The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: Chuck Simmons
Organization: You jest.
X-Mailer: Mozilla 4.61 [en] (X11; U; Linux 2.0.33 i586)
Subject: Re: OP Amp Output Resistance
References: <3D8D30C9.firstname.lastname@example.org> <3D8D4F02.991FC7CA@webaccess.net> <3D8FD91D.2FB63DA8@nospam.com> <3D9073C1.CB29DE92@nospam.com> <3D90807F.C0E4CE42@webaccess.net> <3D90853F.C5598EF7@nospam.com>
Date: Wed, 25 Sep 2002 00:58:23 GMT
NNTP-Posting-Date: Tue, 24 Sep 2002 17:58:23 PDT
Fred Bloggs wrote:
> Chuck Simmons wrote:
> > Fred Bloggs wrote:
> > >
> > > carltons wrote:
> > > >
> > > > In article <3D8FD91D.2FB63DA8@nospam.com>, Fred Bloggs
> > > > wrote:
> > > >
> > > > > carltons wrote:
> > > > >
> > > > > > You are absolutely right. The output is non linear. I have answered this
> > > > > > question many times with non-engineers who don't seem to understand how
> > > > > > the output impedance can go down with something simple like feedback. I
> > > > > > always say with confidence that the feedback causes the output voltage to
> > > > > > try to stay at the same level no matter what the load, which would not
> > > > > > happen without the feedback and all I get are blank stares.
> > > > >
> > > > > Why don't you try explaining that the feedback amplifier is amplifying
> > > > > the scaled error between input and output in such a way as to drive it
> > > > > to zero, and then since the feedback tap is at the output, the voltage
> > > > > drop due to output resistance becomes part of the error and is therefore
> > > > > nulled too. You don't need calculus to understand this basic concept.
> > > >
> > > > I've tried that also to no avail. You can show the offset voltage and how
> > > > it is amplified by the huge gain of the amp and you will still get that
> > > > weird stare in response. I do believe that you have the right approach
> > > > here, but I think that most of the people don't get it until long after it
> > > > is explained, which is okay with me because I want to explain things and
> > > > if there's a time lag, so what. The goal is to educate, after all.
> > > >
> > > > Steve
> > >
> > > Well- operational amplifier theory and understanding is intimately tied
> > > to feedback theory in general. You might try illustrating feedback
> > > principles with some of the more easily understood historical examples
> > > from servo mechanism theory. There is a wealth of interesting circuits
> > > that were actually used at one time involving positioning systems using
> > > batteries and potentiometers with feedback driving wipers for voltage
> > > null. These types of circuits are always interesting, ingenious, and
> > > readily grasped- as long as you avoid minute details like stiction,
> > > contact resistance non-linearities, large excursions, and time
> > > constants. Static graphical analysis is always fun too and is a very
> > > powerful way of illustrating the linearizing effect of the feedback.
> > > Your presentation will be on the sterile side if you omit the historical
> > > background which drove the development of the theory and devices to
> > > solve some very important practical problems.
> > You have to be really careful about selecting historical examples
> > especially from servo systems. Carrier control servos were common from
> > perhaps 1940 until into the 1970s (there was at least one model of HP
> > plotter that used 400Hz servo motors and carrier control). Collins Radio
> > built high power transmitters for the US Army in the 1950s and 1960s
> > that had hysteresis stablized tuning servos (non-liear control).
> > Tachometers are still found in some position servos. Aren't you afraid
> > that someone might ask what the devil the generator is for? Also, the
> > simple minded view of a torque motor, amplifier and pot to null against
> > a reference is obviously only conditionally stable (like a needle
> > standing on its point). A first year physics student can see that
> > without writing any equations. The internally compensated opamp is
> > incredibly simple and tame compared to "simple" servo systema.
> > I would much rather explain opamp applications and operation than
> > explain any typical position servo. If a person is not clear on opamps,
> > about the only thing you can say about most servos is, "Trust me, it
> > works."
> > I really had to mention this because I once spent more than an hour
> > explaining to an engineer why a very simple servo I designed worked. He
> > could not deny that it worked because he had tested it. He could not
> > understand how my servo controlled to better than 0.05 microns when your
> > static analysis would show that I could not possibly control to better
> > than 0.5 microns. People notice things that look strange.
> I agree that a state-of-the-art true servo mechanism is outlandishly
> complex, but so are operational amplifiers outlandishly complicated and
> non-linear. If you can condense the op amp model into the
> super-simplified linear ideal, then you can do the same for most
> servo's. Everything is linear when the inputs are small enough. The
> advantage of mechanical servo examples over electronics is that you have
> something composed of parts for which nearly any engineering student has
> a feel. By servo I mean the essential "slave" model of small input
> causing something else to "slave" to it- and not necessarily carrier
Oh, but most common servo systems are designed with the assumption of
perfect linearity. This may sound silly but when I design a VCM servo, I
jot down F=ma because that is my VCM. The problem I pointed to was that
in a position servo, if you make F exactly proportional to position
error, you have an oscillator. The system is perfectly linear and quite
obviously only conditionally stable. That is an ideally linear position
servo and it won't work. If you idealize an opamp, connecting the minus
input to the output is obviously unconditionally stable.
The difference is in the dynamics. Although an internally compensated
opamp looks more like an integrator than an amplifier, in simple
non-reactive circuits it is pretty close to the ideal infinite gain
opamp. DC motor position servos usually have a double integrator rather
than a single integrator and are, therefore, only conditionally stable
if the feedback is not reactive. A sort of basic difference even when
both are perfectly linear. You can't really ignore F=ma.
... The times have been,
That, when the brains were out,
the man would die. ... Macbeth
Chuck Simmons email@example.com
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup