Robert,
I am sorry that my GIF file had an error in it.
The GIF file should have stated:

  R1=R2

(The Rs in the application notes refers the the source
resistance, which is R1 in my rendition)

If you are not particularly interested in holding the
leakage current down, then you can dispense with R2
entirely and use a direct connection instead.

I usually make R1 a variable resistor and then just
dial up the capacitance value I need. I often use a
precision 1.0 ufd capacitance for C1 instead of the
indicated 10 ufd indicated in the GIF. That gives me
1.0 ufd per kohm of resistance for R1.

By the way, Alice, thanks for giving the URL for the
National applications notes. Please note that the page
numbering for AN29 is slightly different than in my
copy. What I refer to as AN29-11 is AN29-9 in the
web site version.



----- Original Message -----
From: Robert Rolf <Robert.Rolf@UALBERTA.CA>
To: <PICLIST@MITVMA.MIT.EDU>
Sent: Friday, February 23, 2001 5:28 AM
Subject: Re: Most 'All-Purpose' Capacitor?


> The PDF file says "Rs=R3" but there is no Rs listed in the GIF that
> was attached.
>
> The text says that "Rs is the series resistance of the multiplied
> capacitance". This makes sense since R3 is in series with the output
> of the op-amp. Since this "value is relatively large, high high Q
> capacitors
> cannot be realized". However for timing and servo circuits which
> already have a series R element, this circuit is suitable.
>
> Alice Campbell wrote:
> >
> > see
> >
> > http://www.national.com/apnotes/apnotes_all_1.html
> >
> > > The caption in the GIF has an error. You say R2=R3, but
> > > the schematic shows R2=R1, which makes sense when one is
> > > balancing the bias current.
> > >
> > > If one uses a power op amp, and low values for R3, one should
> > > be able to get reasonable currents in the 'simulated' capacitor.
> > >
> > > Thomas McGahee wrote:
> > > >
> > > > Attached you will find CAPMULT.GIF which shows a simple
> > > > circuit that can be used to multiply capacitance. This
> > > > is a useful sub-circuit that can be used at the workbench
> > > > when designing circuits that involve capacitors that
> > > > have one lead connected to ground.
> > > >
> > > > This is based on AN29, page AN29-11 of the National
> > > > Semiconductor Corporation Linear Applications book,
> > > > January 1972. (There's still a lot of good stuff in those
> > > > old books!)
> > >
> > > That there is, which is why I never throw that kind of stuff
> > > away. I still have Fairchild DTL data books since some of
> > > the equipment around here still has those parts in it.
> > >
> > > >
> > > > It is particularly useful when designing timing circuits,
> > > > as you can literally dial up the value you need by
> > > > adjusting either R1 or R3.
> > > >
> > > > As shown, the circuit multiplies a 10 ufd capacitor
> > > > by a factor of 10,000 to yield the equivalent of a
> > > > 100,000 ufd capacitor. The "leakage" current is
> > > > 8 microamps max over a -55 to 125 degree centigrade
> > > > temperature range. Pretty good for 100,000 ufd!
> > > >
> > > > Note that the multiplied capacitance cannot discharge
> > > > instantaneously like a real capacitor, and it only
> > > > actually stores charge in C1, so there is a limit
> > > > to how the multiplied capacitance can be used...
> > > > but in simple RC and constant current charging
> > > > circuits, it works like a charm.
> > > >
> > > > Note also that op-amp speed, etc., affect maximum
> > > > frequency response, etc. You cannot use a 741
> > > > for such a circuit. The LF356 is a high slew rate
> > > > high input impedance opamp.
> > > >
> > > > Fr. Tom McGahee
>
> --
> http://www.piclist.com hint: The PICList is archived three different
> ways.  See http://www.piclist.com/#archives for details.
>
>

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body