Finding the Most Accurate Capacitance Measurement
HTML-код
- Опубликовано: 27 июл 2024
- Do you ever have trouble getting accurate capacitance measurements from your LCR? You think it’s easy to measure capacitance? Well think again…You don’t want to fall in this trap! We have found a tip to always get the most precise measurement every time! Watch the video to learn more.
Have a technical or sales question? Contact us here: www.kemet.com/en/us/technical... 
Наука
![Finesse2Tymes - Pretty Ricky [Official Music Video]](http://i.ytimg.com/vi/xIoP_mhYRT0/mqdefault.jpg)
This video is highly misleading. The issue is NOT that the meter misreads the capacitance, it’s that the capacitance of these MLCC capacitors varies so much with both DC offset AND AC waveform amplitude. The supposedly inaccurate meter reading is actually a correct reading, just under conditions different from the ones that the “nominal” capacitance value is specified for. Here that standard condition is 0V DC, 1VRMS AC, at 1kHz applied _at the capacitor_.
The statement at 0:50 “this [incorrect measurement] is happening due to the Class 2 dielectric exhibiting really low impedance” (15.9 ohms). is just wrong. _Any_ actual 10uF capacitance has an impedance of 15.9 ohms, that’s just Xc = 1/(2 pi f C). It has nothing to do with the special dielectric. What the special dielectric _does_ do is make the capacitance vary widely under different AC (and DC) conditions.
The meter uses (1:01) a simple R-C voltage divider to determine impedance, so of course, with a particular Vsource, the AC voltage across the capacitor will differ according to capacitance; that’s the whole principle that the meter uses to measure the capacitance! And that would work fine for any variety of capacitor whose capacitance didn’t vary with AC amplitude. But since MLCC capacitance _does_ vary with amplitude, and therefore require applying a specific AC voltage to reproduce the spec-value test conditions, it’s necessary to increase Vsource accordingly.
Then at 1:25 "and now we see a much more accurate reading".. also wrong. It's no more accurate than the first reading. It's just that the capacitance has changed due to different test conditions, and those conditions match the standard conditions that the manufacturer uses for specifying the nominal capacitance. Indeed, someone who cares about the capacitance value enough to measure it with fancy equipment actually _should_ measure the cap under a variety of applied AC and DC voltages, just to convince themselves how wildly this type of capacitor varies.
And with regard to the title "Finding the most accurate capacitance measurement" -- it's hard to see how this has anything to do with "accuracy" when the component behavior varies over almost an order of magnitude from zero DC bias to rated voltage (not to mention AC and temperature variations). This is almost nothing to do with accuracy of capacitance value, and all to do with "how to achieve the very particular DC and AC testing conditions that happen to match the manufacturer tests"... but which are unlikely to match how you will use the component.
Hello Graham. Thank you for your comment. What we were trying to do with this video is to demonstrate why some customers measure a Class II or III MLCC's capacitance and the capacitance reads out of spec compared to what the manufacturer specifies. The point of the video is that the MLCC has to be measured under the same conditions that the manufacture measures them at or else they will be out of spec per the reasons that you mentioned in your comment.
@@KEMETElectronics Sure, I get that you were motivated to produce a video that would help customers validate whether the components they received correspond to the spec they ordered. But the main reason that customers would get a capacitance reading that doesn't match the spec is because they have the wrong expectation of how these capacitors behave. And that's a significant problem if they ordered capacitors to populate a board that they designed while unaware of MLCC characteristics. That is the topic that needs illuminating. This video may walk through the exercise of applying test conditions that match the spec, but its incorrect discussion of the mildly theoretical details actually undermines the customer's understanding of the crucial topic of the wild variation in MLCC capacitance under different conditions.
I will repeat more strongly: The statement at 0:50 “this [incorrect measurement] is happening due to the Class 2 dielectric exhibiting really low impedance” (15.9 ohms) is a lie. The measurement is not incorrect. It is not something specially attributable to the Class 2 dielectric. It is not a "really low impedance" in this context, it's the exact impedance that this nominal capacitance would exhibit. And given that "really low impedance" could be construed as a desirable feature (as in "Low ESR"), that is a really disingenuous rationale to suggest.
So while you might have had a fine intent when you decided to produce this video, it instead works to undermine a customer's understanding of what's really going on, and does so in a decidedly self-serving manner.
