Active Discussion About Passive Components

People talk about active and passive components like they are two distinct classes of electronic parts. When sourcing components on a BOM, you have the passives, which are the little things that are cheaper than a dime a dozen, and then the rest that make up the bulk of the cost. Diodes and transistors definitely fall into the cheap little things category, but aren’t necessarily passive components, so what IS the difference?

Resistors, Capacitors, Inductors, Transformers, Diodes*, and Memristors

That’s the list. Those are your passive components. Well, it’s not that easy. Also add in a bunch of types of sensors, because they are still passive. A photoresistor is a sensor but it’s still a resistor, even though its resistance changes based on an external influence. Any sensor whose measurement is a change in resistance, capacitance, or inductance still qualifies as a passive device. Also for fun let’s add a piezo buzzer.

The memristor is weird because it has only recently been proven to exist despite being theorized in the 70s, and is still not quite commercially available. There are now theories about meminductors and memcapacitors, which would also be passive devices, but they don’t exist yet.

It Depends on What Your Definition of Active Is

Part of the problem is it seems people have varying definitions of active. Rather than debunk all the wrong ones and spread bad ideas, here’s what’s correct. A device is active if any of these conditions are met:

  • It is a source of power
  • It amplifies power
  • It acts as a switch

Applying this to the obviously active devices, like microcontrollers, it makes sense. It does all of those things on a GPIO pin. A transistor can amplify or act as a switch. A battery is a source of power.

A circuit remains passive until a single active component is added, so an RC or LC network is still passive. A piezo buzzer has an equivalent circuit of entirely passive elements, so it is also a passive device.

The equivalent circuit of a piezo buzzer is all passive elements.

As a side note, every circuit has at least one active device (a source of power). Also, an electromechanical device like a physical switch is considered passive.

The Diode

There is an exception with the diode. The vast majority of the time, it is a passive device, so it’s handy to just add it to the list of passive devices and mostly forget about it. It wouldn’t be interesting, though, unless we delve into what makes it sometimes active for that single, and rarely used exception, and to do that we have to get into quantum tunneling.

The tunnel diode is very fast (microwave frequencies), and is used in frequency converters and detectors, especially in space where its resistance to ionizing radiation, low voltage, high frequency, and longevity are desirable qualities. There is a specific condition of the tunnel diode in which it has negative resistance so that increasing voltage results in decreased current. Even the tunnel diode acts like a normal passive diode everywhere except this special region.

IV curve of a tunnel diode. The descending section is the area of negative resistance where increased voltage results in decreased current. By Mcguireatneuroticadotcom CC BY-SA 3.0

A charged particle moving across a barrier needs enough energy to get over the barrier or else it can’t cross. With a normal diode there is a PN junction that acts as the barrier. A power supply gives enough energy (called the forward voltage) for the electrons to get over that barrier, and the current flows through it. According to quantum mechanics, though, there is a non-zero probability that the electron will just jump to the other side of the barrier without going over it. This is quantum tunneling. In most diodes the barrier is high enough (controlled by the doping of the PN junction), that the tunneling is unlikely, so no current will flow until there is enough forward voltage to get the electrons over the barrier. In a tunnel diode, the PN junction has a lot more doping, increasing the likelihood of tunneling. These diodes work at much lower voltages than normal diodes because of the high doping.

At really low voltages, the electrons tunnel frequently and there is some current. As the voltage increases, tunneling increases to a peak and starts going down. It goes down because the electrons on one side of the barrier have more and more energy, but there are not the same holes on the other side of the barrier to accept them from tunneling. Once the forward voltage is high enough, the electrons have enough energy to get over the barrier without tunneling, and the tunnel diode acts like a normal diode again. This behavior allows the tunnel diode to act as an amplifier or as an oscillator, which puts it into the active category. We covered negative resistance in the tunnel diode a few months ago, and a post on diodes kicked off the active/passive debate in the comments.

Does it Matter?

Nah, not really. This is well into the realm of the esoteric, and has no practical use other than to annoy people at parties and probably below in the comments. Active and passive are generic terms for components and whether a particular component is classified as one or another doesn’t change how it is used. Quantum tunneling is neat, though, and the fact that we have harnessed it makes me wonder how close we are to warp speed and teleporters.

35 thoughts on “Active Discussion About Passive Components

  1. Actually, the difference is that active devices can exhibit power gain. Voltage or current gain are possible with passive devices (like a transformer), but not power gain.

    And, acting like a switch? That would imply a non-linear device, but not active.

    1. I tweaked the definition to replace “voltage or current” with “power,” as you are correct.

      As for acting like a switch, can you give an example of something that acts like a switch but is not active or electromechanical? The point of that bullet is to say that an active device can control current, turning it on or off.

      1. >an active device can control current

        Any resistor is therefore an active device, because heating or cooling it changes the current passing through it. If you put a heater wire around a thermistor, you can create a “transistor”…

  2. Your wording here is not quite right:
    “In most diodes the barrier is high enough (controlled by the doping of the PN junction), that the tunneling is unlikely, so no current will flow until there is enough forward voltage to get the electrons over the barrier.”
    It should read “In most diodes, the barrier is *thick* enough that tunneling current is negligible, so effectively no current will flow until …”
    The key with a tunnel device is that the barrier region be very narrow. The narrower that region is, the more likely an electron can tunnel through. The actual height of the barrier is mostly irrelevant for the tunneling behavior, needing only to be so high that it prevents current driven by other phenomena. If this were not the case, then scanning tunneling microscopes (which use a lab-grade vacuum as their barrier) wouldn’t work.

