[QUOTE="SHA_S, post: 1410547, member: 632721"]I repeat my first question again.
What losses does this tweezers demonstrate in ceramic capacitors of a small capacity (1-4 pF) at a frequency of 10 kHz?
[/QUOTE]

Here's a measurement of a small (3 pf nominal) ceramic cap at 10 kHz on a bench LCR meter. It was measured with an excitation voltage of 5 volts to minimize the effect of the instrument noise--the measurement current was just a little less than 1 uA. Notice that the value of D is much less than .001 which is a difficult measurement. I'll measure that same cap on a Pro1 later:



Measuring the loss in such small caps is probably better done with a Q meter, but at higher frequencies.

 

Here's a measurement on a nominal 1 pF ceramic cap:

 

Here is the measurement of the 3 pF capacitor from post #42, made on the Pro1 Plus at 10 kHz:



Here is the same measurement, but at 100 kHz:



The Pro1 got a value for D that is substantially in error. The user's manual gives no specification as to the accuracy of the D measurement, but measurement of D for a very low-loss capacitor is difficult even for a highly capable bench model instrument.

 

Considering the 3 fundamental components of circuits, resistors, capacitor and inductors, we have 3 equations relating the voltage across to the current through each. We use upper case for DC voltages, and lower case for AC voltages:

For a resistor, E = I*R

For a capacitor, i = C*de/dt

For an indcutor, e = L*di/dt

https://en.wikipedia.org/wiki/Inductor

v(t) = -L di/dt (you forgot the minus sign)

 

https://en.wikipedia.org/wiki/Inductor

v(t) = -L di/dt (you forgot the minus sign)

Depends on whether you're talking about electron current, or conventional current.

 

Depends on whether you're talking about electron current, or conventional current.

Th en wouldn't the capacitor equation be wrong?

 

Th en wouldn't the capacitor equation be wrong?

Not if the same convention is used for both inductor and capacitor.

 

Should one have a "-" and the other a plus if the convention is changed?

 

Imagine you have a capacitor with one plate grounded, and cause a flow of electrons onto the ungrounded plate; this would result in a negative voltage with respect to ground, whereas a conventional current directed onto the ungrounded plate will result in a positive voltage on that plate. I suppose this means that the appropriate relationship would be -i = C*de/dt for an electron current. I would also suppose that the 3 equations I gave, which are common usage when doing circuit analysis, could be made appropriate for electron flow by simply replacing i by -i.

See: https://www.physicsforums.com/threads/inductor-voltage-drop.539520/

Further discussion of this topic belongs in a thread of its own. If you were to start another thread, a moderator could move the relevant posts so far to that thread.

 

The Electrician, Is it possible to measure the loss in the capacitor 1-4pF at 100 kHz?
P.S. On a normal device naturally.

 

The Electrician, Is it possible to measure the loss in the capacitor 1-4pF at 100 kHz?
P.S. On a normal device naturally.

My ancient Q meter can do that. It can resolve less than 0.1 pF and can measure loss. I don't offhand know how much loss it can resolve. Of course the Q meters can measure at frequencies into the VHF range as well as audio. For very low frequencies an external generator is required.

 

My ancient Q meter can do that...

Photo of losses in the capacitor at 1-4pF at a frequency of 10 kHz, from the ancient Q meter can you show?

 

Photo of losses in the capacitor at 1-4pF at a frequency of 10 kHz, from the ancient Q meter can you show?

I don't think so. First, that frequency is low enough that it is outside the range of the internal oscillator so I'd have to bring in a signal from a separate oscillator. The way it works, the unknown capacitor is set up to resonate with a standard coil. The coil has to have a high enough Q that the capacitor losses can be discerned. The circuit is resonated without the unknown and the Q and resonating capacitance read from the dials and meter. Then the unknown is connected and a new set of readings taken.

The reduction in Q is an indication of the capacitor losses. The new setting of the resonating capacitor is subtracted from the original setting to determine the capacitance. There are two dials, coarse and fine. The fine dial has a range of about 6 or 7 pF so can be read rather closely. The Q is read on a meter. I have measured Q of large inductors at HF to be around 500. A lossy capacitor should drop that figure somewhat and the amount of loss can be calculated. There is a delta Q dial that can be used to measure small changes in Q. Most capacitors will not drop the Q significantly, although ordinary paper capacitors will if the frequency is high.

So it's not a plug and play setup. You have to know what you are doing and be careful how you connect the parts. A loose connection will drop the Q. But it's all basic electronic theory and the Q meter is simply a setup that can be used to make the required measurements with less trouble than otherwise.

Unfortunately the Q meter is fairly large and takes up a big piece of the bench. It measures Q by measuring the voltage stepup at resonance. There are, I think, six frequency bands to cover the range from around 50 kHz to maybe 100 MHz but don't quote me. I don't use the instrument often. There were standard coils made for the unit but I don't have them; they make capacitor losses easier to measure and also provide a calibration for Q. Nonetheless the utility of the instrument remains. It's useful for seeing the parasitic reactance of resistors, for instance, which is well documented but still important to measure for high frequency applications. Wire wound vs carbon composition vs metal film etc.

 

It is not easy to measure losses in a capacitor on an ancient Q meter, modern devices make it easy, but not always correctly. To do this, i asked The Electrician about the possibility of measuring the loss in the capacitor on a serious device.

 

@SHA_S To tag someone on AAC so that they get an alert, use the '@' along with the member's name.

 

New LCR tutorial thread has been created here:
https://forum.allaboutcircuits.com/threads/lcr-meter-inplementation-and-limitations.173587/

 

excellent. I've got a lot to say about it!

 

No any ever capacitor have better tan(delta) as 0.0001. No any have more bad as 0.01.
Q-factor approx is 1/tan(delta). Thermal flux is approx Circulating power in cap (or tank) multiplied with tan(delta).

By the way, I did measurments of many shelve saying smd capacitors range 1...1000 pF at 2 kV and 99% of them so low as over 5 MHz reverted to brilliant coils with no any signs of capacitance. The best few was able up to 100 MHz but over none. Then I realized why damn most of caps at Octopart cost 1-2 cent per piece but some visually identical cost 5...15 USD piece.

 

RE:""Is it possible to measure the loss in the capacitor 1-4pF at 100 kHz?""
One way is robotized "all in one measure" China wonders with ARM inside - limits stay near the 0.1 pF and 1 MHz. Recognizes what that is on the wire and shows all what is measurable for that. Yet rather weaky about any electrical stress, I have burn few just cause forgot to shorten electrolyte feet before measurement.
Other way is robotized Chinese wonders labelled VNA (as nano-VNA as SA1201). The upper limits are about 3500 MHz and 0.05 pF. Note the accuracy then is much higher than for first.
Third way is occillo but that is damn nightmare, and not accurate
Fourth way is dipmeter but is nightmare as well, however better accuracy
Fifth way is Q-meter, the brilliant instrument but not for sub-pF

 

Janis59, there is also a correct way. Take normal tweezers NV-14.