Help with a NPN Enhancement FET Suggestion

Discussion in 'General Electronics Chat' started by Management, Jan 14, 2010.

  1. Management

    Thread Starter Active Member

    Sep 18, 2007
    306
    0
    Hi guys,

    I was doing a little simulation with TINA (TI's SPICE) and I was using a NPN Enhancement Type Fairchild Semiconductor 2N6767. Works great in the simulation but it is old.

    Anyone know of a similar part with similar specs that is available from digikey?

    One that comes in both leaded and surface mountable. The former so that I can do some small testing on a bench.

    Thanks.
     
    Last edited: Jan 14, 2010
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Do you really need a 400V MOSFET?

    If you don't, select one with a lower Vdss rating. Somwhere around 1.3x to 2x the max voltage expected should be about right. You can use Digikey's parametric search to narrow things down. Start by searching for MOSFET, select single MOSFETs, and then N-channel (be sure to check the "In Stock" box) etc.

    The higher the Vdss rating for a given current rating, the larger your Rds(on) and total gate charge will be.

    If you're dealing with >200V circuits, you should really consider IGBT's instead.

    If you post your circuit, perhaps better suggestions could be made.
     
  3. Management

    Thread Starter Active Member

    Sep 18, 2007
    306
    0
    Can't really post it but I really appreciate the suggestions. Is it a rule of thumb to go 1.3x to 2x of max voltage to expect? I'll try to follow that.

    I noticed that in an older design that was followed. FETs with 50V and 60V drain to source voltage (largest supply was 28V). There was one FET though that had 200V Vds. It was used for a current source. It is for industrial applications so I figured that the location of the FET and what it was used for justified the rating. It had a Rds on of .6 Ohms.

    I used Digikey like you said I ordered a few IRFR220 that are made by International Rectifier. Simulation shows it works. I'll see in the lab.

    Thanks for your suggestions.

    What else drives the choice of a FET in current source applications?
     
  4. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    It depends on your application. If you want it more high-rel at the cost of higher gate charge, go for a higher Vdss. You'll have to suffer with a larger gate charge to get the Id back up.

    If you're using a MOSFET in a linear mode, the gate charge doesn't matter so much; you're not changing Vgs very rapidly anyway. The Rds(on) probably won't matter much either; it would just limit the maximum current. What WILL matter is package dissipation. You'd probably want a TO-3 device in a really big copper heat-sink, or cool it with liquid. Water conducts heat 1.4 times better than copper. Aluminum is about 60% as efficient as copper, depending on the alloy.
     
  5. Management

    Thread Starter Active Member

    Sep 18, 2007
    306
    0
    All great points. Didn't know any of it before this. Thanks a lot.

    Operating in the linear region is what I am doing. The issue isn't so much during normal operation but for compliance. Operating in much loooooower currents but the 5x like rating serves as protection. Circuit must be capable of handling situations where unusually high instantaneous currents or voltages occur. That is my understand from a old design.

    In any event, I just learned a great deal so thank you for your quick replies. Typically little knowledge like this comes from experience (like getting burned - not literally).

    Power Dissipation will be an issue so that TO-3 provides the best out of all packages? I'm trying to choose ICs as well that have lower power dissipation.
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Frankly, I can't tell you offhand what package would be the best.

    However, I just did a search on Digikey for N-ch enh MOSFETs that they have in stock with a minimum voltage rating of 350, and power rating of 750W or more, and came up with this for the least cost:
    http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=FDH44N50-ND

    If you want mil-spec components (which wouldn't be a bad idea from a reliability standpoint) your costs will go up considerably, and availability could be a problem.

    Make certain that whatever you decide upon in your design that it is a fairly common package, it exceeds your power/current requirements, and is not planned for obsolescence. Check the manufacturer's site to see if it's on a "lifetime buy" list or the like.

    The TO-3 package is hermetically sealed; the lid is welded on, and the pins are held in glass thru's. TO-3P is plastic, and NOT hermetically sealed. The TO-247 cannot be hermetically sealed, nor can any other package that has plastic bonded to metal. The hermetic seal may be of concern depending upon your intended environment.
     
    Last edited: Jan 14, 2010
Loading...