        1. Oh yeah… All the interconnects are done in a single color of enameled wire… You’ll need all those prayers to God if you ever need to service it! O_o

          The two boards on the middle level, in the center and on the right are stuffed with daughterboards that each contain diodes and diodes and diodes… I think I counted 87 on one single daughterboard. I think I counted 16 or 18 daughterboards… Hard to tell from the angle I was looking On the other side of that black panel running along the top, are the hybrid modules. The top hinges open and you can access them from the front of the unit.

          You can see the two diode motherboards below the hybrid module boards (this is now viewed from the front). A bundle of enameled wire feeds in along the side, connecting to the diodes, and the motherboards are clearly arranged to place all the diode daughterboards in parallel. Heck, I’m assuming some of those are wired to the drawbars. The drawbars are arranged in an 8 x 18 grid of combinations, with any mix and match possible. with 1 switch, you have 8 positions. 2 switches gives you 64 combinations… just 3 switches gives you 512 combinations! Keep going… There are 18 switches! That calculates to 18014398509481984 possible combinations. Technically, it’s 134217728 possible combinations for EACH of the two manuals (the white and black note keyboards)… but the grand total is still impressive. That is an INSANE number of variable states! That doesn’t even count the instrument switches! As mentioned, this thing has two manuals, one above the other, and a set of foot pedals that can play independent instrument configurations. The right pedal is an analog pedal to adjust the volume level.

          And they wired the whole blessed thing with that SINGLE color of enameled wire… -__-
          Tied into bundles… -____-

          http://richfiles.solarbotics.net/eb/GulbransenParagonHybrids.jpg

        1. as time moves forward, the probability increases linearly that you’ll find the value of smd cap you need from fallen parts on the floor.

          such parts are called ‘floor capacitors’.

          ;)

          1. This also applies to small mechanical parts. In one assembly the techs had to install a tiny e-ring into a counter bored hole over a tiny screw with expanding pliers. After a while one was more likely to find an e-ring on the floor than in stock because procurement only ordered the exact quantity for the production run. Extra points given for the spatter-finish tiles that camouflaged them. They were stainless, so magnetic brooms could not work and regular brooms, which would work, would gather too much dust compared to the component size.

      1. The equipment is broken, but what about the part you want inside?
        The part is both broken AND functional, and you don’t know which until you test it.

        Actually, that sounds like every component in my parts drawer.
        Some of the CMOS logic chips date back to the 80’s. Might be good, might be dead.

    1. You could also subdivide in components that need to be a more exact value and ones that do not. Like a resistor for instance needs to be a value but with switches you have a wide range of options.

    1. Many shops I’ve looked for items at have their online catalog divided in active and passive component sections.

      As for the red or blue pill, my advise is to not accept pills of any kind from strangers.

  3. In network theory, the distinction between active and passive is defined in terms of ports: a pair of terminals where incremental current can flow, and where it makes sense to measure voltage.

    Being able to measure a port’s response to voltage and current means you need another port to drive it, and we have two general models for a driving port: the Thevenin equivalent circuit (an impedance in series with a voltage source) and the Norton equivalent circuit (an impedance in parallel with a current source). For the sake of discussion, let’s say our driving port is a variable voltage source in series with a 1k resistor.

    A passive component is controlled entirely by the driving port. If we set the driving port’s voltage to 5V and connect a 10k resistor, a 1uF capacitor, or a 1uH inductor between the free end of the 1k and GND, we can calculate the voltage and current response. If we turn the driving port voltage into a sine wave, we can calculate the response of the driven resistor, cap, or inductor.

    That ‘controlled entirely by the driving port’ behavior is what makes a component passive in network terms.

    An active component like a transistor has at least two ports (gate-source and drain-source for a mosfet) and a gain, transadmittance, or transimpedance relationship where the voltage/current at one port changes the voltage/current behavior at the other port.

    If we connect the 5V-1k Thevenin equivalent circuit across a mosfet’s drain-source port, we don’t have enough infrormation to predict how much current will flow. We also have to know the voltage at the gate-source port and the mosfet’s transconductance. If we connect a sine wave voltage across the gate-source port, the current/voltage of the drain-source port will change independently of the 5V-1k driving port.

    That ‘changes independently of the driving port’ behavior is what makes a component active in network terms.

    A diode has two terminals and one port, and its current/voltage relationship is Id=Is(-1+e^Vd/nVt). If we connect the 5V-1k driving port to a diode, we can calculate the diode voltage and current. If we turn the driving port voltage into a sine wave, we can calculate the response across/through the diode.

    That makes a diode passive in network terms.

    The diode’s current-to-voltage function is nonlinear, but so are the I-V functions of a capacitor and inductor.

    A diode is also non-reciprocal though: the I-V function changes if you reverse the polarity of the driving voltage. A reverse-biased diode acts like a capacitor, but is still passive in network terms.

    Non-reciprocal components break the math we use to calculate the response of passive networks, and so do active components. Diodes and transistors both break the math, but they do it in different ways.

  4. A diode often acts as a switch, is it really to be called passive?

    I don’t know what a “meminductor” or “memcapacitor” are supposed to be, but inductors with a ferromagnetic core and capacitors with a ferroelectric dielectricum have memory features and are used in memory devices. The old fashioned core memory, experimental MRAMs and already commercially available FRAMs.

  5. It’s amusing that while MOST coils are passive devices, saturable core reactors are merely a collection of coils, will do power amplification and can be classified as active devices. Also I’ve always considered diodes active devices for various reasons but that point is moot.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.