# 50Hz Filter

Discussion in 'Electronic Basics' started by Ed, Aug 3, 2004.

1. ### EdGuest

Hi,

I've made a simple circuit which consists of an ultrasonic transmitter and
reciever. I send out 40KHz pulses and when they return I amplify the them
and put the amplified signal into a comparator (this creates a nice square
wave signal for a microprocessor). The problem is, when I'm not sending out
pulses the receiver picks up the 50Hz signal from the mains electricity
supply (I'm based in England, hence the 50Hz). Should I use a notch filter
to remove the 50Hz signal or should I implement some sort of band-pass
filter, which only accepts 40KHz? What would you all recommend and which is
the easiest to implement?

Thanks for any help.

2. ### PGuest

Hi, I made such things too several years ago.

Make a simple band-pass filter, it does not have to be very steep, at
40 kHz you are many decades away from 50Hz, some RC filtering will do.
Notice that you probably also have some 100, 150 Hz components.

Regards,
Pieter Hoeben
http://www.hoeben.com

3. ### Active8Guest

Screw the BP and notch approach. Usee a HP filter and set the cutoff
freq well above say, 150 Hx just in case. There's bound to be
equipment around with rectifiers in it that will produce harmonics
that will get back into the power lines and radiate. IME, in a 60 Hz
system, I'll find plenty of 120 Hz noise. I can filter it out and
still pass a 100 Hz signal, though.

4. ### EdGuest

Does that just consist of a resistor and capacitor? If so, what values do I
select for the cap and resistor? I've done a google search but can't find

Thanks.

5. ### Active8Guest

It could, but it depends on how much noise we're dealing with. You
might need an active filter like a Sallen-Key or better.

In either case, the cutoff freq is 1/(2piRC)

You also want to keep the input conductors to the amplifier as short
as possible and examine the current flow of the input. Keep the loop
as small as possible.

6. ### Jonathan KirwanGuest

Tthat would still leave some rather broad band noise from where the high pass
begins up through "daylight," yes? How hard is a simple bandpass, anyway?

But what crosses my mind is that the input picks up 50Hz too easily, perhaps.
Using a resistor divider (and an 'HP' capacitor) at the input might help.
Something like:

+V
|
\
/ 2M
1uF \
|| |
,-----||-------+-+--> first gain stage
| || |
--- \
| | receiving / 2M
--- transducer \
| |
gnd gnd

Just a thought.

Jon

7. ### JamieGuest

pass the square wave into a one-shot 555 timer and AND gate.
if the incoming wave does not expire on the high side before
the one-shot timer does then using a gated circuit won't
trigger a pulse.
this means that time base wider than 40 kHz will not be detected/

8. ### Active8Guest

Sure. Try something simple like this first.

9. ### Henry KolesnikGuest

Sounds like a rx shielding problem and the need for a high pass filter.

10. ### Watson A.Name - \Watt Sun, the Dark Remover\Guest

Neither. The two freqs are so far apart that you should be able to get
what you want by using a high pass filter. Change the coupling
capacitor in the preamp to a smaller value, so that the lower audio
freqs can't get thru. Or else add a high pass filter somewhere near the
input.

11. ### Rich GriseGuest

Watson A.Name - "Watt Sun, the Dark Remover" wrote:
[yet another HPF suggestion]
Everybody here is kind of barking up a tree, from what I see.

There is no 40 KHz ultrasonic transducer that will give a perceptible
signal in response to 50 Hz in the audio band.

The guy's circuit is picking up hum electronically, from the ambient
50 HZ EMI.

The whole circuit is acting as an inductive pickup loop.

Find the ground loop(s), fix it, and move on.

(admittedly, a filter to get the 50 Hz interference out of a 40 KHz
signal is kind of a naptime design project. ;-) )

Good Luck!
Rich