## What kind of filter is that?

13 posts / 0 new
Author
Message

Hello, guys.
What kind of filter is that?
I could not find information on it, its properties and the calculations. Where it can be read?
Thanks.

## Attachment(s):

It's looks like a all-pass filter to my untutored eye.
Assuming the top left rectangular object is a resistor, then the input signal goes throught the resistor, through a length of wire then out the other side.
The op-amp circuit does nothing that I can see.
Is this post a couple of days early?

Is it an AF filter?

Four legs good, two legs bad, three legs stable.

Quote:

The op-amp circuit does nothing that I can see.
Is this post a couple of days early?

30+ years ago I remember Elektor publishing a "zero volt reference" circuit in their April edition. Laugh? We almost did!

A bit of analysis later and it looks like it has two poles and no zeroes. So it's a really strange low pass filter. I made an assumption that Iout = 0. I don't think my analysis will remain accurate if that is not true, but I didn't see another way of solving it otherwise.

Where did you see this thing?

I understand, that it is an active low-pass filter.
Advantage of this schematic: offset free (constant component of the signal does not influence on OpAmp) with high impedance output and steeply falling response.
Couple years ago friends give me high-voltage power supply unit from old optical system (laser source from 1980s). There was voltage multiplier (near 20kV) and stabilizer, and filter, that was build under this scheme (little more complicate in fact) to reduce the high-voltage amplitude ripple. I did not understand the principle of operation, but noted the possibility of such implication of amplifier. Yesterday this scheme appear in local forum. Some professor in university gave it to student as a research project.
This is not an audio-frequency filter.
And Â«The op-amp circuit does nothing that I can seeÂ» statement is not true.
But I canâ€™t find any literature references on this: many â€œcollections of OpAmp shemesâ€, textbooks, etc. Now I plan to use this scheme in the new project to reduce total capacity.
But I do not want to rely only on results of the computer simulations.
Are there any references to Â«analog geniuses of the pastÂ» publications? :)

Ignoramus, you mean that gerator could have two poles, according to nleahcim, and more - complex poles? And what about response - it depends from load resistanse.

Its a no-offset low-pass filter. I think that Linear Tech has one or more IC implementations of it. The advantage is that it is insensitive to the offset of the op-amp. So, there is zero drift and zero offset (at DC) through it.

While I have not done an analysis of it, I think the conclusion that it has two poles and no zeros may not be correct. I suspect that "no zeros" s true only if the op-amp has infinite bandwidth. At very high frequencies, the op-amp gain is low, so the second pole may "go away".

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

Quote:
And Â«The op-amp circuit does nothing that I can seeÂ» statement is not true.

It is totally and irrefutably true. Although it may be a reflection of my lack of understanding, rather than a flaw with the circuit.

Four legs good, two legs bad, three legs stable.

Look at the bottom of the right-hand capacitor; it's feeding into a virtual earth. Which means there must be a frequency where the impedance of the series resistor equals that of the capacitor, so it is at least a low pass filter at the output.

I made a little hand calculation. Looks like a second
order lowpass. Simulation with LTspice confirms that:

It's a second order lowpass. The exact characteristic
depends on the components, they can be used to adjust
for example the overshoot in the frequency response.

I used a -100000 amplifier in the simulation.
Behaviour at high frequencies with real OPamp
may be interesting, probably the attenuation might
be limited by some opamp parameters thus as output impedance.

I looked up on the LinearTech web site. The part I remembered is no longer there. There is a similar one that uses switched-capacitor technology.

The goal of the circuit is to avoid gain and offset problems at ADC inputs. These both affect the calibration of an ADC input. By avoiding them (where you need only filtering, and not gain), you can reduce calibration costs as well as active filter costs (no need for a low offset amplifier and tight tolerance resistors).

If you don't have DC calibration issues, then there are "better" active filters. Sallen-Key is one.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net