Design an Active Bandpass Filter

[HelpingHand]

Design an Active Bandpass Filter
Center Frequency=50kHz
bandwidth of 10 kHz
passband gain 7


How would I complete a pencil and paper design and then simulate the design in Multisim.

After that I am to record my finding and data?? Confusing on how to start the design to begin with.

 

[crutschow]

Here are some references on designing active filters.
One thing missing from your requirements is the filter order (the filter sharpness or rolloff rate at the filter edges).

 

[HelpingHand]

I was told I could use a low pass filter to an IC with an output into a Highpass filter.
Or any type of Highpass as long as it isn't simple, is what I was told.

 

[crutschow]

I was told I could use a low pass filter to an IC with an output into a Highpass filter.
Or any type of Highpass as long as it isn't simple, is what I was told.

The simplest filter is a single RC, 1st order filter. The next would be a 2nd order bandpass filter, so apparently a 2nd order filter would meet the "it isn't simple" requirement. (?)

 

[HelpingHand]

The way you explained things is already better understood. Why wouldn't the prof. just say 2nd order filter? But yes the requirements would be met with a 2nd order bandpass filter.

 

[MikeMl]

The way you explained things is already better understood. Why wouldn't the prof. just say 2nd order filter? But yes the requirements would be met with a 2nd order bandpass filter.

So you will use two cascaded opamps; one to create a highpass, the other to create a low-pass response?

 

[HelpingHand]

This is correct Mike.

 

[JoeJester]

Here is a reference for you ....

 

[HelpingHand]

Attached is a picture of my multisims but I still have not obtained a gain of 7. Could you please help again?

 

[JimB]

I still have not obtained a gain of 7

Use a newer op-amp than the 741, try a TL071.

Your input signal is 15v RMS, a gain of 7 would give an output signal of 105v RMS.
Your supply is only +/- 15v so your input signal is FAR to big. Try 1 volt or less.

Some of the resistor and capacitor values look a bit odd, maybe worth re-checking.
Actually, looking again at your circuit, it looks all wrong.
Read Joe Jesters application note and try again.

JimB

 

[MikeMl]

You need to figure out how to get that toy simulator you are using to do a Bode plot (frequency response) . You are driving your proposed circuit with 60Hz, instead of sweeping the frequency from ~1KHz to about 1MegHz.

As Jim said, you are overdriving the input.

A 741 amplifier does not even come close to having sufficient gain-bandwidth product to do this job.

The values used in the input and feed back at the opamps are way off. Try to constrain them to be between ~10K and 1MegOhm.

There is no need to have a third opamp to get the gain. Set up the HP and LP sections so each has a small gain such that Glp*Ghp = 7.

I took your circuit, changed one resistor, and here is the Bode Plot showing a gain of 7 in the pass band. You still must change the design to get more reasonable values for R1,2,3 and 6. Note that I am driving the input with 1v and getting 7V out. How did I do that?

Also, if you look at the Bode Plot, you are not getting the filter rolloff sufficient to be at 0.707*7V at 45kHz and at 55kHz. You will need to use a at least a 2-pole filter for each the HP and LP sections, for a total of 4 poles...

 

[JoeJester]

Mike,

his "toy" simulator is multisim.

I agree with you that he needs to know how to make a bode plot.

I also think he needs to lookup multiple feedback bandwidth filters for this assignment. He can read more about them at https://www.circuitstoday.com/band-pass-filters

 

[MikeMl]

...

his "toy" simulator is multisim...

I know. Having to place those stupid little "instruments" like scopes and dvms just gets in the way of doing what you need to do...

Look how simple, intutitive, clean my LTSpice sim of his proposed circuit is compared to that horrible Multisim schematic

 

[JoeJester]

Mike,

We both know, from reading this type of forum, that some people do not wire their schematics "cleanly". If they do their breadboards in the same manner, they are back asking "why doesn't it work ... just like the schematic "didn't work".

