot easier (and more proper, from an academic standpoint--learn to http://en.wikipedia.org/wiki/Citation" !) to just point us to the source you originally got the circuit from. For instance:

Pulse Width Modulator, from "Op Amp Circuit Collection", National Semiconductor, 2002, p. 9. Available online at:
http://www.national.com/an/AN/AN-31.pdf

Still, what are your thoughts on how this circuit works? And that circuit is awfully complex. Perhaps you should start with the Free Running Multivibrator on p. 8?

Yeah, definitely National's style of schematic from their App Notes. Good find -- how in the world did you find that with a search?

And do you see any way it could be called a PWM circuit? I'm at a loss on that.

 

[MATLABdude]

Yeah, definitely National's style of schematic from their App Notes. Good find -- how in the world did you find that with a search?

And do you see any way it could be called a PWM circuit? I'm at a loss on that.

Right off the bat, I knew it looked familiar, and was probably from a datasheet, either TI's or National Semiconductor. I Google'd the part number, got NatSemi's page, looked through the datasheet and found nothing, and then, on a lark, clicked on one of the handily-situated App Notes. What can I say? I'm good at finding stuff.

I'm a little rusty on more advanced op-amp circuits, and I haven't the foggiest idea how that circuit works, what it's taking in as an input (square wave? either 5 or -5V? PWM of a smaller voltage?) or what duty cycle it's returning. I wish they'd annotate that a little better, because it's probably just a circuit diagram cribbed out of one of their other datasheets without any of the principles of operation or even expected input/output.

 

[swty todd]

Thought I recognized the style... It'd be a whole lot easier (and more proper, from an academic standpoint--learn to http://en.wikipedia.org/wiki/Citation" !) to just point us to the source you originally got the circuit from. For instance:

Pulse Width Modulator, from "Op Amp Circuit Collection", National Semiconductor, 2002, p. 9. Available online at:
http://www.national.com/an/AN/AN-31.pdf

Still, what are your thoughts on how this circuit works? And that circuit is awfully complex. Perhaps you should start with the Free Running Multivibrator on p. 8?

you got the source right.
I got it from there.
As far as multivibrator is concerned we will already be doing that in our labs and the teacher wants us to get a circuit which we won't be performing in our labs.
I don't how the circuit works (if it does).
I think +-5 volts refers to an AC input.
Maybe my next step should be leaving that circuit and searching for some circuit.

 

[swty todd]

I tried finding a suitable circuit with a understandable theory and design but I couldn’t get one.
So I decided to use allaboutcircuits simple bargraph driver circuit (where the component values were not shown) and decided to use my own values.
For the opamp I have decided to use LM741 (since it is easily available). For the resistors which are connected in series with the diode I decided to use 1k ohm resistors ,so as not to harm the LEDs.
I have decided to use the red LEDs.
For the resistors I which are directly connected to the V+ I have decided to use three 1k ohm resistors.And I will take V+ as 15 volts. Now the problem is how to calculate V1 and V2 .Since V+ is getting divided into two parallel branches one into the three 47 ohm resistors and the other into the opamps, I wonder how to calculate the resistance of the opamp branch.
I think I am stuck up. Experimenting in the lab is not a good option right now since my college is far away and its holidays for 2-3 days now.

 

[berkeman]

The values of the resistor divider on the left are not important. Just shoot for maybe 10mA in that divider. The resistors in series with the LEDs most certainly can't be 1M Ohm as you've drawn. What forward current do you want in the LEDs (look at their datasheet)? What is the LEDs' typical forward current? What voltage will be across the LED+resistor when it is on? That sets the value of the LED resistors.

 

[swty todd]

I tried the circuit in my lab and unfortunately it didn’t work, but hopefully after a bit of troubleshooting it will.
My final circuit that I am using is below.

Actually I am supplying +15 volts to pin 7 of IC741 and -15V to pin 4 of the IC.
Even without giving any supply to the +V & Vin both the LEDs glow.

Actually when pin 7 & pin 4 were connected as above, +V was at 5V & Vin was at zero, both the LED’s glowed. Using multimeter I noticed that the voltage across the LED2 was 7v .When I increased Vin to 0.1V the voltage across LED2 went to 13.3V & further increase in Vin didn’t lead to further changes.

At the same time the voltage across LED1 was 9.5V(when Vin was at 0V) & 13.19 V (when Vin was at 0.1V).
Increasing Vin further didn’t change the voltage drop across LED2.

Increasing the values of resistors only decreses the intensity of light of the LEDs.
I fail to understand why it’s reaching saturation so early, how should I modify it?

 

[berkeman]

Several issues with that circuit. First, LM741 opamps are not a good choice for comparators. Second, comparators should have some positive feedback for hysteresis, or they will oscillate when the + and - inputs are close to each other (which will happen a lot in a bargraph application like this).

Third, since the input and output are ground referenced, you usually would use a single supply opamp or comparator that includes ground as a valid input. Since you have a -15V supply, you need to look into what the maximum reverse voltage is that the LED can tolerate. LEDs generally do not have high reverse standoff voltages.

Look into single-supply comparators, to see if they would work better for you, and be sure to add hysteresis to all of your comparators (or opamps used as comparators).

 

[swty todd]

First, LM741 opamps are not a good choice for comparators.

Is 741 really imp?Because here op-amps are expensive and the place where they are sold is far away. So I need to be very clear about which op-amps I should be getting.I performed a search and got this list for single supply op-amps.http://search.datasheetcatalog.net/key/SINGLE+SUPPLY+OP"
Out of them LM7131 seems to be the simple one.What other characteristics should I be looking out for?

you need to look into what the maximum reverse voltage is that the LED can tolerate. .

Actually I don't anything about my LED except that they are red in color. I just found them at my home.But a search on google revealed red LEDs having max reverse current flow from 10uA to 100uA.Ironically I don't remember burning more than 4 LEDs in the 15 or so times we rebuilt the circuit and tested it.

comparators should have some positive feedback for hysteresis
.

Someone suggested me that since the opamps are going into saturation , I should use a negative feedback. To me using positive and negative feedback both seem correct.Anyways never used positive feedback,so what value of resistor should I be using or how do i calculate it?

And thanks.

 

[berkeman]

I googled comparator hysteresis tutorial, and got lots of good hits. Here is a good one from Analog Devices:

http://www.analog.com/static/imported-files/tutorials/MT-083.pdf

.