Hello im building a circuit with that works as a switch, my load requires a current of 300mA, and its an inductive load so i selected a diode that Is placed in parallel to It.

I have searched for BJTs and selected the most common one, which Is 2N2222 that can deliver almost 1A so its more than fine.

for the designing part i read that in saturation region u usually select a Beta gain of 10, since the one shown in datasheet are shown but they consider the BJT working as amplifier.

So now here the calculations i made:

If i have 5V driving the base then what u will have Is 5-0.7/RB=30mA, which means RB = 143 Ohm ( I also have seen on datasheet that he can handle that current at the base )

Is like that correct? Now if i consider that im driving this transistor from a mcu there are considerations to make?
Secondly i selected the diode , the one i chose is 1N4148, which has a reverse voltage of 75V ( supply voltage is at 12V so in kinda far from such voltage )


Thanks for the help ( im kinda a newbie, so sorry in case its too simple for ya, i gotta start from somewhere)

 

I also have seen on datasheet that he can handle that current at the base

Looks ok, but who is "he" that can handle the current?
Can the 5V source deliver 30mA?

 

A lot of MCUs aren't going to like delivering 30 mA. If you want to stick with using a BJT, you could consider using a Darlington arrangement, which would reduce the load on your MCU by a factor of 10 to a more reasonable 3 mA.

Also, note that there is nothing magical about a beta of 10 for operation in saturation -- that is really a convention that took hold in the early days of transistors. For ease of comparison, manufacturers specify saturation performance by measuring the parameters at a forced-beta of ten.

You might also look at using a logic-level MOSFET as your switch.

 

Looks ok, but who is "he" that can handle the current?
Can the 5V source deliver 30mA?

the transistor i meant, he can handle 30mA at the base.

Yeah the 5V source can deliver 30mA btw!

 

A lot of MCUs aren't going to like delivering 30 mA. If you want to stick with using a BJT, you could consider using a Darlington arrangement, which would reduce the load on your MCU by a factor of 10 to a more reasonable 3 mA.

Also, note that there is nothing magical about a beta of 10 for operation in saturation -- that is really a convention that took hold in the early days of transistors. For ease of comparison, manufacturers specify saturation performance by measuring the parameters at a forced-beta of ten.

You might also look at using a logic-level MOSFET as your switch.

But taking a beta of 10 Is the right "approach" like as a hw designer u would have done that?

Btw yeah i can try with MOSFET too, since they are controlled in voltage, and they dont require current at gate. Thanks

 

Btw yeah i can try with MOSFET too, since they are controlled in voltage, and they dont require current at gate. Thanks

Well, the gate has a capacitance you need to charge to fully turn on at a given voltage above its voltage threshold. Guess how a capacitance is charged?

 

Well, the gate has a capacitance you need to charge to fully turn on at a given voltage above its voltage threshold. Guess how a capacitance is charged?

Yes, there is some transient current -- that is true even if nothing is connected to the output pin of the MCU since everything has some capacitance to the rest of the universe. For a device suitable for this application, the total gate charge is probably on the order of a few nanocoulombs, which a microamp current would supply in a few milliseconds. Putting a 5.1 kΩ resistor in series with the gate would keep the peak current below 1 mA and still turn on the transistor in a few microseconds.

 

My answer was more related to the fact that is common to read that MOSFETs are voltage controlled/driven, meaning (or inducing to think that) no current required, which is, in general, wrong, despite for a simple switch application might be not so relevant.

Otherwise gate drivers wouldn't have sense to exist.

 

you could just add another transistor to get overall higher gain... then the input current will be low.

 

The BC817-25 Hfe≈ 250 as a hearty transistor to handle 500mA motor and murphy's law. The motor field, not like an led load.
Low side motor switch mode arrangement is NPN, the arrow points down like a switch points down. There is no mosfet here.
The MCU 5V io pin connects to Rb=2.1k calculated as 2mA Saturation that is why I mentioned Hfe ≈250 (gain, using switching region)
The Motor has 1N5822 across it, why so big? because 12V inductive kick might exceed 1N5819 20V rating and human intervention torque.

The diode has a silver band indicating +cathode and it is reverse biased connecting + to + On low side switching cathode faces up.
If EMI such as ringing is problem it can be handled with a snubber, the 100nF cap on 12V and series 30Ω R to GND, also
on the MCU Vin may need 10uF for current droop mentioned earlier. Since it is a shared GND. that is; the MCU shares 12V GND only. called star GND.
A small base current switches a large emitter current. 50k R from Base to emitter keeps the zero state, that is the motor remains off when 5V io goes low.

 

You won’t get 30 mA directly from the micro. I would suggest using using a Darlington.

What is limiting the collector current? You cannot expect it to be limited by the base current as this is not very well controlled, especially if you use a Darlington.

How fast is it switching? The inductance will limit the current if the on time is low enough.

dI/dt = V / L

So, let’s take an inductance of 1mH as an example. With 12V across it, the currentvis

dI / dt = 12 / 0.001 = 12000 A / sec

That is 12000 uA / usec = 12 mA / usec

So, if you leave it in for 25 usec it will reach 300 mA.

 

But taking a beta of 10 Is the right "approach" like as a hw designer u would have done that?

10 is a safe value for 2N2222 under any circumstances, but other values are possible.

Here's a thread where I posted a saturation experiment for half a dozen different transistors. Table from that thread attached for convenience.

Two randomly selected 2N2222A both saturated using beta of up to about 130 (for a collector current of about 30mA - beta typically reduces at higher collector currents). I wouldn't recommend using a value higher than 10 without cherry-picking them (i.e. testing before using). While also making a notation on the schematic that they were cherry-picked in case the transistor ever needs to be replaced.

 

10 is a safe value for 2N2222 under any circumstances, but other values are possible.

Safe if something else is limiting the current. I think the TS believes he can limit the collector current by limiting the base current. This is a very bad idea.

His load is an inductor. The only thing limiting a static current is the resistance of the coil. For 12V applied, the resistance would have to be 12/0.3 = 40 Ohms. At that point, it is more of a resistor than an inductor.

See my earlier post to see how the inductance can limit the current.

 

The sketch provided by the TS implies that the inductive load is, in fact, a motor.

 

I missed that.

 

you could just add another transistor to get overall higher gain... then the input current will be low.

U mean like a darlington pair driving a second transistor?

 

The BC817-25 Hfe≈ 250 as a hearty transistor to handle 500mA motor and murphy's law. The motor field, not like an led load.
Low side motor switch mode arrangement is NPN, the arrow points down like a switch points down. There is no mosfet here.
The MCU 5V io pin connects to Rb=2.1k calculated as 2mA Saturation that is why I mentioned Hfe ≈250 (gain, using switching region)
The Motor has 1N5822 across it, why so big? because 12V inductive kick might exceed 1N5819 20V rating and human intervention torque.

The diode has a silver band indicating +cathode and it is reverse biased connecting + to + On low side switching cathode faces up.
If EMI such as ringing is problem it can be handled with a snubber, the 100nF cap on 12V and series 30Ω R to GND, also
on the MCU Vin may need 10uF for current droop mentioned earlier. Since it is a shared GND. that is; the MCU shares 12V GND only. called star GND.
A small base current switches a large emitter current. 50k R from Base to emitter keeps the zero state, that is the motor remains off when 5V io goes low.

How u usually choose the RC costant of the snubber?

 

U mean like a darlington pair driving a second transistor?

darlington is one option...

it not necessarily the option i would lean towards since it has pretty large voltage drop of about 1.4V. relays are not too sensitive so it still should work fine. however, even slight change in circuit would make that aspect better