Difference Between MOS And BJT

I know the difference between them, a BJT acts as a constant current source, with the current controlled by the base voltage, and normal ones (like the ones that are not in EC) suffer from the Early effect which makes it so that at a higher voltage it conducts a little more making it slightly like a resistor, although it's still a good constant current source, and MOSFETs act like variable resistance resistors that are controlled by the gate voltage, although it is a little bit exponential but that doesn't matter too much, and if the drain voltage is higher than (lower than for P-channel of course) the gate voltage it starts to act like a BJT with Early effect. I didn't know that BJT's acted like pentodes, not that I knew much about pentodes anyway, but that is kind of interesting. Don't pentodes have five terminals because pent means five? What are the other two terminals used for? If there are just three terminals than why the pent?

The BJT's can be viewed as a current controlled current source, and MOSFETs - voltage controlled current sources. Vacuum tubes are a whole other topic, I can write pages and pages about. Basically a triode has 3 terminals (and 2 for the heater). It has an anode, a cathode, and a control grid. The anode can be viewed as a collector (drain), the grid is the base (gate) and the cathode is the emitter (source). Triode tubes suffered from the MILLER effect (not Early), and had a limited gain at high frequencies, often requiring 6 or 7 tubes working in tandem (and fine tuned) to operate adequately. That's when the tetrode tube came into play. By separating the anode and the control grid by a so called screen grid, the Miller effect dissappeared, and the HF gain improved, but another nasty effect came to be. The screen grid is often polarised at a very high voltage (close to that of the anode), thus it accelerated the electrons towards the anode too ''aggressively''. At low plate voltages, the anode couldn't sink in all the electrons, so some of them bounced back to the screen grid (which as an attractor absorbed them). This is called a secondary emmision, and resulted in screen grid current flow and a negative resistance kink in the tetrode characteristics at low anode voltages. While this made tubes an excellent choice for a stable negative resistance oscillator, it made them a bad choice for an amplifier, often resulting in parasitic oscillations. That's when the pentode was invented. By adding yet another (suppression) grid between the anode and the screen grid (often biased at ground or cathode potential), it acted as a repellent for the bounced electrons, thus hurtling them back to the anode, and removing the kink problem. There were a few tetrode variations, resembling a pentode, most notably the kinkless tetrode (designed primarily to bypass the pentode royalty fees Mullard requested).

Ok. If the first grid was connected to the grid terminal and the last grid was connected to the cathode terminal, then what was the middle grid connected to? You said that it was "near" the anode voltage, but that implies it was less than the anode voltage thus requiring another terminal. What's that terminal used for?

It's close to the anode voltage because its supply voltage is usually filtered out by an RC chain, and since it draws some mA of current, the resistor causes a voltage drop. Usually it's negligible, unless the designer chooses a different grid line voltage for his project. This grid acts as a screen between the anode and the control grid, nullifying the Miller Effect to a large extent. High Frequency gain and stability suffer without this grid.

What happens when this grid is hooked up straight to the anode? Will it's effect be removed?