![]() Better matched parts are selected from the same wafer, and the best possible matching is possible when transistors are on the same silicon die, i.e. If one wants matched parts, one should purchase hFE-grouped transistors. Some variations also exist within the same batch. In most textbooks and tutorials, the common explanation is something similar to the following: The transistors are not coming from the same batch, so there will be variations in the manufacturing process. In this case, RE should be removed.It's common sense that the current gain β (hFE) in a bipolar junction transistor is somewhat unpredictable, and one must design a discrete transistor circuit to be insensitive to the variation of β. When you want to use only one single resistor for DC biasing, implement voltage-controlled current-feedback and connect RB not to the power rail but to the collector (of course with a recalculated value). This effect cannot be (partly) compensated by voltage feedback (provided by RE). And this current is highly dependent on the beta-value which has very large tolerances. Therefore, we speak about "current injection" into the base. Here, we have a large resistor RB (mostly in the high kOhm or even MegOhm range). However, in the shown circuit that is not the case. Remember that the most severe temperature influence is caused by the temperature sensitive current Is that appears in Shockleys well-known expoenetial expression Ic=Is. Such a feedback scheme works best when the DC base voltage is kept as constant as possible (using a low-resistive voltage divider at the base). This resistor provides negative current-controlled VOLTAGE feedback. It is the task of the resistor RE to stabilize the DC operating point - mostly against temperature variations. The circuit incorporates something like a contradiction. Panic attack - the problem with the shown circuit is the following (and that is the reason, this scheme is not recommeded): It's bad design to make the circuit unnecessarily expensive and complex when something simpler and cheaper will meet all the specifications. ![]() You can reduce the bias point variation with temperature to almost any desired amount with the addition of extra circuitry, up to and including using op-amps to force a bias point. Whether it's good design or not really depends on the design parameters. Maybe you don't care much if that radio starts to distort a bit if someone is operating it at -40°, only that it works well from 10☌ to 35☌, and costs the minimum amount. ![]() You're also going to have to account for beta variation with temperature (and aging, especially at very high temperature, and radiation exposure), which depends on the required operating range, quite significant for military temperature range, relatively modest for consumer products. If you're amplifying an AC signal up to (say) 1V you don't much care about the DC voltage at the emitter. So your maximum swing is reduced a bit for those units at the extremes, but if your required signal levels are not pushing the limits of the supply you're fine. ![]() You build a batch of 10,000 pieces and you happen to get some transistors at the extremes. So you design the circuit for a beta of 100. Suppose beta can vary from part to part from 75 to 150 (2:1), which is practical for normal inexpensive binned Asian transistors. Now the reason is given as because it is dependent on beta.īut here what is meant by beta dependence? Does that mean two same models have different beta or does that mean beta is changing by time for the same unit? Now above assumes beta is known equal to 100 and is fixed and never changes. Using the KVL equation with this becomes: I neglect Vbe since it is small relative to the 7.5V I also take Ie=Ic=100*Ib. I will use beta and write KVL as follows: So the aim is to obtain around 7.5V(half-rail) at output when there is no small signal input. ![]() But I will still solve this by assuming we have a constant unchanging beta: Now I read that this is not a good design. What I understand that, in a fixed bias BJT scheme the problem is beta dependence. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |