Filter bandpass

Thread Starter

hero22222222222

Joined Dec 22, 2019
78
Hi i want to design a filter that will pass a signal with frequencies between 45-55 hz without attenuating the signal. And frequencies outside this range to be attenuated. I have connected the most basic hpf and lpf to create a bandpass filter using resitors and capacitors.

The issue is when i use f=1/(2*pi*r*c) for a band pass of 45hz and 55hz, the signal will be attenuated by -3db... how can i make a filter that wont attenaute my frequency band of 45 to 55

I will be working with a 300V 50hz signal which i need it to pass through the filter without any attenuation

Please help
 
Last edited:

crutschow

Joined Mar 14, 2008
27,212
How steep a roll-off do you need outside the passband?
What is the voltage and frequency of the signals you want to attenuate?
 

Thread Starter

hero22222222222

Joined Dec 22, 2019
78
How steep a roll-off do you need outside the passband?
What is the voltage and frequency of the signals you want to attenuate?
Hi thank you for your reply.

As long as the filter doesnt attenuate my voltage signals between 45-55hz it will do the job for me. So the roll off outside the band is not a big concern to me (it can be anything as long as it has been slightly attenuated)...

I will be working with a 300V 50hz signal which i need it to pass through the filter without any attenuation
 
Last edited:

Papabravo

Joined Feb 24, 2006
16,132
For such a narrow bandwidth, cascading the HPF and LPF is not the way to go,
Since the signal is 300V, an active filter is not a viable option. There is no way to make a passive filter at that voltage level without attenuation in the passband. The inductors for such a filter will be huge.
If the rolloff outside the passband is not a concern, then it is hard to imagine a purpose for such a filter.
Take some time to think about what you are asking and see if you can refine your requirements.
Consulting a textbook on filters might be helpful in this regard.

https://www.amazon.com/Analog-Filte...ds=Analog+filter+Design&qid=1613672261&sr=8-6

A bargain at $17
 

Thread Starter

hero22222222222

Joined Dec 22, 2019
78
Hi thank you for your reply.

As long as the filter doesnt attenuate my voltage signals between 45-55hz it will do the job for me. So the roll off outside the band is not a big concern to me (it can be anything)...

I will be working with 300V
For such a narrow bandwidth, cascading the HPF and LPF is not the way to go,
Since the signal is 300V, an active filter is not a viable option. There is no way to make a passive filter at that voltage level without attenuation in the passband. The inductors for such a filter will be huge.
If the rolloff outside the passband is not a concern, then it is hard to imagine a purpose for such a filter.
Hi thank you for your reply

I am thinking of ideas to prevent someone from suppling a 50hz rated load with a voltage source that is outside the 50hz band. I can always step down the voltage if the voltage is the issue.... then add some circuitry with an led to light up if the frequency is outside the rated frequency
 

Papabravo

Joined Feb 24, 2006
16,132
Hi thank you for your reply

I am thinking of ideas to prevent someone from suppling a 50hz rated load with a voltage source that is outside the 50hz band. I can always step down the voltage if the voltage is the issue.... then add some circuitry with an led to light up if the frequency is outside the rated frequency
I think you may have some unrealistic ideas about what a filter can and cannot do. They are used primarily with signals that carry information, not power. Using Active Filters with information signals is restricted to a range of roughly 0 to 36 volts or ±18 volts. These are typical supply voltages for operational amplifiers. For Radio Frequency (RF) and Power circuits we use passive filters consisting of inductors and capacitors. These components do not have the same voltage and current restrictions that operational amplifiers have.

So a power conversion scheme where you change from 300 VAC to some lower value would be of limited utility, especially if you wanted to restore the original voltage after the filter. There is an imutable rule about power conversion schemes and it is this:
The output power will always be less than the input power. Sometimes it will be a great deal less.

My sense of what you are trying to do involves checking the frequency of an input and allowing it to pass to the output without modification. If the frequency is above or below the allowed range then you want to open a switch and allow nothing to pass. This is an inherently digital process rather than an analog process.
 

Thread Starter

hero22222222222

Joined Dec 22, 2019
78
I think you may have some unrealistic ideas about what a filter can and cannot do. They are used primarily with signals that carry information, not power. Using Active Filters with information signals is restricted to a range of roughly 0 to 36 volts or ±18 volts. These are typical supply voltages for operational amplifiers. For Radio Frequency (RF) and Power circuits we use passive filters consisting of inductors and capacitors. These components do not have the same voltage and current restrictions that operational amplifiers have.

So a power conversion scheme where you change from 300 VAC to some lower value would be of limited utility, especially if you wanted to restore the original voltage after the filter. There is an imutable rule about power conversion schemes and it is this:
The output power will always be less than the input power. Sometimes it will be a great deal less.

My sense of what you are trying to do involves checking the frequency of an input and allowing it to pass to the output without modification. If the frequency is above or below the allowed range then you want to open a switch and allow nothing to pass. This is an inherently digital process rather than an analog process.
Thank you for your detailed reply. I did not know this about filters... so thank you for sharing your knowledge.

Yes you are spot on with your last paragraph.

What would be the best way to implement this.
 

Papabravo

Joined Feb 24, 2006
16,132
If it were me I would start with a pair of solid state relays, one for the hot and one for the neutral. On the side of the SSRs closest to the source power I would place a very light resistive load in series with an opto-coupler. You can configure the opt-coupler to output a square wave that is high for the positive half cycle and low for the negative half cycle. At this point you can either use a microprocessor to measure the frequency or you can have the high going portion of the cycle trigger a retriggerable one-shot with a timeout of say 3 cycles at 45 Hz. As long as the oneshot does not timeout you know the frequency is above the minimum. You can maybe do something on the high frequency side as well.

You could also count cycles with a counter and then use a pair of digital comparators to see if the frequency was inside or outside the bounds.

I prefer the processor solution because it appears to be more flexible
 
Top