Transistor Biasing Revisited
 

In the tread 'Video Polarity & Direct Injection' by John (vts1134), there arose some contention as to whether a resistor is needed in the emitter-to-ground leg of the transistor. :deal::bash:

The argument was that the resistor is necessary to set the operating point, in the same sense that a cathode-to-ground resistor sets a tube's operating point (for Class A or linear point operation in either case).

But there's a major difference.:nono:A tube's grid is biased negative, while the transistor's base is biased positive.* Thus NO resistor is required in the emitter-to-ground leg to develop base bias. Rather, the bias is derived via a resistor from the positive supply rail.

I know this is the way I built transistor projects back in the day, all using the grounded-emitter configuration, and they all worked fine. Also, John mentioned that his circuit worked with grounded emitter right off the bat. And then the issue arose from other posters on the need for a emitter resistor.

http://www.electronics-tutorials.ws/...r-biasing.html

While refinements of the basic circuit for stability DO include a emitter resistor, that resistor isn't primarily for base bias.
-------------------------------------------------------
*The discussion pertains to NPN transistors. Rarely seen PNP types use negative supply voltage, hence negative base bias.

I don't think anyone (myself included) claimed that the emitter resistor was *necessary* to bias the transistor, only that omitting it was bad practice for stability and repeatability. The equations in the link you posted clearly show a strong dependence on the beta of the transistor, so biasing one individual circuit will not guarantee that duplicates using transistors with an unavoidable range of characteristics will all produce the same collector current. This may not be a killer in stages that work at low levels and do not have a large voltage swing on the collector, but for stages that need to be biased near center no matter what transistor you pull from the bin and put in there, you need the emitter resistor to make it predictable. You would never go without the emitter resistor if you needed to make multiple copies of a circuit that performed the same as each other. Even in circuits with a deliberate variation in emitter current (like AGC'd IF stages), you would make the range of current predictable with a (AC bypassed) emitter resistor, because it affects the range of gain adjustment and unfortunately also the loop stability. If it were feasible to build products without emitter resistors, TV set makers would have done it very early on, as saving five cents per set on 10 million sets a year would be worth it.