From: Sar Saloth (sarsaloth_at_yahoo.com)
Date: 2002-03-08 15:30:39
Hi Andreas,
>To be honest, the control theory course was something I just took and
>forgot. :-p This is Joerg Hansmanns design and he has done thorough
>SPICE simulations (and real testing? Joerg?) that shows it is OK,
>though not with the opamp I'm using. Personally, I raised the same
>question as you do a few months back and maybe I should do
>simulations of my own. Unfortunately, I'm swamped in work on other
>parts of this OpenEEG project, so I'll have to ask you to show me the
>hard numbers. ;o)
I had two courses in control theory, but to be honest it was mostly algebra
and I never quite got it down pat as my algebra wasn't adequate. My
understanding of stability of the circuits is based upon experience, even
though I haven't been keeping up to date these last few years, which I will
use as an excuse for not knowing about those OpAmps.
My "experience" also tells me that SPICE simulations on a regular OpAmp
models that are excessively loaded down will not give meaningful
results. It really depends on the models, and for many OpAmps, the people
that designed the OpAmp Spice models wouldn't be concerned with that regime
of operations. Few models are on the transistor level, so the area of
validity of the model depends on the model and how it is used.
Since the data sheet for the LT1114 doesn't claim stability for that type
of circuit (47uF direct output load at unity gain), I don't why it would be
worth the time to prove that it is unstable. I would rather spend the time
finding a component that does claim stability in that type of circuit.
To honor your request for not wasting your time, I will volunteer to do the
following two things since I am to spend the next few weeks searching for
components anyway
1. provide a circuit with 3 extra components that will work with ANY OpAmp,
but can be tuned or by using a jumper would be identical to the current
circuit. The regulation performance at mid-frequencies might be
compromised, but then again, by adding the components it can be played
with. Regulation is something that I might be willing to simulate.
2. Search for the least expensive component that claims stability in that
type of circuit.
Speaking of simulators, I have been somewhat disappointed. I had been
using Electronics Workbench 5 which I found to be quite accurate (but I
never did go into "strange" regimes). However, EWB 6 now renamed to
Multisim2001 has been totally rewritten and I have produced completely
impossible simulations. I gave up after that. That is annoying as it is
the 2nd or 3rd "CAD" program I have paid for. I wonder if my license for
MentalCad and for EWB are transferrable? I could donate those.
<snip>
> > TI makes a part that is a "virtual ground rail", one version of
> > which is a constant 2.5V or so above the negative rail and the
> > other part is a rail-splitter. I have never used either of them.
> > They were supposed to be around a dollar and they are designed for
> > this sort of application.
>
>Yup it is a nice circuit. Unfortunately, the wideband noise figure
>indicates it is suitable for 12-bit applications. Correct me if I'm
>wrong, but any high frequency noise on the virtual ground will be
>injected directly into the signal path as common mode noise.
Yes, that is a very important point, thank you.
Do those parts tolerate large capacitive coupling? How does that impact
the total noise figures. I have no need for those parts, so I won't
investigate them. Since there seem to be OpAmps capable of capacitive
loading, I don't see the use for these parts.
> > What is the reason for not using commercial instrumentation
> > amplifiers? TI has a new low voltage single-supply CMOS InAmp
> > that uses very little power (probably only in SMT) They send
> > free samples.
>
>It is because the ADC uses differential inputs, so the signal path
>from electrodes to converter can be kept differential. The other
>designs use regular in-amps (LT1167 and INA114).
Thank you.
I appreciate the education
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:40 BST