[Prev][Next][Index][Thread]

Re: Measuring Capacitance

    [The following text is in the "ISO-8859-1" character set]
    [Your display is set for the "US-ASCII" character set]
    [Some characters may be displayed incorrectly]



----------
> From: Thomas McGahee <tom_mcgahee-at-sigmais-dot-com>
> To: tesla-at-pupman-dot-com
> Subject: Re: Measuring Capacitance
> Date: Sunday, February 02, 1997 10:34 PM
> 
> 
> 
> ----------
> > From: Tesla List <tesla-at-poodle.pupman-dot-com>
> > To: Tesla-list-subscribers-at-poodle.pupman-dot-com
> > Subject: Re: Measuring Capacitance
> > Date: Sunday, February 02, 1997 8:25 PM
> > 
> > Subscriber: sfalco-at-worldnet.att-dot-net Sun Feb  2 17:39:56 1997
> > Date: Sun, 02 Feb 1997 10:12:16 -0500
> > From: Steve Falco <sfalco-at-worldnet.att-dot-net>
> > To: tesla-at-pupman-dot-com
> > Subject: Re: Measuring Capacitance
> > 
> > > By the way, methods #4 and #5 can both be checked for error very
> > > simply: After measuring the voltages across each component, add these
> > > values together and compare them to the actual measured voltage
across
> > > BOTH components. If they don't add up within a few percent, then your
> > > meter is loading down the circuit.
> > 
> > I think this error check works for method #5 but not for method #4. 
> > When I measured a 550k resistor and a 4500 pF capacitor in series
across
> > 120v -at- 60Hz, I read roughly 0.8 volts across each component.  I believe
> > that this is due to the different phase shifts across the resistor and
> > capacitor in circuit #4.
> > 
> WHOA! .8 volts? This isn't a phase problem, this is a current problem.
> 
> The frequency is too low for this size capacitor! Your meter is loading
> down the circuit because it is gobbling up all the current. The current
> THROUGH the two elements must be MUCH greater than the current diverted
to
> the meter or you will get just such weird results! It is NOT a phase
> problem but a current problem. What kind of meter are you using? Most
DVMs
> use a 10 Megohm input impedance. At 100 volts that corresponds to 10
> microamps of diverted current. For example, if you were to measure the
> voltage across two 10 meg resistors in series with 100 volts you would
get
> a reading of 33 volts instead of the actual 50 volts. Your very low
> readings tell me you are NOT using a DVM. Your meter's impedance is
> gobbling up all the available current and giving you really gross errors.
> 
> Always try to make the current flow THROUGH at least 10 times what you
are
> diverting. And if you are trying to measure the voltages with a
D'arsonval
> type meter, be aware that they usually have a rating of about 10K/volt.
In
> other words, it diverts about 100 microamps for a full scale deflection.
> Now do you see why you get readings of .8 volts instead of 60 volts?
> 
> Use a DVM, and if the two don't add up to the total, then use a sine wave
> generator and go for a higher frequency. Just make sure the meter is
> designed for the frequency you use. And don't use a D'arsonval meter!
That
> sucker adds inductance into the circuit! 
> 
> > As another example, the book I got this technique from had a 4k
resistor
> > and 0.7 uF capacitor in series across 115 volts, with Vr = 80 and Vc =
> > 100.  Thus the voltages may add up to less than the line voltage, or
> > more than the line voltage, again because of the phase shifts.
> > 
> 
> The author made a boo boo. Trust me on this one. If this was a resistor,
> coil and a capacitor circuit, then phase effects are definitely involved.
> But NOT with just a resistor and a capacitor. Try using a high impedance
> DVM and see what a difference you get. I bet the author was using a meter
> that added inductance to the circuit. Thus when he made the measurements
he
> was actually changing the circuit from a RC circuit to an LCR circuit,
> which is a different type of animal entirely.
> 

Sorry, *I* am the one who made a boo boo here! It is true that the sum of
the voltages as measured by a meter will not always add up to the source
voltage. I wrote my reply at about 10:30 at night, I was tired, and I
didn't bother to read what I wrote with a critical eye. When I read my own
post this morning I couldn't believe that I said the phase made no
difference. Of course it does. Steve, my apologies on this one. When I'm
wrong, I'm wrong! Only in a purely resistive circuit can we ignore phase. I
tell my students that all the time, but that doesn't mean that occasionally
I don't forget to apply that knowledge rigorously. My apologies to anyone
else out there who got confused because of my erroneous reply. 

I ran an experiment with a 560K resistor in series with a 4700pf, which is
the closest I had to the values mentioned above. With a 10megohm impedance
DVM I got 
Er=90V and Ec=69V   at 120 VAC 60Hz.  Then I tried putting a 4700 ohm
resistor in parallel with the DVM to cause it to have an effective
resistance of 4700 ohms and measured again. This time I got both readings
to be about 1.1 VAC. This leads me to believe that the meter originally
used to get the .8 volt reading must have been on say the 5 volt scale of a
1000 ohms per volt multimeter.

So it looks like my paragraph about current loading is in fact correct, but
I was wrong about the phase thing.

> > 	Steve Falco
> > 	sfalco-at-worldnet.att-dot-net