The circuit I have is simply a comparator, 1 resistor and 1 capacitor - It seems very simple, but I can't think of the correct search string for Google to find more info.

The 2 comparator inputs are joined by a resistor and one or other of the inputs has a capacitor to GND.

The general idea is; the voltage on the capacitor becomes Vref, but because of the path through the resistor to the input, the capacitor voltage tracks the input - I can't remember if that's average, mean or whatever, but you get my "drift".

The whole point being; the comparator ignores slow changes and only responds to events that deviate rapidly from the drifting level.

All I need is; what its called so I can Google.

Thanks.

 

Not sure if it has a name.
It's a fairly simple circuit.
What aspect of it do you need to Google for?

 

I know what you're thinking about but I don't know a name for it. I wouldn't say the capacitor tracks, I would say it lags. The first duty is to make one input lag behind a change so the change triggers the comparator to change its output. The second duty of the capacitor is to track the voltage to the peak voltage so the first input no longer trips the chip. This would mean the compartor does not respond to voltage changes in one direction and it responds to voltage changes in the other direction, but only for an amount of time until the charge on the capacitor catches up to the input voltage.

 

Also not sure if the whole thing has a name, but try googling 'moving average circuit threshold'.
Presumably something like this is used in the ubiquitous PIR sensor.

 

Not sure if it has a name.
It's a fairly simple circuit.
What aspect of it do you need to Google for?

The circuit came from a collection of appnotes, it may have been the NS linear applications handbook, maybe the Jim Williams collection or The Art of Electronics book.

Pretty sure it was named; "data slicer" - but Google doesn't recognise that.

 

I know what you're thinking about but I don't know a name for it. I wouldn't say the capacitor tracks, I would say it lags. The first duty is to make one input lag behind a change so the change triggers the comparator to change its output. The second duty of the capacitor is to track the voltage to the peak voltage so the first input no longer trips the chip. This would mean the compartor does not respond to voltage changes in one direction and it responds to voltage changes in the other direction, but only for an amount of time until the charge on the capacitor catches up to the input voltage.

Obviously a capacitor is going to lag any change, but AIUI: the time constant needs to be a *LOT* slower than the events of interest.

The trick is probably; making the lag long enough - but no too long. Too much lag could trigger the comparator in response to drift rather than events to be separated from the drift.

 

Input on non inverting. Output, through a resistor, to inverting and a capacitor to ground.

It is an amplifier.

 

No, it isn't. There is no resistor from the output to either input, which you can see on...wait...um...

Note to the TS with 3,645 messages -- nice schematic.

It is a comparator (see post #1) with the reference input derived from the long-term average value of the signal input. This is a circuit that is not in very many app notes, but most analog designers re-invent at some point. In the pre-internet days, I thought I was a genius. Later, in the post-internet days, I tracked it back to the 40's. I used this to detect syllable boundaries in speech audio. It way outperformed an AGC followed by a static comparator.

ak

 

The circuit I have is simply a comparator, 1 resistor and 1 capacitor - It seems very simple, but I can't think of the correct search string for Google to find more info.

Sounds like a, DC offset elimination circuit. Figure 10, page #9 of this app note - http://cache.freescale.com/files/microcontrollers/doc/app_note/AN3552.pdf

 

Obviously a capacitor is going to lag any change, but AIUI: the time constant needs to be a *LOT* slower than the events of interest.

The trick is probably; making the lag long enough - but no too long. Too much lag could trigger the comparator in response to drift rather than events to be separated from the drift.

Data slicer?

 

Sounds like a, DC offset elimination circuit. Figure 10, page #9 of this app note - http://cache.freescale.com/files/microcontrollers/doc/app_note/AN3552.pdf

Seem to be getting nowhere fast!

Found the circuit in that AN and copied the description into google.

The results were a closer match than my first attempt - but still pretty much all over the place.

 

No, it isn't. There is no resistor from the output to either input, which you can see on...wait...um...

Note to the TS with 3,645 messages -- nice schematic.

If you can't visualise 3 components without a schematic..........................................

 

Not sure if it has a name.
It's a fairly simple circuit.
What aspect of it do you need to Google for?

Anything I can find out about it.

 

Anything I can find out about it.

What part of the two circuits that match your schematic don't you like?
In other words.... What do you want the circuit to do? Or are you in search of a name you recognize?

 

If you can't visualise 3 components without a schematic...

I'm not the only person on this forum.

ak

 

What part of the two circuits that match your schematic don't you like?
In other words.... What do you want the circuit to do? Or are you in search of a name you recognize?

The original appnote I found it in was just the circuit and the shortest possible description.

There were no example calculations for the time constant or how to dimension it for drift rate and pulse rise time.

I don't care what its called - I just need something that Google recognises.

 

https://www.google.com/search?q=dat...#tbm=isch&q=data+slicer&imgrc=JnCkOZoZi5uSyM:

 

I don't care what its called - I just need something that Google recognises.

English not being my birth language, I can perfectly relate to what you're saying. Sometimes I spend more time thinking about the right words for my query than reading and evaluating the results afterwards.

 

English not being my birth language, I can perfectly relate to what you're saying. Sometimes I spend more time thinking about the right words for my query than reading and evaluating the results afterwards.

The results I got for; "data slicer" were mostly about medical stuff.

The link posted by ronv is a *VERY* random collection of images - a couple I found in all that lot were the schematic I already have.

Sometimes Google *IS* broken.

 

The results I got for; "data slicer" were mostly about medical stuff.

The link posted by ronv is a *VERY* random collection of images - a couple I found in all that lot were the schematic I already have.

Sometimes Google *IS* broken.

Click on the pictures:
https://www.maximintegrated.com/en/app-notes/index.mvp/id/3435
http://www.dabrahams.net/site/?page=projects-articles-rfdata
http://electronics.stackexchange.co...-circuit-to-generate-a-digital-pulse-from-ldr
http://electronics.stackexchange.co...ing-the-output-of-an-opamp-into-a-square-wave
What frequency do you want to decode?