High-pass filter. Ceramic or Electrolytic capacitor?

Hello everyone,

I need to high-pass filter a signal before feeding it to an ADC.

Properties of the signal:
- Frequencies in the 1-100Hz range. I am mostly interested in the 1-50Hz range.
- Low ampltude (a few 1-10mV).
- Always positive.

I was considering a 20uF capacitor in order to filter the frequencies < 1Hz.

What kind of capacitor would you recommend to use?
20uF is not that big and well in the range of the ceramic capacitor. It is also smaller, cheaper and has a longer life. Is there any reason to consider the electrolytic?

kalifed said:

20uF is not that big and well in the range of the ceramic capacitor. It is also smaller, cheaper and has a longer life. Is there any reason to consider the electrolytic?

Click to expand...

Right.

But why do you ask? You have any other parameter in mind?

The last parameter would be performance/quality. I see a lot of examples of high-pass filter with electrolytic capacitor, I was wondering why. I was also wondering if there was some sort of trade-off between size/price/life and quality.

What you need is a LOW PASS filter, with the pass band 0-50Hz. A good op-amp filter is what you want to use. You are seeing examples with electros because MLC capacitors in higher values haven't been around all that long. The ceramic cap is the one to choose for a filter, mainly because of the ESR.

SLK001 said:

What you need is a LOW PASS filter, with the pass band 0-50Hz. A good op-amp filter is what you want to use. You are seeing examples with electros because MLC capacitors in higher values haven't been around all that long. The ceramic cap is the one to choose for a filter, mainly because of the ESR.

Click to expand...

Yes I also need a Low pass filter. But here I'm talking about the high-pass filter.

kalifed said:

Yes I also need a Low pass filter. But here I'm talking about the high-pass filter.

Click to expand...

What do you need to high pass filter? The only place I can see in your description, is you need a 0-1 Hz high pass.

SLK001 said:

What do you need to high pass filter? The only place I can see in your description, is you need a 0-1 Hz high pass.

Click to expand...

? The whole topic is about high-pass filtering.

At these low frequencies maybe you don't care. Getting
a good manufacturer's SPICE model that includes the
electrolytic's inductance would let you assess with some
realism. But I doubt inductance matters at 1Hz. Polarity
however might, depending on application (signal, and the
surrounding circuitry - if bias can be positive or negative
then you need a nonpolarized cap, while most of the high
value bulk electrolytics are unipolar (meant for supply
filtering, not a potentially-reversing signal).

A 47uF 10V ceramic is smaller than any electrolytic I've
ever seen, if this matters to you, and will be better in
every electrical & environmental aspect. Cost, maybe
not, but you're still talking not-much.

Hi,

You say voltage is 1..10mV. Im not sure what you mean.

There is
* the signal voltage amplitude of interest (1..50Hz)
* the low frequency voltage you want to filter. (0..1Hz)
* your ADC input voltage range.

Please define all three.

Klaus

KlausST said:

Hi,

You say voltage is 1..10mV. Im not sure what you mean.

There is
* the signal voltage amplitude of interest (1..50Hz)
* the low frequency voltage you want to filter. (0..1Hz)
* your ADC input voltage range.

Please define all three.

Klaus

Click to expand...

Hi,

- The low frequency voltage has an amplitude of 1-10mV
- The higher the frequency of the signal the lower the amplitude. The amplitude decreases as a power law of the frequency. The 30-50Hz range is at least 10 times smaller that the <1Hz
- The ADC is irrelevant here because I amplify the signal before. But its range is about 20% larger than the <1Hz full range. (So the signal should not become negative)

Don't hesitate to tell me if I am still unclear.

Thank you!

Hi,

Summary.
* you need a high pass filter
* the "low frequency" you want to filter, amplitude before filtering is 1...10mV
* you are interested in the high frequency, mainly 1..50Hz, unknown amplitude
* you need an amplifier. With unkonown input voltage and unknown output voltage

* you say "always positive". After a high pass filter...DC is attenuated...AC remains...how can it be "always positive"?
Or do you mean the signal before the filter is always positive.

* You talk about ADC. But you say it is irrelevant. You want to attenuate the frequencies below 1Hz....so they don"'t reach the ADC..but you say the ADC input range is 20% larger than the attenuated signal. And what has the ADC range to do with "signal not become negative"?

In my eyes there are a lot of contradictions. I really don't know what you want to say.

Maybe you should draw some pictures. Frequency charts, signal flow, values...

Klaus

Consider that high permittivity ceramic capacitors have microphony issues due to the piezoelectric behavior, may be unwanted for low voltages signals in some cases. If you require low noise + long term capacitance stability, go for film capacitors. If some capacitance variation can be accepted, choose electrolytic.