Ideal Diodes

What do you mean by an 'ideal diode'? A diode connected transistor does a pretty good job at following the 'ideal diode equation' Ic =3D Is ((exp (eV/kT )+1)

George H.

I'm guessing he is more interested in zero forward voltage

-Lasse

During forward conduction of an ideal diode, there's no power to steal!

"Ideal rectifiers" are used in, say, AC voltmeters, but they aren't

You could make an almost-ideal 2-terminal diode, but it would have a battery or equivalent inside.

John

Well, since it is possible to make a truly ideal diode there will be some power to steal. This would be part of the leakage of the diode. FETS are pretty low power devices and would require very current to be used. The problem with fets is the high gate voltage. So you would need some type of floating amplifier to amplify the few mV of drop to drive capable levels.

Fets used in memory supposedly can work off a single electron. So suppose if we could store an electron on the gate when current is in one direction and remove it when current is in the opposite direction. The issue is getting the electron on the gate and removing it.

I suppose, somewhat based off your idea, we could charge small cap and then use that as a battery source. It doesn't seem impossible but probably not worth it when one can achieve the same effect by using a much simpler 3 terminal device.

You can get close to ideal in terms of Vdrop by using an opamp and relay. But it might fall slightly short on the 1MHz spec :)

NT

LTC4358 for example - not 2 terminals, but worth a look.

Cheers

Not two terminal, but conceptually equivalent:

Problem with the LTC4358 is Vmin of 9V.

If it leaks, it's not ideal.

If you want to do a leaky diode with super low forward drop, you could do it with a mosfet and a micro-power comparator. The power for the comparator could come from a lithium battery (20 year life or so) or you could charge a capacitor when the diode was reverse biased, and use that energy to run the comparator, but that has gotchas.

I once made a 2-terminal negative 1K resistor as a school project, but it had batteries inside. It was fun to play with... just plug negative R values into all the usual equations.

John

Um, do you not realize that nothing is ideal? Show me one real electronic component that is ideal where ideal means it approximates the standard physical model found in textbooks/physics and not some made up useless model that you attach the term ideal to.

Even a switch will have a forward drop and a non-zero leakage resistance.

That's cool. If there were a two terminal version that operated at a lower voltages it would be exactly what I'm looking for.

It doesnt look too hard to roll your own

NT

=A0 =A0 ...Jim Thompson

=A0 =A0| =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

Close to what I'm thinking. I don't see why it can't be done with two terminals.

maybe a pin diode will do what you want? You can take a advantage of that in the higher frequency operations otherwise, it acts like a normal diode..

Or maybe you need a precision rectifier?

Jamie

You asked for an ideal diode, and now you're lecturing me about nothing being ideal?

John

I guess he jumped to the conclusion that you were intelligent enough to implicitly realize that he meant he wanted an approximation to an ideal diode since it is impossible to construct an ideal diode and anyone with at least a decent high school education would innately understand this? Sorry for the run on sentence but I think it was appropriate here.

I think it was the OP's responsibility to specify more accurately what he was looking for.

An arbitrarily leaky ideal diode, I think.

John