BJT as electronics switch

[bananasiong]

Hi,
I want to learn more about using bjt as electronics switch. Usually I will connect the load to the collector, the emitter to GND and the signal to the base (for NPN). But I don't know which NPN BJT to be used, I always use different transistors to try, until I got it. This is because I don't know how to read the datasheet.
Last time, i use a tip41 to switch a relay with 5v, i feel i was so lucky for that.

If i want to use a 2N3904 as a switch, with 5V, what should the load resistance be? Can anyone teach me to read the datasheet? Let's start with 2N3904.

https://www.electro-tech-online.com/custompdfs/2006/08/2N3903-DPDF.pdf

thanks!

 

[akg]

the first thing to decide is how much you expect it to switch , (say if 80ma ) , then select a trans that can pass more then that current, here your 3904 can handle 100ma (200ma Max) , now look at datasheet to check what will be the hfe at that current because it will decide the base current, you can see that it is 30 , so design the base ckt as per this , now at 100ma ,and max vce(sat) is 0.3v , so the max continous power dissipation u expect 30mW , require a heatsink ? u deside ?. remember during switching , before reaching saturation the transistor is going through the active region where the power dissipation will be very high . so go for max switching speed(ref: switching char: ), and minimum switching

 

[audioguru]

Hi AKG,
The hFE rating isn't used for a transistor acting as a saturated switch. Its saturation rating of max 0.3V is with a collector current of only 50mA and might be double at 100mA, and with a base current of 1/10th the collector current.

 

[akg]

oohh.. thanks AG

 

[akg]

it means we'll require more base current , and the power dissipation will be also more than what i said above.

 

[bananasiong]

say if 80ma , then select a trans that can pass more then that current, here your 3904 can handle 100ma (200ma Max)

Which part of the datasheet mention that it is 100mA and 200mA max? Is it at 'ON characteristics'---(IC = 100 mAdc, VCE = 1.0 Vdc) ? I don't know the way to read the datasheet... sorry about that..

now at 100ma ,and max vce(sat) is 0.3v , so the max continous power dissipation u expect 30mW , require a heatsink ? u deside ?. remember during switching , before reaching saturation the transistor is going through the active region where the power dissipation will be very high

Usually how much is the power that considered as high and need heatsink?


So, the point is, determine how much current to be switched but not voltage of resistance?
Then the current required to the base is 1/10th of the amount of the current u want to switch??
How if i don't know how much current to switch for example a DC motor?
Measure the resistance of the DC or what?

 

[akg]

Which part of the datasheet mention that it is 100mA and 200mA max?

Maximum ratings -> collector current continous ( remember it should be read along with the max device temperature)

Usually how much is the power that considered as high and need heatsink?

i donno excatly , but it will depend on the surface area of the device , the thermal resistance , device operating temp , ambient temp etc .
the objective is to always keep the temp of the device within normal limits

So, the point is, determine how much current to be switched but not voltage of resistance?
Then the current required to the base is 1/10th of the amount of the current u want to switch??

the current will depends on all the above factors . the current handling and dissipation of the semiconductor is focus.

How if i don't know how much current to switch for example a DC motor?
Measure the resistance of the DC or what?

yes you can , but remember the running current of a dc motor will be far less than the current when starting (ref: back emf).
I assume you have noted the AG's post above .

 

[audioguru]

Usually a small transistor doesn't use a heatsink.
The Thermal Characteristics spec's on the datasheet for a little 2N3904 transistor has a max Total Device Dissipation of 625mW when the ambient air is at 25 degrees C and it says to derate at 5mW/degree C when the ambient air is warmer.

Its absolute max continuous current is 200mA. If you want it to survive then don't use it to turn on a motor that draws more when starting. The resistance of a motor indicates how much current it will draw when starting.

 

[philba]

the issue of "what's high power or not" is dependent on the package's ability to dissipate heat. You didn't include the thermal characteristics but fromthe fairchild 2n3904 datasheet, the to92 package has a thermal resistance (junction to ambient) of 200 C/W. That means each watt of dissipation will raise the device temperature by 200 C. So, you just multiply this times your computed dissipation (Vcesat * Ic) and add the ambient temperature and you have the device's temperature. example: if you are dissipating 100 mW, your temperature will be 20C higher than the ambient temp. At 25C, that would make it 45C. nothing to worry about. In fact, with a low Vcesat like the the 2n3904 has, you will probably blow the transistor long before it needs a heat sink.

Those heat calcs apply to any device. Try it for a darlington transistor which has a high Vcesat to get a sense of where it really makes a difference.

 

[audioguru]

If the transistor has a high current load and it switches slowly, then it will smoke and die.
A 20V supply and a 100 ohm load makes a turned on current of about 200mA which a 2N3904 can drive. But if it switches slowly then it spends time with 10V across it and 100mA through it which is a power dissipation of 1W! Its absolute max power dissipation is only 625mW.

 

[akg]

'bananasiong' , the above posts should have cleared your questions

 

[bananasiong]

Oh.. thank all.. but i'm still confused. If i want to switch a DC motor or relay, the first thing to do is measure their resistance, then the voltage supplied, then use I=V/R and check which transistor to be used, right? Not only the current, I need to make sure that the power dissipated is supported by the transistor.

*correct me if I'm wrong*
thanks a lot..

 

[Nigel Goodwin]

Oh.. thank all.. but i'm still confused. If i want to switch a DC motor or relay, the first thing to do is measure their resistance, then the voltage supplied, then use I=V/R and check which transistor to be used, right? Not only the current, I need to make sure that the power dissipated is supported by the transistor.

*correct me if I'm wrong*

Partially wrong, and partially right.

It's fairly pointless measuring the resistance of them, particularly the motor (which will be meaningless), and the relay will have it's resistance printed on it, or be in the datasheet.

Dissipation shouldn't be a problem, as the transistor should be saturated (low voltage drop), and any transistor able to pass the current is likely to have more than enough dissipation.

Motors are VERY different though, and their current requirements vary MASSIVELY according to load - and if you stall the motor it becomes huge, expect ten to twenty times the normal running current!.

 

[audioguru]

Hi Nigel,
Don't you think the current of a stalled or starting DC motor is due to its DC resistance?

 

[Nigel Goodwin]

Hi Nigel,
Don't you think the current of a stalled or starting DC motor is due to its DC resistance?

Certainly in part, but you don't spec the driver transistor to continuously handle that current - or you would certainly have some seriously large transistors!.

 

[audioguru]

Of course a motor driver transistor must be spec'd to handle the max current of the motor.
The absolute max current for a 2N3904 is 200mA. for a 2N3055 it is 15A. Their max current rating is the same for continuous or for instantaneous current.

 

[Hero999]

Bear in mind the starting current will be slightly lower than what you'd expect due to the inductance.

 

[Nigel Goodwin]

Of course a motor driver transistor must be spec'd to handle the max current of the motor.

You have to limit the current in some way, even if it's just the internal resistance of the batteries - or do you think the ULN series motor drivers are rated at ten or twenty amps?.

 

[audioguru]

do you think the ULN series motor drivers are rated at ten or twenty amps?

I wouldn't use a dinky little ULN2003 to drive a motor that draws more than its max rating of only 500mA. I would use it or some other driver to drive a power transistor or Mosfet with a high enough current rating.

 

[Nigel Goodwin]

I wouldn't use a dinky little ULN2003 to drive a motor that draws more than its max rating of only 500mA.

Have you ever seen a motor which draws under 500mA when stalled?.