How are processors designed today?

[lidnariq]

Ah, so within a small self-contained group it's okay to have a few transistors feeding into each other, but you generally need to isolate each logical unit by drawing from the power source again?

Yeah, it's better engineering to isolate everything, but abusing that can do better (smaller, faster, lower power, &c)

MOSFETs care about the gate-to-body voltage, while BJTs care about the base-to-emitter voltage. This difference actually helps quite a lot. Relatedly, some logic (especially XOR3, as used in adders) can be implemented smallest using "transmission gates", which don't regenerate the voltage through them.

When I was taking the class on VLSI, I think I remember being told "up to 5 stages without a regenerative stage"

Are resistors actually used in gates these days? I thought they took up too much room on the die.

No, absolutely not. But I figured it was easier to explain with a resistor than two (or four...) different kinds of MOSFETs.

I was mostly wondering how many different parts are involved, and something like paging (whatever that is) probably isn't a standalone part, but it's made of a bunch of transistors and things, just put together in a clever way to make that function.

Yeah, kinda. Have you tried reading the wiki's Visual Circuit Tutorial ? Once again, 1970s level technology, but it'll give an idea for the start of how to build up these abstractions.

[Bregalad]

Are resistors actually used in gates these days? I thought they took up too much room on the die.

Not only that, but in analog chips (such as AD converters or voltage regulators) they also uses transistors to fake resistors and capacitors where they would normally be used, it increases the complexity of the circuit a lot but allow to reproduce it in a smaller area.

I remember a while back Kevtris made a small computer entirely out of NAND chips: http://blog.kevtris.org/blogfiles/NANDputer/

There is no reason to limit you to do that, because using many NAND gates where you could use another type of gate just slows down the circuit for no particular gain.

[Drew Sebastino]

I think I found something awesome. http://www.partsim.com/

[Drew Sebastino]

Wait, I just realized something. I don't even know the difference between voltage and current.

Edit: I really just realized that I don't know anything about electricity.

[Sik]

Quick oversimplification:

Voltage (volt) = how fast the electrons go
Current (ampere) = how many electrons go
Resistance (ohm) = how hard it's to go through
Power (watt) = Voltage × Current (・＜・)

[Drew Sebastino]

Yeah, I was reading this just then, and it's pretty much what you just said. https://learn.sparkfun.com/tutorials/vo ... d-ohms-law

What's up with + signs and - signs and grounding and stuff? I also don't know what AC and DC power is, but I know that the wall outlets in houses supply DC power and the annoying power box things in power cords for computers and other things transfer it to AC power. Also, if power gets lost down the line, would you need to put a bunch of places in the circuit board that are connected in the main power supply, or is the power loss neglicable?

[tepples]

An electron has a charge of -1.60217657 × 10^-19 coulombs. The sign is negative for historical reasons because charge was discovered before the particle that carried the charge. (Google "conventional current".) An ampere is one coulomb of electrons per second.

Just as height has no useful absolute "zero", voltage also has no "zero". Things are only higher or lower voltage than other things. So "ground" is a voltage chosen as the zero reference for a system. It was named "ground" because early experiments used a rod driven into the earth as ground, and this is still true of residential wiring that uses the water main as ground.

[Sik]

Note that here the sign is used to indicate the direction in which the electrons are going (negative voltage basically means the electrons are going the opposite way you'd expect).

Also you get AC and it gets converted to DC, not the other way. DC is basically the current going straightforward, while AC has the current constantly alternating between forwards and backwards. The only reason AC really exists is because it has less loss when carried over long distances (e.g. between the power plant and your house), pretty much anything electronic needs to convert it to DC (electric but not electronic stuff like light bulbs can use AC directly just fine).

[Drew Sebastino]

Note that here the sign is used to indicate the direction in which the electrons are going (negative voltage basically means the electrons are going the opposite way you'd expect).

So the side with the negative is the side they are going out, and the side with the positive is where they are going in?

So "ground" is a voltage chosen as the zero reference for a system. It was named "ground" because early experiments used a rod driven into the earth as ground, and this is still true of residential wiring that uses the water main as ground.

Don't you actually have to have it to where part of the circuit is going into something that will pick up the electricity for some reason? My grandfather works on cars and houses and a billion other things, and I actually saw him attach a wire from the battery I think to the frame of the car. What's the purpose of this exactly?

[Sik]

So the side with the negative is the side they are going out, and the side with the positive is where they are going in?

Depends on whether it's conventional current or real current! =D

Rule of thumb: in a circuit diagram, diodes point in the direction where electrons will go. In the case of batteries, the small side is the direction electrons come out of (the large side is where they go into).

