Science desk
< August 18	<< Jul \| August \| Sep >>	August 20 >

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

August 19

how do commercial capacitance meters (DMMs with capacitance function) work

I can think offhand of several methods of measuring C:

measure voltage drop across AC of known frequency
measure frequency of RC or Hartley osc composed of a reference resistor/inductor and the cap under test
measure phase shift
charge reference cap, discharge into cap under test, the ratio of the voltages will be proportional to their capacitances
charge cap from constant current source, the time to charge to a certain threshold voltage will be proportional to the capacitance
???

But how do commercial multimeters do it? 80.171.81.10 (talk) 02:09, 19 August 2017 (UTC)[reply]

Another method: charge cap, then discharge into resistor and measure the RC time constant, the time it takes for the voltage to fall to 36.8%. Note that the article implies (by omission) that the fall is linear with respect to time, but that is not the case. The article really needs a graph showing the non-linear fall. Akld guy (talk) 03:35, 19 August 2017 (UTC)[reply]

This capacitance meter circuit comprises two 555 timer ICs. It directly reads capacitance in the range 100pF to 10µF. IC1 and IC2 operate as an astable (with frequency above 80Hz) and as a monostable multivibrator respectively. Time period of the monostable multivibrator is determined by the resistors (R3 to R7 and VR1 to VR5) selected by switch S1 and unknown capacitance Cx. Replace the 1mA moving-coil ammeter with a digital meter if preferred. Blooteuth (talk) 05:45, 19 August 2017 (UTC)[reply]

All of the above. When capacitance meters were a separate thing, they were usually an AC bridge circuit, and even then they needed a variable drive frequency to cope with the wide range of capacitance values encountered. As a crude modern circuit (the cheap, inaccurate, but highly useful $10 "any component testers" found on eBay), they're mostly based on some sort of charge pump where a monostable drives a series of constant current pulses and the voltage rise is measured. This circuit is used because it's cheap: it can be built from low frequency DC circuits attached to the pins of a cheap microcontroller, rather than some AC circuit which would have to be designed and built onto the circuit board with dedicated components. The downside is that it's limited in the range it can measure, and it measures "charge storing ability" (including that charge decaying from leakage over time) rather than an idealised capacitance. Leaky old capacitors in ancient radio circuits highlight just what a difference there can be with those two measurements, over a short or long time period. Andy Dingley (talk) 09:08, 19 August 2017 (UTC)[reply]

The 555 works fine in the µF range. Testing caps in pF range, another circuit is necessary, sometimes another oscillator, followed by a circuit converting the frequency to a voltage. --Hans Haase (有问题吗) 15:22, 19 August 2017 (UTC)[reply]

Thank you, everyone (OP.) 78.53.241.64 (talk) 08:28, 20 August 2017 (UTC)[reply]

Bridge circuits were mentioned but usually require manual balancing so they do not yield a direct reading. Wheatstone bridge is the basic configuration used to measure resistance. For measuring capacitance the bridge types are Schering Bridge and Wien bridge; the former is independent of frequency while the latter measures capacitance in terms of resistance and known frequency (or can be used to measure frequency when the capacitance is known). Blooteuth (talk) 13:00, 20 August 2017 (UTC)[reply]

Once the microprocessor era was established, there were any number of automatic self-balancing bridges around as test gear. Even before this though, the classic HP 4270 was available from 1970, using digital control circuits and a dozen boards of TTL. It used four oscillator frequencies from 1kHz to 1MHz and measured from sub-pF to μF. Nice article here: Hitoshi Noguchi; Takeo Shimizu; Koichi Maeda (May 1970). "Measuring Capacitance Automatically" (PDF). Hewlett-Packard Journal: 14–20. Andy Dingley (talk) 20:20, 20 August 2017 (UTC)[reply]

Huge dandelion

What is the maximum size the common dandelion can grow to assuming ideal conditions and no competition? — Preceding unsigned comment added by 61.7.175.177 (talk) 14:55, 19 August 2017 (UTC)[reply]

Well, to start, what do you mean by "size"? Height, dry biomass mass, wet volume? Here [1] is a very tall dandelion, here [2] is Guinness'current record holder. Here [3] is a scholarly research article on dandelion demography that can give some context. Here [4] is a study that specifically examines the effects of competition on dandelions. Here [5] is a study specifically about plasticity in a few dandelion species, which is the idea that the same species can grow rather large or rather small, depending on conditions.

Here is an article [6] all about increasing dandelion biomass in agriculture.

I suggest that nobody really knows for sure. You can try some experiments yourself, or you could hire someone to work out some estimates based on models of plant growth, possibly building upon some of the research linked above. But lacking that, the records above are probably pretty good approaches to maximum height. SemanticMantis (talk) 16:39, 19 August 2017 (UTC)[reply]

The first image is not taraxacum officinale. It looks to my fairly experienced eye like either lactuca virosa or lactuca serriola, both of which belong to the lettuce family. Let's get the sheep sorted from the goats before we start comparing. Richard Avery (talk)

Yes, thanks, I had meant to mention that it didn't seem that way but in my haste to find scholarly articles I forgot to. I am a little confused. I didn't think "dandelion" was used for anything other than Taraxacum. Maybe the photo is mixed up? The Guinness bit is perhaps more reliable. SemanticMantis (talk) 16:40, 20 August 2017 (UTC)[reply]