17-04-2012, 06:29 PM
Hi Al,
I'm sure you know I was being slightly flippant :-)
In the audio world, with just a couple of rare exceptions, bipolars are always capable of producing better results in terms of linearity and noise and cost. That's compared to JFETs and MOSFETs. There are people who argue otherwise, but in purely objective terms, it's not difficult to settle the argument - though obviously subjectivism is often an insurmountable factor in audio...
MOSFETs have their place - an obvious one being switched mode power supplies. Their freedom from secondary breakdown is nice. Their ESD tolerance is less ideal though
Specialist applications involving high impedance is clearly where JFETs and MOSFETs can excel. although it's possible to get pretty high with bipolar circuits when you start playing games with bootstrapping.
Joe, the best analogy I've heard for how JFETs work is to consider the drain-source path as being like a hosepipe that's been crushed, and as you "pull" at the gate, it opens up the pipe. Of course, you should have no problem with these because they are generally implemented just like valves, holding the gate at 0V and using Rs to develop a voltage a the source that results in the required -ve Vgs. JFETs work in so-called "depletion" mode, whereby you need to do something to them to make them stop conducting...
MOSFETs generally work in "enhancement" mode, where, like a bipolar, you need to supply an input to make them do something. Like a bipolar, that input is a voltage, but unlike a bipolar, that voltage is not very well defined.
In a bipolar, base current flows (and this is a minor inconvenience), but it's the base-emitter voltage that matters. In other words, the all three of these transistor types are voltage-operated. Many people mistakenly believe the bipolar transistor to be current-operated, but that model of operation is rather imprecise and only used for approximations.
All the best,
Mark
I'm sure you know I was being slightly flippant :-)
In the audio world, with just a couple of rare exceptions, bipolars are always capable of producing better results in terms of linearity and noise and cost. That's compared to JFETs and MOSFETs. There are people who argue otherwise, but in purely objective terms, it's not difficult to settle the argument - though obviously subjectivism is often an insurmountable factor in audio...
MOSFETs have their place - an obvious one being switched mode power supplies. Their freedom from secondary breakdown is nice. Their ESD tolerance is less ideal though
Specialist applications involving high impedance is clearly where JFETs and MOSFETs can excel. although it's possible to get pretty high with bipolar circuits when you start playing games with bootstrapping.
Joe, the best analogy I've heard for how JFETs work is to consider the drain-source path as being like a hosepipe that's been crushed, and as you "pull" at the gate, it opens up the pipe. Of course, you should have no problem with these because they are generally implemented just like valves, holding the gate at 0V and using Rs to develop a voltage a the source that results in the required -ve Vgs. JFETs work in so-called "depletion" mode, whereby you need to do something to them to make them stop conducting...
MOSFETs generally work in "enhancement" mode, where, like a bipolar, you need to supply an input to make them do something. Like a bipolar, that input is a voltage, but unlike a bipolar, that voltage is not very well defined.
In a bipolar, base current flows (and this is a minor inconvenience), but it's the base-emitter voltage that matters. In other words, the all three of these transistor types are voltage-operated. Many people mistakenly believe the bipolar transistor to be current-operated, but that model of operation is rather imprecise and only used for approximations.
All the best,
Mark
