Need help dissipating more power with transistors!

Discussion in 'General Electronics Chat' started by Konduction2, Sep 14, 2014.

  1. Konduction2

    Thread Starter Member

    Feb 12, 2014
    52
    0
    THE GIST: I want to know the MOSFET most similar to the TIP31c or a higher power BJT similar to the TIP31c, particularly one that can output more than 150 watts. Thanks!!

    I have successfully used the TIP31C as an amplifier. However, I just attached more power-sucking devices and need my amplifier to output more power. I do NOT want to put BJT transistors in parallel, as that never works too well. Therefore, I need help either finding the MOSFET that is the most similar to the TIP31C, OR finding a similar BJT that has more than 40W max power CEO (collector-emitter output). If there are many BJTs similar to the TIP31C that have higher power CEO, I'd like to know all of them that output between 40 and 200 watts please!
     
  2. ScottWang

    Moderator

    Aug 23, 2012
    4,853
    767
    Try the mje3055(npn) and mje2955(pnp).
     
    Konduction2 likes this.
  3. bertus

    Administrator

    Apr 5, 2008
    15,647
    2,346
    Last edited: Sep 14, 2014
    Konduction2 likes this.
  4. crutschow

    Expert

    Mar 14, 2008
    13,002
    3,231
    Note that it may be the power supply limiting the amp power, not the output transistors.
     
    Lundwall_Paul likes this.
  5. Konduction2

    Thread Starter Member

    Feb 12, 2014
    52
    0
    Everyone, sorry for the super long response time - I've been extremely busy and there were a lot of datasheets to go through. I'm still looking for a power FET similar to the tip31c.
    Particularly, I am looking for a FET that has:
    -a TO-220 package (or something similar that also works well with breadboards and can be easily attached to heat sinks)
    -a base-emitter (on) voltage - i think for FETs this is called the gate-source threshold voltage? - of preferably 1.8V like the tip31 or otherwise up to 12V.
    -a total power dissipation of at least 30W
    -drain to source voltage of at least 12V

    I think I may have found one, the IRF540N, but there are some things on the datasheet that confuse me. The sheet says the min and max gate-source threshold voltage are 2V and 4V. Does this mean that it is off when the gate receives 2V? More specifically, is its behavior between 2 and 4 gate-source voltage the equivalent of the TIP31C's behavior between 0 and 1.8 base-emitter voltage? If yes, are there any other power FETs with a minimum on gate-source voltage of 0V instead of 2V? Please explain if I've made any stupid assumptions :)

    ScottWang: Thanks, I'll use them in the future! Also, do you know any power FETs with the characteristics I just listed above?

    bertus: Thanks for the table! I've gone through it as much as I could but I did not find anything, and would still much appreciate your expertise with my questions updated above in this message. Also, I still value your direct expertise more, since I noticed there are many transistors missing from your table as well as an inconvenient lack of sorting options (no gate-source threshold voltages directly listed, or whether it is a FET vs BJT).

    crutschow: My power supply is an ample 650 watt computer PSU, so that is not my bottleneck.
     
  6. ScottWang

    Moderator

    Aug 23, 2012
    4,853
    767
    Could you attach your circuit and labeling the power voltage?
     
  7. Konduction2

    Thread Starter Member

    Feb 12, 2014
    52
    0
    Is it important? I am looking for a power FET to use for many different circuits. Sorry I am just curious why - I can draw a circuit easily if necessary.
     
  8. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    Transistors don't output power they control it.

    There is always a cost to be paid in that some of this power is lost by dissipation in the transistor itself. How much depends upon the circuit configuration. However the % lost can approach 100%.

    There ares six things to consider when designing to handle high power.

    1)Power
    2)Voltage
    3)Current
    4)Temperature / cooling
    5)Safe operating area (Otherwise known as the SOA)
    6)Time

    Failure to get any one of these right can result in destruction of the transistor.
    Correct operation is set by the configuration and values of the rest of the components in the circuit/ physical assembly
    We can discuss these if you like.
     
    Last edited: Sep 25, 2014
    Lundwall_Paul likes this.
  9. ScottWang

    Moderator

    Aug 23, 2012
    4,853
    767
    The bjt was controlled by current and the mosfet was controlled by voltage, you can't replacing them each other directly, can you design the circuit by yourself or the circuits already design completed?

    The different circuit could be request the different mosfet, so you can't using one kind of mosfet to suit all the circuits, maybe two or three different specs, they depend on how the voltages and current of the circuit.
     
    Konduction2 likes this.
  10. Konduction2

    Thread Starter Member

    Feb 12, 2014
    52
    0
    studiot: Yes, I completely agree. Below my gist, you can see that I did specify what exactly I was talking about: "max power CEO". And I also do understand that many parameters must be right in order to reach the max.

    ScottWang: Ah, I see - thank you! I can change the circuit. I have attached an old schematic from when I was making a hard drive speaker. What I have now is essentially the same, but instead of the capacitor and hard drive, I have 25 meters of LED strips (they are SMD5630, which take a ton of power). The op-amp shown is an LM386 audio amplifier. The LED strips sync fine with the music with only 5 meters, but I know my BJT will die if I attach the remaining 20 meters. This is why I would like to get some power FETs to put in parallel, to replace my BJT.
     
  11. Konduction2

    Thread Starter Member

    Feb 12, 2014
    52
    0
  12. tcmtech

    Well-Known Member

    Nov 4, 2013
    2,034
    1,653
    I see now. It's a LED string driver and not a speaker driving amplifier you are working with. Given that most any power transistor with a higher HFE gain value and higher power handling capacity would likely work just fine.

