7
\$\begingroup\$

I am a software developer (using high level languages like .NET,C,C++ etc) trying to understand how computers work at a lower level. I never studied any Electrical Engineering at university. I did GCSE Electronics.

I have experience using discrete capacitors and transistors to build circuits on breadboards and PCB's.

I am trying to understand how computers use capacitors to store information. I understand the difference between SDRAM (flip flops) and DRAM (transistor/capacitor pair).

What do you call the capacitors in a computer that store one bit of information? (there are billions of these in a computer). Every time I Google Capacitors or Capacitors in computers; Google returns information about discrete capacitors.

\$\endgroup\$
5
  • \$\begingroup\$ I wonder if they are moscaps. They're fabricated on silicon and it is easy to make very large arrays of them. Try searching with moscap dram as a keyword. \$\endgroup\$
    – HL-SDK
    Commented Jan 29, 2014 at 20:32
  • 1
    \$\begingroup\$ Look up metal insulator capacitor (MIS) or CCD (Charge coupled devices) \$\endgroup\$ Commented Jan 29, 2014 at 20:34
  • \$\begingroup\$ You mean "SRAM (flip-flops)". SRAM stands for static RAM, while SDRAM stands for syncrhonous dynamic RAM. \$\endgroup\$
    – Dave Tweed
    Commented Jan 29, 2014 at 21:16
  • \$\begingroup\$ SRAM cell (4 transistors) and flip-flops (commonly the D flip-flop) are different. \$\endgroup\$
    – RawBean
    Commented Jan 29, 2014 at 21:24
  • \$\begingroup\$ @RawBean - A SRAM cell implements one type of flip-flop. \$\endgroup\$ Commented Jan 30, 2014 at 9:37

3 Answers 3

2
\$\begingroup\$

The capacitors don't have a special individual name; the basic unit of DRAM is known as a DRAM cell and the capacitative storage in Flash is also a cell or a floating gate.

Large numbers of devices are fitted on a integrated circuit using Photolithography.

\$\endgroup\$
2
\$\begingroup\$

The big drawback of capacitors is that they leaking, due to their internal parasitic resistance (its conductance is never null). In common IC technologies (based on silicon), they are not a good choice for storing a bit of information. That's why we use them in conjunction with a transistor in the so-called DRAM memory.

As you know as a software engineer, DRAM is refreshed multiple times per seconds to keep its state. Non-volatile memory like Flash (NAND, EEPROM ...) doesn't use capacitors. For example, EEPROM use transistors only, and the capacitive effect of a part of the transistor, the floating gate. Which is, as far as I know, different than using a pure capacitor.

You could read this page: http://en.wikipedia.org/wiki/Regenerative_capacitor_memory

\$\endgroup\$
4
  • 1
    \$\begingroup\$ 1: It's not that the resistance accross a capacitor is never null, but that it is never infinite. 2: Flash does use capacitance to store bits. The capacitor is the floating gate of a MOSFET against its body. This gate is insulated with basically glass all around, so the leakage resistance is very small. Tunnelling is used to add to or drain charge from it. \$\endgroup\$ Commented Jan 29, 2014 at 21:33
  • \$\begingroup\$ I am trying to understand how a computer can have billions of these capacitors on a silicon sheet. Thanks. \$\endgroup\$
    – w0051977
    Commented Jan 29, 2014 at 21:34
  • \$\begingroup\$ Thanks + 1. Is there a name for these capacitors? There are lots of different names quoted in the comments. Not sure which is right. \$\endgroup\$
    – w0051977
    Commented Jan 29, 2014 at 21:42
  • \$\begingroup\$ @OlinLathrop: thanks for your valuable comment. I've corrected my mistake about the resistance which is never infinite. Finally I decided to use the term conductance that is not null instead of resistance that is not infinite. About your point on flash techno, and the floating gate of transistor used to store bit, I've decided to say that it's not a pure capacitor, but the capacitive effect (kind of parasitic effect somehow) of the transistor which is used. There are plenty (infinite) parasitic effect everywhere in ICs so... let's make a clear difference between to two. \$\endgroup\$
    – RawBean
    Commented Jan 30, 2014 at 7:43
0
\$\begingroup\$

There is a style of digital logic circuitry that uses capacitance to store information: dynamic logic. If you have a processor that specifies a minimum operating frequency then it almost certainly uses dynamic logic. The advantage is that you replace a latch with an inverter and a pass transistor, where the pass transistor is used to move charge on or off of the input capacitance of the inverter.

The disadvantage of dynamic logic is that it is much more susceptible to noise, and if you aren't careful about charging the storage nodes at just the right time the logic levels will be degraded by charge sharing. And, of course, you have to keep the clock running all of the time.

\$\endgroup\$
3
  • \$\begingroup\$ I am trying to understand how a computer can have billions of these capacitors on a silicon sheet. How do they all fit? Thanks. \$\endgroup\$
    – w0051977
    Commented Jan 29, 2014 at 21:35
  • \$\begingroup\$ Well, they are very, very tiny...no bigger than a single transistor. Transistors now have dimensions on the order of tens of nanometers so a billion of them will fit on a small silicon chip. The amount of capacitance may be measured in attofarads. \$\endgroup\$
    – Joe Hass
    Commented Jan 29, 2014 at 21:39
  • \$\begingroup\$ If it helps any, these are not tiny discrete capacitors. They are made, like everything else in an integrated circuit, by lithography. It is analogous to a photographic negative being used to expose an image on photographic paper. Layer by layer, it starts out as a "picture" of a the plates of a capacitor. \$\endgroup\$ Commented Jan 29, 2014 at 22:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.