Them being not tidy isn't a function of the simulator. They will soon learn and all we can do is provide the example for them. Soon they will take the time to produce clean work ... and it's up to us, as individuals, to address that issue. Using the virtual DMM really isn't the test equipment of choice either.

No one has ever addressed acceptable standards with respect to presenting information here. It was never mentioned at the last rules review, so I guess the general population's give a damm is busted. On the other hand, individuals can choose to respond in whatever manner they feel is necessary when answering a posting.

I typically choose not to respond to untidy schematics ... they waste my time tracing all the wiring. It could come down to deciding on the importance ... waste time tracing their wiring and "take a nap." In that case, a nap is a better time.

 

[MikeMl]

Actually, I was criticizing the MultiSim user interface and use model rather than the OP's drawing style. At least his schematic flows from left to right...

 

[crutschow]

Multisim may have a user interface that you don't like but it has handy features such as switches that can be operated from the keyboard during simulation and LEDs and logic probes that indicate their on or off state on the schematic during the simulation. It has a number of controlled sources (VCVS, VCDS, etc.) that are easier to set up then in LTspice. Being able to easily start, stop, and then continue the simulation is handy. And those "stupid little instruments" have a number of functions that are easier to use then in LTspice. I certainly wouldn't call it a "toy" simulator.

 

[MikeMl]

Multisim ...has handy features such as switches that can be operated from the keyboard during simulation

LTSPICE has a voltage-controlled or current-controlled switch where you can specify a full set of operating parameters, Trip voltage, Hysteresis, On resistance, Off resistance, smooth transitions, etc, etc. You can specify the times when the switch operates (in the time domain).

and LEDs

LTSpice does not have an animated LED, however, it is trivial to click on a LED (or any other component in the schematic) to view the current flowing through it. Zero current means the LED is off, small finite current means the LED is lit, and too much current means you just smoked it. To find out how much current is flowing through the LED in Multisim means you have to place an ammeter in the schematic?

and logic probes that indicate their on or off state on the schematic during the simulation.

LTSpice shows the voltage at any node during DC analysis, just hover the cursor over the node in question. During AC or TRAN analysis, clicking on a node shows its time history.

It has a number of controlled sources (VCVS, VCDS, etc.) that are easier to set up then in LTspice.

Debatable. LTSpice has a rich set of Behavioral sources that allow you to synthesize arbitrarily complex functions; it turns it into a programming environment. The ability to pass variables to sources, components using Params produces much more flexibilty than "recanned" instruments. When you add waveform arithmetic, waveform integration, averaging, FFTs, LTSpice beats Multisim hands down...

Being able to easily start, stop, and then continue the simulation is handy.

LTSpice can do this, but it involves saving the state to a file.

 

[JoeJester]

LTSpice has it's advantages, as does all the other simulation programs out there.

The real problems relates to the models.

Does LTSpice allow you to fully simulate a PIC, including using the ASM code?

LTSpice assumes a few things that the other programs don't.

Everyone has their favorite. LTSpice's biggest point is the price.

 

[crutschow]

Attached is a picture of my multisims but I still have not obtained a gain of 7. Could you please help again?

That is a 1st order bandpass filter. I thought you needed at least a 2nd order filter?

 

[audioguru]

You have the lowpass filter with a 30 ohm resistor and the resistor for the highpass filter is only 35 ohms but the datasheet for a 741 opamp and for almost all other opamps show a MINIMUM load of 2000 ohms. Your load is 30//35= 16.2 ohms.

A simple calculation shows that your input resistor of 1k ohms and feedback resistor of 30 ohms produces a gain for a POWER amplifier (not an opamp) of 0.03 times so the output will be 0.45V for low frequencies.
If the opamp for your highpass filter is also a POWER amplifier then its gain will be 1k/35 ohms= 28.6 times for high frequencies. But it must be an amplifier that works at 50kHz. A 741 opamp is not a power amp and it does not work above about 9kHz.
Your output opamp would have a gain of 8 if you selected an opamp that works at a frequency as high as 50kHz.

I do not know why you are feeding 60Hz to the circuit.