Hi Graham. Thank you for the video suggestion. We appreciate your comments.
@@Graham_Wideman Indeed sir, thank you for your insight and taking the time to think and make a comment, I was just watching casually and already knew what you stated, yet it escaped me. What should be stated is how the manufacture rates the components, which are high quality and I’ve personally used them many times. I agree though that by calling out the Keysight instrument as not measuring correctly is misleading, especially if the target audience doesn’t have an advanced understanding of the complex physics involved. This will lead to poor product design and customers choosing an improper component. What is odd is that this behavior is useful in many situations so it’s not a negative for the product. I do appreciate the effort made educate their customer base, but that effort should be geared more toward universal information that can be applied to all capacitors.
Yeah, my pretty much next question would be : Then what is that MLCC's actual capacitance when measured *at* 100mVRMS+0Voffs at 1kHz.
Also, how large is the uncertainty of that measurement setup.
Knowing about the equipment used, I'd say it's pretty much spot-on.
I'm also wondering if the MLCC's capacitance would be in-spec after a reflow's heat cycle.
I know that ferroelectric MLCCs usually increase their capacitance upon reflowing.
I'm not sure if that is a related phenomenon to what we are seeing here with higher AC amplitude.
Great tip .Thank you. If we are measuring pcb trac capacitace with FR4 material , should we turn on ALC?
Thank you. Finally, I can do the short measurement on a 4268A. It is the ALC problem!
So glad to hear!
@@KEMETElectronics Hi, how can you measure parallelled X7Rs? For individual measurement, I got correct value with ALC on. When I parallel 2 47uf X7Rs, the result is way off. Any tips?
Hi Po-Ting - We have a couple of questions - 1) Are you measuring both capacitors at the same time, or are you measuring them individually when connected in parallel? 2) What is the resulting capacitance that you are measuring? If they are being measured together, the resulting total capacitance should be be around 94uf. - Sam@KEMET
@@KEMETElectronicsHi! I soldered 2 47uf X7Rs in parallel, and measured it as one larger cap. For unknown reason, my 4268A won't show any meaningful result (0.0001nF on display /w ALC on). This was done on a 16044A type test fixture. Today, I also try a 16334A type test fixture (tweezer type) and finally it shows 103uf which is within the spec. I couldn't afford a genuine 16044A or 16334A. They are simply out of my reach. I don't understand why my 16044A type test fixture won't work on paralleled caps. Do you think it was the test fixture? Could you also do a similar test on your device? Thanks.
Can you tell us whether or not the 16044A is measuring correctly when 1) only one of the capacitors are being measured and 2) what the results are when measuring the two parts in parallel when ALC is off? If the test fixture is not giving the correct results for a single capacitor or when ALC is off, then it is probably an issue with the fixture, especially since the 16334A works just fine. On the datasheet for the 16044A, I noticed that there is a size limitation for a capacitor, so I suspect that having two capacitors might not fit in the DUT probe pins correctly leading to the off readings. Also, looking at the capacitance meter that you are using, 4268A, it looks like the default conditions for testing the capacitor are 70 uF at 1 V rms at 1 kHz. 120 Hz measurement ensures the constant 1 rms test signal for up to 600 uF. Please be sure to check the datasheet/specsheet to see how we measure the parts. Some of our X7Rs at 47uF are measured at 120 Hz 0.5Vrms. -Sam@KEMET
Wish I had a LCR Meter like that.
What budget capacitor meter do you recommend to measure smd capacitor in circuit
It's very hard to measure capacitors in situ. A budget meter would be good at anything but class II ceramic caps because of the piezoelectric nature of class II ceramics. For tantalum, film, class I ceramics, and aluminum electrolytic any RLC meter should be good.
what if I want to measure a 0201 SMD component? what fixture should I use?
Hi Francisco - Sam@KEMET has provided a few options for you here, but any similar tester would do the job. You can choose from below, or any comparable product of your choice.
- www.hioki.com/en/products/detail/?product_key=5906
- www.testequipmentdepot.com/instek/accessories/test-fixtures/test-fixture-smd-chip-components-lcr-15.htm
- www.amazon.com/HPS28004-Digital-Precision-Capacitance-Inductance/dp/B08P8GCSC3
cake?