So, I understand the basic principal of why this is done, and the general idea about how it is done. And I have made several circuits that use a transistor as a switch/amplifier. However, I just realized that I may have been doing this wrong the entire time. Hopefully someone here can shed some light on this for me.
I have been using a low side switch similar to first picture below (typically with NPN type). However, I have been doing some reading and have read that low side switches are typically used with PNP type transistors, and high side switches are typically used with NPN type transistors. Examples of a NPN High Side and a PNP Low Side are below as well.
As far as I can tell my circuit was working fine. But now I'm reading something different. So what gives?
T1 is a high side switch and can only work if the voltage across the load (and therefore at the collector) is significantly below the voltage of the base (Or rather at least 0.7V). You don't appear to have a base resistor in any not in fact a current limiting resistor, which is why T1 works, but will burn out the LED I would have thought.
T3 is correct, except you have no base or current limiting resistor.
The terms "low side" and "high side" are used to describe on which "side" of the load, the switching device is.
If the switching device (transistor or FET) is connected to ground, it is "low side" and correspondingly if it connects to the positive supply, it is "high side".
You do not need a base resistor for an emitter follower (the first circuit) and indeed, you must not use one, but you do need one for "common emitter" circuits such as the other two (that is, the emitter is connected directly to a supply terminal). If using a FET, a series resistor is used to limit the capacitive charging current of the gate.
Obviously you always need a current-controlling resistor for a LED. Your third circuit shows a "kickback" diode as you would (only) need for a relay or solenoid.
Firstly your circuits all lack a current limiting resistor for the LED, so they will all
destroy the LED (and possibly transistor) from over current.
The flyback/freewheel diode is needed for an inductive load, motor, relay, solenoid,
servo, but not resistive loads or LEDs.
Your first circuit "myway" is an emitter follower - no good for switching as the
transistor cannot saturate and will get very hot driving any sort of high power load
as the collector-emitter voltage will be a volt or so minimum.
When a transistor saturates the collector emitter voltage can be as low as 0.1 or 0.05V,
leading to much less heat generation in the transistor.
To saturate you must drive the base with quite a lot of current, perhaps 5% of the load
current.
To saturate you use a common-emitter circuit like the 2nd of your diagrams, but you
must have a base resistor to control the current to the base and prevent damaging
the transistor or the Arduino driving it. Perhaps 180 ohm resistor is a reasonable
choice for switching upto 0.5A.
Your first circuit will only be able to get the transistors emitter up to about 4V if
driving from a 5V arduino, since the Vbe drop will be 0.7 to 0.8V, and the droop in
the Arduino output voltage under load could be as much as 0.5V.
As for PNP circuits they are just a mirror image of NPN, flip everything vertically. This
allows a high-side switch, note, as NPN switching circuit is on the low side since
its a common-emitter circuit.
Just assume that there is a resistor on base. I left it out of the schematics for the purpose of expediting the post. As for the LED, I used that simply for the purpose of illustrating a load. It could be anything, a motor, an LED, etc.
To clarify my question: I have been using T1 method (with base and limiting resistors as needed) in my applications, and it seems to work fine for what I have been doing with them. But in a effort to better educate myself (since this is a hobby/learning experience for me) I have been doing reading. And that is when I discovered the high vs low side switch. So the question is, when using a NPN transistor (and assuming I am using a base and current limiting resistor as applicable) is there any reason why I shouldn't use the method described in T1 rather than T3?
Paul, that is kinda what I was thinking. Typically, lets say I was using a TIP31 transistor and a DC motor, I would use a base resistor to adjust for the gain, to provide X current to the DC motor (load) that was connected to emitter. However, the article I was reading last night (I don't have a source at the moment) stated that circuits like this should be connected with the DC motor on the collector side. And that is exactly why I was confused. I have been using example T1 as my basic digital switch/amplifier circuit ever since I began toying with electronics without any issues. I'm mostly concerned that there may be a better way.
You are getting some confusing messages here. I stand by precisely what I stated before.
The emitter follower circuit is useful - under certain circumstances. Its purpose is to "amplify" current but not voltage. You lose the 0.7V base-emitter drop of the transistor. You never want to use a base resistor as this will simply lose more voltage - and then fail to regulate the voltage reliably.
It is very useful if you want to power something - such as LEDs - from a separate supply such as 12V but do not want to supply them with more than the supply voltage of the Arduino (less that 0.7 V or so) because the transistor dissipates the extra voltage rather than requiring a regulator to do so - of course it then needs a heatsink if any significant power is involved.
To then actually control the current to the load, you use a resistor between the emitter and the load.
Ok, Paul, I think I understand what you're saying. If I were running a 5v load, and needed to amplify current, but NOT at a higher voltage, the emitter follower circuit with any necessary resistors would be the way to go. However, if I were wanting to control a higher current, say 12v, by use of the switch then something more akin to the second schematic would be the proper way of doing so. Am I correct?
Yes, the "low side" switch allows you to switch the full voltage of the supply.
Now the limitation of the "high side" switch as it relates to microcontrollers, is that the "switch" itself is at the full supply voltage, so whatever you use to control the switch must be able to switch (off) that voltage in order to turn it off. In general, you need a low side switch to control the high side switch.
Ok, I think I see where the confusion had come it. I have only used a single supply voltage in my circuits. Which I guess is good enough for that application. So when I used a transistor to switch on a a load of a different voltage I would use something like this, but with a common ground, correct?
