I'm experimenting with BJT logic circuits and am interested in learning about improving power consumption and stability (Fully saturation when I want it, consistently). The simplest circuit to help describe my question would be the basic NOT gate, where the collector is connected to the positive rail through a pull-up resistor, the base is connected to a resistor which is then connected to your input, and the emitter is connected to ground. The output is located at the collector.

Ok, from what I understand thus far, the biggest thing determining if the transistor fully conducts (saturates) when it's supposed to is the ratio between the resistors at the base and collector. Or, in other words, the ratio between how much current there is going through the base and how much current the collector resistor can supply. My question is, what effect is there on changing the actual values of these resistors? What's the difference between a 1k and 10k resistor setup (where the collector is 1k and base is 10k) and a setup with 100k and 1Mohm resistors? The ratios are the same, so the switch should behave in much the same way, right? I'm guessing there are 4 things that happen as the resistor values get higher. 1) you have a much smaller output signal to work with. 2) Beta goes down as Collector current goes down. 3) the possibility of interference increases. and 4) the switching speed goes down because of internal capacitance and such.

Were my assumptions correct and/or indicating correct working knowledge of BJTs? Are these the ONLY things I need to worry about when determining how high I can go with my resistor values? Any helpful advice related to this?

 

The next stage. The load. A logic gate doesn't just switch itself, it switches the next gate or some kind of load, or it doesn't have a purpose in life. For every 10k/100k gate you make, it has to drive the next stage, which is a 10k load (if you make them all the same). You can't stop your brain at the collector. There is always the next stage.

 

The next stage. The load. A logic gate doesn't just switch itself, it switches the next gate or some kind of load, or it doesn't have a purpose in life. For every 10k/100k gate you make, it has to drive the next stage, which is a 10k load (if you make them all the same). You can't stop your brain at the collector. There is always the next stage.

Yes, this is very true. Basically you're saying that the higher the resistor values, the smaller your output will be. I mentioned this in my original post.

"1) you have a much smaller output signal to work with."

Usually I have this in mind when designing the gates, I figure out the most current (or the lowest load resistance) I can have whilst still operating correctly, then I simply never exceed this limit when fanning out my outputs or driving LEDs. Typically I use a buffer stage to drive the LEDs. I'm ok with this because I like the idea of having an ultra-low power circuit the vast majority of power is in the LEDs and not the logic itself. I've seen circuit designs where the base and collector resistors are both under 1k, and that just seems wasteful. It's one of the reasons I was so curious about this topic, what's the advantage to using such low resistance? It seems to me like they'd have a reason. Maybe it was just designed that way so that beginners who build the circuit wouldn't have to really worry about it... idk. It just makes me uncomfortable when I'm the only one I can find using such high value resistors in my gates, there's gotta be something wrong/bad that I'm not catching.

 

I think you've got it whipped. The last unanswered question seems to be, "why are there different designs". Some chips are all about using barely enough power to get the logic switching done and some are actually called, "buffers" because they are intended to drive a heavier load. There are so many families of logic chips that I don't even know them all. Each family has a strength that justifies its existence. If you study each family of logic chips, you will know why each family has a purpose that is not filled by the other families of chips.

 

I think you've got it whipped. The last unanswered question seems to be, "why are there different designs". Some chips are all about using barely enough power to get the logic switching done and some are actually called, "buffers" because they are intended to drive a heavier load. There are so many families of logic chips that I don't even know them all. Each family has a strength that justifies its existence. If you study each family of logic chips, you will know why each family has a purpose that is not filled by the other families of chips.

Excellent! I completely understand what you're saying. Transistor-Transistor logic, transistor-resistor logic, diode/transistor/resistor logics of all sorts! I feel like I'm finally starting to grasp the big picture now, things are coming together nicely! Thanks for the feedback.

I'm fascinated with transistors (they're my favorite component, second is capacitors), you'll likely see quite a few more questions about BJTs and eventually FETs