Don't you actually have to have it to where part of the circuit is going into something that will pick up the electricity for some reason? My grandfather works on cars and houses and a billion other things, and I actually saw him attach a wire from the battery I think to the frame of the car. What's the purpose of this exactly?

It's the end of the circuit basically.

[Drew Sebastino]

Wait, I thought where the electrons go in (or the back of the battery) was essentially the end of the circuit. At what point in the circuit should there be something that goes off for grounding?

Hate to be a jerk, but there's another category I made under "General Stuff" that I'd prefer got more attention...

[Khaz]

Personally I CANNOT BELIEVE we haven't reconciled the difference between conventional and scientific electric charge. MAKE ELECTRONS POSITIVE. DONE. WHY DOES IT MATTER OH MY GOD.

Seriously. People carrying on conflicting obsolete systems like that causes screwups, and screwups easily lead to deaths in the electricity game. Should have been resolved 100 years ago.

Wait, I thought where the electrons go in (or the back of the battery) was essentially the end of the circuit. At what point in the circuit should there be something that goes off for grounding?

If your circuit goes from a battery to a load and back to the battery, it's an ungrounded circuit. Lots of DC systems are ungrounded. The benefit of such a system is that if you develop a ground fault (accidental path to ground, ie/ wire accidentally touching something) anywhere in the closed loop, it won't (shouldn't) screw up the system but you can still detect that fault and correct it. If you have TWO ground faults in such a system, you're screwed.

In any system that IS grounded, if you get a fault anywhere else in the system, you're screwed.

[Sik]

Personally I CANNOT BELIEVE we haven't reconciled the difference between conventional and scientific electric charge. MAKE ELECTRONS POSITIVE. DONE. WHY DOES IT MATTER OH MY GOD.

Or just stop using conventional current so we just have one direction to deal with, sign quirks be damned (it isn't so hard to understand that current goes from the minus end to the plus end, seriously). Good that symbols use the real direction so at least when reading them you don't have to fight common sense.

[Drew Sebastino]

So what you're saying is that you can either have the back of the battery (or whatever), or a spot to ground the electric current? What happens if you don't do either of these? It's not like it's going to explode, is it?

Also, what's the difference between a P Channel and an N Channel Mosfet? I see that both only have 3 prongs instead of 4, so I'm guessing it's how the 4 prongs are grouped to make 3.

[Khaz]

Or just stop using conventional current so we just have one direction to deal with, sign quirks be damned (it isn't so hard to understand that current goes from the minus end to the plus end, seriously). Good that symbols use the real direction so at least when reading them you don't have to fight common sense.

Personally I'm biased toward the scientific side so I'd agree, I'd say teach electricians electricity is negative and move on. As a more practical person tangentially involved with electrical industry, I'd say that actually making that transition is going to be a lot harder than it sounds. You would have to have a way of identifying explicitly which drawings are the old way and which are correct, because you can't just burn every electrical drawing ever made. Sure, you could come up with a plan, but attempting that is "error-likely", which means it's basically certain SOMEbody's gonna die over it if not hundreds of people. The world of science would probably have an easier time adjusting without killing anybody.

After all, at the end of the day, "positive" and "negative" are completely arbitrary.

So what you're saying is that you can either have the back of the battery (or whatever), or a spot to ground the electric current? What happens if you don't do either of these? It's not like it's going to explode, is it?

Okay so, I'm out of practise so please correct me if I make a mistake. But if you have a circuit with just a battery and a load, that's a closed circuit which is the simplest case, like a flashlight. (That's also the scenario I was talking about with ground fault detection, but that's generally for much bigger systems.) If you short the battery across the terminals, bypassing the load, then that's when you risk things melting or exploding or something else undesirable.

If you connect one terminal of the battery to ground and do nothing else, nothing should happen. The battery is essentially storing a difference of voltage between the two terminals. There should maybe be a quick miniscule discharge while that terminal aligns itself to ground (?), and then nothing because the one connected terminal is now AT ground voltage so no current will flow. If you connect both terminals to the SAME ground that's just shorting it. If you connect each terminal to a totally separate, but equivalent voltage of ground... Uh. I don't think anything happens? That's a weird scenario. I dunno just trying to cover all the bases here and I'm taking some liberties with the definition of "ground".

Also, what's the difference between a P Channel and an N Channel Mosfet? I see that both only have 3 prongs instead of 4, so I'm guessing it's how the 4 prongs are grouped to make 3.

I'm guessing that's referring to whether it is a PNP or NPN transistor? Basically just inverted polarity...?