Newbie OTA question

[Zoomby]

Hi,

I'm trying to understand a single 1-pole filter from this circuit: http://www.uni-bonn.de/~uzs159/rs20.png. I think everything is clear, except for the negative OTA-feedback. I read it is used to make the filter unity gain, but how are the feedack resistors chosen? If I compare it with opamp feedback, there seems to be other rules. Can someone explain it?

Bye,
Chris

[blue hell]

Chris.

[Zoomby]

Hi Blue Hell,
thanks for the reply. Yes, I read the article before. But it only teaches me how the differential input is converted to current. No information about negative feedback. My problem is I'm new to electronics and there seems to be no tutorials about OTAs in the internet that are useful for the newbie hobbyist. What I like to know is, i.e. why is there a voltage devider used in the right lowpass-filter to get unity gain? In contrast to an opamp where you use direct connection to get unity gain (like a buffer). I'd be glad if you'd give me some hints to understand it. Or some other newbie tips which guide me in the right direction.

Bye,
Chris

[zipzap]

Hi
Well, i feel i´ve learned so much in the last year (can go fast sometimes), but this is still a mysterie to me. Just this much: A Ota can not really be compared to an opamp when it comes to feedback and buffering. That´s where things go different. With an opamp the buffering can be done with a resistor too, one one the input, one of the same size in the feedback.
Lot´s of confusing stuff...
As i said, i can´t understand the function of the ota in the shematic, so i can´t explain.
This article might help a little. I´ve learned a lot about opamps, diodes and transistors from these tutorials and the forum.
Sorry guys around the world, those are only for the ones from the land of the poets and the thinkers - that´s what germans call themselfes - crazy bullshit, isn´t it?
http://www.elektronik-kompendium.de/public/schaerer/otalim.htm
For more basics check your library. Francis has some good books for example, i have better understanding reading on pages than on the screen.
sers

[Zoomby]

Thanks for the link zipzap.
I noticed the OTA article from "Das Elko" before, but I thought it was to advanced. I'll give it a try.
Maybe someone can enlighten us about OTA feedack anyway.

[blue hell]

It's not really easy to explain, but I'll have go at it.

First off all the unity gain has to be set not just over the OTA but over the structure consisting of the OTA the capacitor and the opamp (the latter just being a unity gain voltage buffer here).

In the rightmost structure (*1) when you would leave out the the resistive divider and apply a direct feeback to the OTA's negative input the voltage on the OTA's negative input would simply be too high and it would produce distorion.

The reason why this is possible is that the structure's voltage gain can be set to be arbitrary high by choosing a sufficiently low value for the capacitor (as long as the capacitor has good quality). So one can afford the losses in the voltage divider, and they are needed as well to avoid distortion.

In order to see why the structure's gain can be made arbitrary high imagine the capacitor to be a frequency dependent resistor (with a value of 1/(2.PI.f.C) ). For high frequencies it will have low resistance and for kow frequencies it will have high resistance.

The lower the capacitor value the higher this resistance will be (for any given frequency). As the OTA's output is a current the voltage can be arbitrary high when the reistancence is choosen arbitrary high (as V=IR), and hence the capacitance arbitrary low.

So it is a sort of a trade off really, the bad input properties are compensated by an other property of it, the possibility to have arbitrary gain in certain circuits.

Opamps usually have better inputs that can be driven over the full supply voltage range without problems, so for opamp circuits you'll not usually see such constructs.

The part where I said that a capacitor can be regarded to be a frequency dependent resistor is not really theroretically sound, but its a simplification that serves its purpose for this part of the circuit's explanation - not for the entire workings of the cicuit though.

Does that help ?


(*1) For the leftmost circuit it's complicated a bit by the resonance control feedback, but the same reasoning still applies.

Edit : for the capacitor's resistance also see : http://www.st-andrews.ac.uk/~jcgl/Scots_Guide/info/signals/complex/react.html

[Zoomby]

Thanks for the explanation Blue Hell! Makes it a little bit clearer. I think I have to accept it for the moment, and concentrate on more basic stuff.
BTW. Does someone know a good source for learning to understand feeback in general? (not only basic opamp circuits, but something to better understand what's going on under the hood)

[piedwagtail]

A fine book to try and get secondhand is
the IC Opamp Cookbook by Walt Jung.

Full of schematics and with definitive discussions of opamps,OTAs and current differencing amplifiers ie.the venerable LM3900 of arp filters and serge waveshapers.

'Under the hood' is a red herring only chip designers need to know that,opamps are often extremely complex but all try to obtain the ideal blackbox.

Here's the first two pagesupsidedown(1974 er .....copyright)which set out the ideal blackbox assumptions.


(Note on one side we have an original T.Henry publication...3080 book which is the ultra description of OTA mechanics including his discussion of absolute room temperature,this is available from magicsmoke?
And on the other some details of Cocoa futures traded in Lasalle st Chicago..... )

Robert

---

pict1327.jpg
Description:
Filesize:	426.03 KB
Viewed:	218 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

---

pict1328.jpg
Description:
Filesize:	426.22 KB
Viewed:	231 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[Zoomby]

Thanks for the tip. I'll try "IC Opamp Cookbook". Ordered it very cheap from amazon.