Ideally the MOSFET is a current source in saturation mode, but I don't understand why is it modeled with an (ideal) infinite inner resistance. How would the current flow?
6
-
1\$\begingroup\$ Because it has thee connections, not two. \$\endgroup\$– OldfartCommented May 29, 2018 at 11:08
-
\$\begingroup\$ Current sources have infinite impedance - go back to basics and forget MOSFETs. Voltage sources have zero impedance. \$\endgroup\$– Andy akaCommented May 29, 2018 at 11:09
-
\$\begingroup\$ Ah, so in order for it to flow it searches the path of least resistance, so everything outside of it basically. Would that be correct? \$\endgroup\$– tadm123Commented May 29, 2018 at 11:22
-
\$\begingroup\$ it searches the path of least resistance Nope, when a 1 ohm and a 1000 ohm resistor are placed in parallel with 1 Volt across them, will all the current flow through the 1 ohm resistor? If you think that is the case then explain why Ohm's law suddenly does not apply for the 1000 Ohms resistor anymore. The Ro simply becomes part of the whole network so you have to find the current same way as for any network. \$\endgroup\$– BimpelrekkieCommented May 29, 2018 at 11:28
-
\$\begingroup\$ Thanks for responding. Well I'm not saying everything would flow through it but the vast majority, right? \$\endgroup\$– tadm123Commented May 29, 2018 at 11:33
|
Show 1 more comment
2 Answers
\$\begingroup\$
\$\endgroup\$
An ideal current source will have an infinite output impedance. You could connect a resistor in parallel with the current source but that resistor would need to have an infinite value R = infinite ohms because if the resistor value was any lower than infinite, current would flow through it and the current source (including resistor) would not be ideal anymore. The current would then depend on the voltage across the current source.
To model a non ideal current source the simplest way is to give that parallel resistor a finite value. That is what is done for a MOSFET in saturation mode as it is a non-ideal current source.
How the current flows can simply be determined by network equations. If you know the Vds of a MOSFET you can determine how much current flows through Ro.
answered May 29, 2018 at 11:26
\$\begingroup\$
\$\endgroup\$
tadm123 - you must see the difference between static DC resistances and dynamic/differential resistances.
As an example - an IDEAL current source delivers a constant current - independent on the connected load. Therefore, there is a DC current flowing through the source - however, because it is a CONSTANT source,the internal dynamic resistance (against any changes of the current) is assumed to be infinite in all of our calculations.
Transistors are treated as (non-ideal) current sources (as far as SIGNAL currents - i.e. current changes - are concerend) - and, of course there is a DC current through the device.
However, the output characteristics of the transistor Id=f(Vds) exhibit a rather small slope (nearly horizontal) which is equivalent to a relatively large DYNAMIC resistance. This means, that we have only rather small current changes (signal current) for relatively large signal voltages across the device. This effect allows us to speak about a rather large internal (dynamic) resistance rds=d(Vds)/d(Id).
(Note that the static DC resistance Rds=Vds/Id is not very large, but this quantity does not appear in our gain calculations, because signal parameters matters only.)
