Andre Sant'Anna: Great video.
Like many people have said, when you are measuring such a tiny tiny value
capacitor, then there are a ton of other stray capacitances that confound
the measurement - the probe, the wiring, the breadboard, they all have
values in the pF range so you are actually trying to do something pretty
difficult. These stray values don't affect the measurement of the bigger
capacitors much because the stray values are so much smaller in comparison.
When measuring something small like this, you might want to connect a bunch
of them in parallel (the capacitances will add) and then divide by the
number of capacitors you used to get the answer for a single one of them.
Curvian Vynes: Hey, 5.99E^-11 is not 5.99^-11. It's 5.99*10^-11.
Anyway, great video! Thanks!
Mike O'Neill: Could the error be caused by the noise induced? Noise is just another wave
form placed over the one you are looking at. So, you should be taking your
voltage readings either from the top of the noise, or the bottom. Not the
top and bottom of the entire wave form.
Flapjackbatter: "How to measure a capacitor with an oscilloscope."
Look at the time. 46 minutes. You babble way too much about things that are
Dustin Lewis: I've tried to watch a few of your videos, but you talk so much, its
difficult to string the important bits together between the babbling.
Jeff Burris: I have o-scopes but I don't have a dedicated function generator per say (as
a test instrument made for the purpose). Until I do (like tonight, now, in
a pinch ) perhaps this video can get me testing capacitance nonetheless,
using something on hand, like maybe a receiver or transmitter receiving or
transmitting a plain carrier or consistent signal (not Beethoven's 5th) :)
I'm not trying to calibrate instruments, rather, I just need to get
approximate readings on hand-made air capacitors for value. I want to make
a rotary but instead of flat vanes use concave soda can bottoms (novelty,
for antenna tuning with a coil)... can bottoms for both stator & rotor
vanes (I know I can't offset them but spindle through the middle should be
fine :) :) o man I can maybe measure inductance of the coil, too, without
buying an LCR meter this week (my multi has neither) THANK YOU
(subscribing). Between you and Ray Heffer I'm in high cotton.
The Lightning Stalker: Stray inductances are going to affect your measurement, particularly with
small value capactiors. This should be taken into account. Oscilloscope
probes also have a capacitance which will also affect the reading. Your
breadboard also has a capacitance.
TheThore: So you need a calculator for dividing by 1000? ;)
tim master: How large of capacitors could you use this technique with?
Rod Dormido: the possible source of error is that you forgot to factor in the
capacitance of the oscilloscope probe.
charles bell: Just Great !!
Phil Brown: You explained that very well ... Subscribed
Chakir El yattafti: Thank you very much. I do believe that the problem you have with 4.7 pF is
due to the reactance of the capacitor become significant big due to this
formula reactance is Xc=1/W.C the reason you did not have a problem with
47nF is because the reactance was low. your Circuit Impedance is
Z=sqrt(R^2+1/W^2*C^2), in order to compensate for the reactance of the
capacitor try to increase the frequency until the term 1/W^2*C^2 in Z
equation get smaller in order to get the impedance as pure resistive as
possible. I hope that helps.You do really great job.
Round House: Reducing the lead on the resister will also help. Nice video.
tim master: I wonder if a cage would help reduce the noise.
Barry Moore: An alternative measurement method is to use a sine wave and a two channel
scope, Place the channel 2 across the cap, and the channel 1 across the
resistor (R) and capacitor (C) in series. Adjust the frequency (F)
until channel 2 is 0.707 of the input (channel 1 value), At this frequency
the capacitive reactance is equal to the resistance of the resistor i.e.
R=1/(2*pi*F*C). As we know R and F we can calculate C. A digital storage
oscilloscope will display accurate peak voltages of the channels. The phase
angle could also be measured as it is exactly 45degrees.
Gabriele Barbaraci: finally I have seen someone explaining the things how must be done!!!!
Nouman tajik: Thanks for tutorials, It is Really Helpful, Good Luck :)
Paul Collins: I happy to say I enjoyed the video very much
Anthony Stewart: Dear MjL,
The probe capacitance must be observed when measuring capacitance <
100pF. Probe settings for 10:1 will give a higher impedance than 1:1 and
thus lower capacitance.
Ahmed Abdalsalam: thanks for your effort.
Harish Kumar: nice one! thank u
Chris Gonzales: This was a great exercise. Even with a very basic scope I got reasonable
results ( with practice ). If I had to identify a bag of small "mystery"
caps I would put a bunch in parallel to reduce the error.
Bartolomeo Cianciatella: So when the square wave is down to 0, the capacitor is discharging itself
through the resistor and INSIDE the function generator ? Is that ok ? What
would be happen with higher voltages/currents ?
I suppose you could put a diode in series, and take its voltage drop into
Just asking because I'm pretty new to these things.
