Maker Pro
Maker Pro

Building a custom notch filter

mallyajiggs

Feb 6, 2010
18
Joined
Feb 6, 2010
Messages
18
Hey guys

I am building a 60 hz notch filter and have the components for that, but i am not sure on what i can build the circuit ?
This will be connected directly to the mains( residential ac power outlet).

I assume i cannot use breadboards to build this circuit on ?
any suggestions ? other than using a wooden board ?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Aaagh, OK -- A notch filter at 60Hz is designed to attenuate 60Hz from the incoming signal.

I presume you're in the US or a country with 60Hz mains.

If you attenuate the 60Hz component of mains power then you will effectively be stopping any power from getting through. A switch may be a better option.

On the other hand, maybe it's a narrow bandpass filter, and you want to only allow the 60Hz mains through, and eliminate noise and other signals.

In the latter case I'd certainly be hoping that all the components are rated for mains use and can carry the current required. I'd also be wary that the device doesn't act as a low impedance at any frequency which might cause it to act as a short across the mains.

Whatever you build it on will be at mains potential and has the distinct possibility of killing you. I'd make sure it's in a suitable enclosure and fused.

If you can post a circuit diagram, then maybe someone can assure you that it's OK to place across the mains.
 

mallyajiggs

Feb 6, 2010
18
Joined
Feb 6, 2010
Messages
18
i need to filter the AC frequency so that i can increase the dynamic range of sampled data... esentially this notch filter goes before a band pass filter which is like a power interface circuit and then connected to a oscilloscope to sample ..

So what can i use as the base for this notch filter ? also any suggestions for the entire circuit ?
attaching the circuit diagram Link
22340603.jpg
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
It looks to me that the circuit you have provided (although too small to make out any real details) is designed to provide signal outputs, probably for extracting control (or maybe broadband?) signals that are superimposed on the mains.

From what I can tell, the 60Hz output is a signal too, probably for timing - i.e. a clock.

In your first message, you suggested that it would be connected to the mains, in the second that it would be connected to a signal. Is your signal on the mains? Or is it affected by mains hum?

You can't filter the mains, reject 60Hz, and then power something from what is left.

You can filter the mains to reject 60Hz, and look at the resulting data as a signal.
 

mallyajiggs

Feb 6, 2010
18
Joined
Feb 6, 2010
Messages
18
ok, just to make things clear , i am not a electronic guy ...
its just that i need to reproduce this circuit to analyse data.

So yeah i need to reject 60Hz and look at the resulting data as a signal, also the signal to which it is connected is the mains.

attaching a larger circuit image

imgng.jpg
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Ok. I wonder if an LC filter would be easier. Unless you're looking for frequencies close to and both above and below 60Hz then a notch filter may not be appropriate.

The circuit you have looks like it will provide isolation of the signal output from the mains, but do these frequencies correspond to what you're looking for?

It looks like the transformer is used to determine the range of frequencies, but again, I can't read all the component values (although this copy was a great improvement)
 

mallyajiggs

Feb 6, 2010
18
Joined
Feb 6, 2010
Messages
18
i am looking to analyse noise signals from electrical devices .. hence filtering other noises even that of the AC mains
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
Twin-T filters are very effective for removing a fundamental frequency and passing all harmonics etc. unaffected. Properly tuned this should work nice for your application.
What you build it on is not so important as how you build it (layout, distances, mechanical stability) and how you enclose it.
It goes w/o saying that the capacitors must be mains rated.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
It looks to me that if you remove the 60Hz port from that circuit, and get the correct transformers, you may well be able to do what you want.

It is important that all of those capacitors are mains rated. For the deepest notch, you need to ensure that the capacitors and resistors are as close in value as possible. I would recommend either close tolerance devices and/or measuring a batch to get the closest match.
 

mallyajiggs

Feb 6, 2010
18
Joined
Feb 6, 2010
Messages
18
Thanks Steve,

Let me just summarize quickly what i am trying to do, i guess i wasn't clear enough earlier.
This circuit diagram is part of something, i just need to reproduce this to analyse data.
It is very important for me to stick to the circuit and not change any components.

Essentially this is an interface circuit for the USB oscilloscope which will be connected to either of the 3 ports depending on the analysis i would require. Also the purpose of 60 Hz notch filter here is to eliminate continous noise produced by devices with motors.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
OK. The 60Hz port just seemed superfluous since it was that which you were trying to eliminate. It might also affect the depth of the notch.

The 60Hz filter will only remove 60Hz and leave the noise -- I thought it was the noise you were trying to measure?
 

mallyajiggs

Feb 6, 2010
18
Joined
Feb 6, 2010
Messages
18
Yes, but i want to analyse signal which has 60 Hz component and then once without the 60 Hz component
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
OK, that circuit (as far as I can tell) will only give you one or the other, but not a 60Hz + noise.

I'm not sure if you can combine the outputs again, but getting their relative levels correct would be an issue.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I can't really help you with the choice of transformer for that. The transformer you have selected would certainly work for 60Hz signals, but whether it has the right impedance and turns ratio is another question.
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
No, that's a 12VA 2x115V output. You don't want to apply that much into you USB 'scope..
Use the lowest power & lowest voltage you can find, certainly not much more than a 1VA 5V transformer.
 

mallyajiggs

Feb 6, 2010
18
Joined
Feb 6, 2010
Messages
18

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
That one will at least be safe for your 'scope but is still way more powerful (+ heavy & costly) than neccessary.
Use this (single primary) or this (dual primary).
Digikey doesn't say what the single primary voltage actually is, but the datasheet reveals it's 115V (surprise-surprise). Even if you're in a 115 country it won't matter if you get the dual primary and run this transformer at half voltage.
You'll have to be careful anyway to stay away from the lowest range on your 'scope. (5V x 1.1 x 1.24 x 1.414 = max 9.65Vpk)
 

mallyajiggs

Feb 6, 2010
18
Joined
Feb 6, 2010
Messages
18
ooh... thats great help... i was a little concerned too.
whats this deal with series & parallel output voltage ?

i'll be taking the voltage out on parallel ryt ? according to my circuit diagram specified on page 1. ( Transformer 3 - 60Hz port one )
esentially my parallel output voltage should be 5V ryt ?

and which would you recommend single or dual primary ?


Again.. thanks a lot......
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
Both have dual 5V secondaries that can be wired in series for 10V, used independently, or wired in parallell for 5V full power. You need only use one of the secondaries for your application.
With dual primaries you can wire them in series (for 230V use) if you want, but still apply 115V to it, resulting in only 2.5V output. This might also have the advantage of less waveform distortion since the core is nowhere near saturation. I'd get the dual primary just for the added flexibility. More pins to worry about but the same price.
 
Top