Actions
Forum Thread Details
  • Replies 11 replies
  • Subscribers 474 subscribers
  • Views 993 views
  • Users 0 members are here
Related

Need help choosing a PNP transistor

Catwell

image

 

When I power the base, I am cutting off the current flow through the Collector and Emitter.

 

However, when the base is off, I will need to conduct 4.2W through the transistor.Now, I am having trouble sourcing a transistor for this situation.

 

This is for a automotive part, in the end. So, if anyone has ever worked in that industry, what % do you go over the bare minimum ratings per transistor? No one here at this job seems to want to help me. The curse of being a contractor.

 

Cabe

 

Engineers are more than just designers, talk about your day to day here in the EDE Life Discourse Group.

Parents Reply Children
  • in reply to Catwell

    Cabe,

     

    This transistor is going to a load right? If that is the case the transistor is not dissipating all 4.2W of power otherwise there would be none left for the load. Maybe you're doing it as a safety margin sort of thing, but you need to be specific because in the long run overspecing can cost you money.

     

    Looking at the datasheet the VCE saturation voltage for this transistor is not specified, however a CONSERVATIVE assumption of 1V VCE drop (most transistor have a Vce drop 0.2-0.6 in saturation), you can determine how much power is lost the the transistor as heat.

     

    In your case 1V x 280mA= 280mW of power lost in the transistor. Looking at the NTE397NTE397 data sheet, it says that maximum power dissipation at 25C is 10W, it specifies to derate this value by 57mW/C therafter. I don't know how hot a car can get but assuming (Extreme Case, I know) the transistor was operated with at 125C(257F) ambient the transistor could still dissipate 4.3W (10W- 100C*57mW/C) so you're golden.

     

    You can safely use the NTE397NTE397, and you'll still have plenty of margin.

     

    Best Regards,

  • in reply to Jorge_Garcia

    Jorge,

     

    This circuit is driving LEDs on a car, so I want to make sure nothing overheats. After some deliberation, I am going to go with a MOSFET alternative. The price of the transistors I chose before is just too high.

     

    Cabe

  • in reply to Catwell

    That's fine Cabe,

     

    I've worked with HB LEDs my whole career up to this point. When you build these circuits take measurements of the voltage across the CE(SD) of your transistors. If you are using them as switches (not regulators) then you will find that the power dissipated across the transistors is not equal to the power coming into the circuit.

     

    I just don't want you to spend more money than you have to because of overspeccing your components. Paying an extra 5 cents for a transistor with a higher spec is irrelavant for low quantities, but in a commercial application with 100,000s of units 5 cents per component adds up.

     

    I get the feeling that my calculations are being blown off, and that's ok you don't have to agree with me. But take measurements of the circuits you have that's the greatest proof (or disproof) of anything I've stated here. Measure the current going through your transistor and measure the voltage across CE of your transistor this will give you the power dissipated as heat by the transistor.

     

    Just check it out, mosfets are great I've used them as current mirrors in my 40W LED boards and have had no problems( I use Aluminum PCBs for max heat transfer).

     

    hth,

    Jorge Garcia

  • in reply to Jorge_Garcia

    Jorge,

     

    I am actually changing my design based on your price suggestion. I was originally going to use a $3.00 USD transistor, but now I am switching to a $0.04 Mosfet. So, thank you.

     

    My next step is sizing the Mosfet appropriately.

     

    Cabe