Anonymous News: what is all the noise on the scope
supershwa: An impressive arsenal of tools on your workbench, and lots of know-how! I
think a 5 minute version based on the [title] of the video would be
helpful, but the long version does show you're qualified as a
Marcos Ramirez: thank you so much, electronics is amazing!!
zinou sid: thank you so much (good job) (y)
logicsystems paddy: GREAT . VERY NICE WAY EXPLAINED
Aaron Hayes: The capacitance of the cables leads and breadboard may well higher than
that 4.7pF capacitor.
Mark Cummins: I suggest The reason for the noise is because the test circuit has such a
high total impedance and there is no shielding. I suggest that the value of
the resistor and the impedance of the capacitor need to be approximately
the same to get a readable slop. In this case the resistor is about one
meg-ohm then the total circuit impedance will be more than 2 meg-ohms which
was much higher than the clean waveform test circuit impedance. You will
need shielding to get a noise free waveform. You would need to make sure to
decade resistor box was shielded as well. Or you could implement a bandpass
filter on the scope to remove the line power frequency noise and the higher
frequency noise. A much cleaner waveform could be derived if you raised the
testing frequency. This would allow a lower resistor value and because the
capacitor value would be lower in impedance the total circuit impedance
would be lower. This would load down the noise that is induced more and
thus clean up the waveform. But you would have then to be worried that the
capacitance would not be linear. That is to say the capacitance may not be
the same at 1 kHz as it is at 100 kHz. Linearity of the capacitance is
sometimes encountered with old or defective capacitors.
Round House: If you overlap vs with vc, you will see that pass the 63% mark, the current
is slowing in transit.
COUZINITROCKS: Another way to do this is to get the wave to fill up 8 divisions on the
scope then just count 5 divisions and read voltage at 5 divisions thats
your one time constant it works because 5/8 is about 62.5 percent and
is an oldschool trick from analog scope days
spectrum1844: Thanks for explaining - how to measure capacitor using scope.Are there
scopes with built in function generators too? What would benefit people
more is if you could present a simple block diagram on paper and show how u
connect the wires for doing the measurements.
jeremyhall420: Yes the capicitor is considered fully charged at 5 time constants.
sanjursan: Well, it was instructive, but it should have been titled: "How jto measure
a capacitor with an oscilloscope AND a lot of other expensive equipment."
John Miller: Persistence, very interesting video. Thanks
Thijs de Bont: The basic concept is sound, but a few comments though:
- you swapped 63.2 and 62.3 percent at some point.
- you have to take the output impedance of the signal generator into account
- the error margin of the measurement cursors is way to big at the given
- 4.7pF will be swamped by the capacitance of the cables and passive
probes. That's why this method with the given equipment doesn't work on
small value capacitors.
kimarchergy: Just a suggestion, measure the stray capacitance of the leads just as you
have them position on the bench without the Capacitor .Then insert the
Capacitor and test to get the total capacitance and subtract the stray
capacitance value to get the value of the Capacitor under test.
constance washington: its a pop quiz like ""speed"" and measurments of velocity...craisens
aaron medina: Great
James Holmes: lekker
ducklandwikeno: Earn Some more biscuit Money what are you taking about
Selim Okutan: thanks for usefull recomandations
ช่างตั้มมือถือ รับซ่อม iPad iPhone 0846726028, 0832735824, 0895010100: good
constance washington: It sounds a little weird--that's just based on the specifications already
existing--why do you have to do the measurement to a specification that
already exists if you can gather the % rate of loss through
electricity...its simple = of transmitting electricity and loss over *_*
metal...wire and transduction**Transduction can refer to: Signal
transduction, any process by which a **_____** converts one kind of signal
or stimulus into another. Here is what I am trying to say.....mmm....that
is specifications and so is = gains --circuit breakers...and
loss---connector-wire-gage-type of metals conductive to energy
transfer...complete circuit...so your calculation would have been already a
specification in the =....im not sure if I said that right ?.......so
physics--mechanical--signals--amplification--are all measurements in the
specifications already included--even in thought**.
movax20h: Instead of using 63.21% (t = rc), you can use 86.47% (t = 2rc), because
slope of voltage on capacitor is smaller there, so it is easier to find out
exact spot for it. After that you do the same, but devide result by two. c
= t/(2r). Also it is use as wide time division as possible.
Other value to try is 75.00% (t = 1.3863 * rc), or 80.00% (t = 1.6095 * rc)
which should be also easy to setup on a scope.
Another approach is to take data to computer, and fit exponential function
(or linear after taking logarithm). This should be much more accurate and
also give you estimation of error.
Frank Berry From Upstate NY: Think about adding some "adsorbing dark materials" in your videos...you're
awash in white illumination Bro!
Greg Cunns: Very good. I'm curious, did the probe cable capacitance affect the accuracy
the 47pF test?
As some one else also asked, can you do a ESR tutorial with a scope and
TheDkjayjay: hey, u are cooler than cool
got myself a Tektronix TBS 1022 and i love it.
School: Thanks for teaching us how to fix it. The problem with most Medical assistant in california is usually the cost of education. Hopefully it'll pay off. Thanks again.