    As far as package size goes the TO-247 and similar case sizes are fairly easy to re fit to a TO-220 package replacement.
     
  13. ScottWang

    Moderator

    Aug 23, 2012
    4,853
    767
    LED Type: 5630 LED
    Input Voltage: DC 12 Volt
    Length: 16.4 feet 5m
    Power: 150 W
    http://www.homeled69.com/faq_17.html

    According to a product made from SMD5630, 5 meters have 150W, so if you using 25 meters maybe have 750W, the power voltages is +12V, then the I=W/V=750W/12V=62.5A.

    If the above values was correct, and then you have to choosing the specs of mosfet as 2.5 times of 12V, at least 3 times of 62.5A, the total Ids current about 187.5A, because you want to in parallel with all the mosfets, so choosing the Vds big than Vds=12V*2.5=30V is the priority, I was listed some mosfets were matched what I mentioned about the V/I, and the Rds more less can be afford more current.

    BUK9507-30B_Nch_30V75A157W_7mΩ_Vgs5V.

    IPP042N03L_Nch_30V70A-79W4.2mΩ.
    IPP096N03L_Nch_30V35A_9.6mΩ_Vgs4.5V.
    IRL3803_Nch_30V140A200W_6mΩ.
    STB60NE03L-10_Nch_30V60A_10mΩ.

    FDP8440_Nch_40V80A306W_2.4mΩ_Vgs4.5V.

    IRF1010EPbF_Nch_60V84A200W_12mΩ.
    STH260N6F6-2_Nch_60V180A300W_2.4mΩ_Vgs10V.

    IRF3205_Nch_55V110A200W_8mΩ.
    IRF3205s_Nch_55V110A200W_8mΩ.
    IRF1405_Nch_55V169A330W_5.3mΩ.

    IRFP2907_Nch_75V209A470W_4.5mΩ.
    FDP5800_Nch_80V60A_6mΩ_Vgs10V.
    IRFB4710PBF_Nch_100V75A_14mΩ.
    IRFPS3810_Nch_100V170A580W_9mΩ.
     
    Last edited: Sep 26, 2014
    Konduction2 likes this.
  14. Konduction2

    Thread Starter Member

    Feb 12, 2014
    52
    0
    tcmtech: yep. Since I'm planning on putting FETs in parallel, power handling capacity of each FET doesn't really matter. I was more concerned about the gate-source threshold voltage (see question below).

    ScottWang: Thanks! I'm really sorry, but I should have mentioned that I've multimetered these LED strips and the power load is wayy lower than what they should be (but I don't care.. they were same price as weaker ones anyway). Each strip of 5 meters only takes 0.8A at 12.3V. Whatever - I can still make good use of the concepts from the info you gave me though! However, could you still help me answer the question below please?

     
  15. ScottWang

    Moderator

    Aug 23, 2012
    4,853
    767
    The IRF540 is not a logic level mosfet as 4.5 Vgs, so you just don't focus on the 2V or 4V threshold voltage, the datasheet shows that the Vgs=10V when the Vds is on.

    RDS(on), Static Drain-source On Resistance, Vgs = 10 V, Id = 11 A, Rds type=0.055 Ω, Rds(max)=0.077 Ω.

    The bjt was drived by the current, but the mosfet was drived by the voltage, if you need a logic level Vgs as Vgs=4.5V then you can choosing from the parts I listed.

    If you want a low level voltage as Vbe of bjt then you can using a npn bjt to drive a P type mosfet.
     
    Konduction2 likes this.
  16. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,791
    1,103
    A threshold range of 2V to 4V means that, because of manufacturing tolerances, Vgs(thr) could be anywhere from 2V to 4V. You should design for the worst case and assume you need 4V at least. At the Vgs threshold a FET is only beginning to conduct (a few uA or mA). To ensure a FET is turned on fully, Vgs should be several times the threshold voltage. For most FETs that means ~10V.
     
    Konduction2 likes this.
  17. Konduction2

    Thread Starter Member

    Feb 12, 2014
    52
    0
    Can I put two different FETs in parallel for higher power capacity, if all of the Vgs values are the same?(assuming I don't go over the max voltage values for the weakest FET)

    ScottWang: Ah, okay, thanks! I think I will use your last suggestion.

    Alec_t: Ohh I see, that made it so much clearer for me - thanks!
     
  18. ScottWang

    Moderator

    Aug 23, 2012
    4,853
    767
    For the Vgs of mosfet even the same ID number also could be different, so if you can just don't in parallel with the different mosfet.
     
    Konduction2 likes this.
  19. crutschow

    Expert

    Mar 14, 2008
    13,002
    3,231
    You can parallel BJTs with a small emitter resistor to equalize the currents in an amplifier but the variation of Vgs between MOSFETs, even of the same model, precludes easy paralleling of them. The MOSFETs would have to have their Vgs values matched to within a few mV for that to work.
     
    Konduction2 likes this.
  20. Konduction2

    Thread Starter Member

    Feb 12, 2014
    52
    0
    Whoa whaaat. That is surprising and disappointing haha. Before I started this thread, whenever I did Google searches of putting BJTs in parallel, I just kept seeing recommendations to put FETs in parallel instead. Were those recommendations bad, or are there special power FETs that are manufactured with higher consistency that can be paralleled? All of the FETs that I've just acquired (the IRF540N, IRF511, IRFP140A) or have seen so far have a big range of possible Vgs
     
Loading...