High Amperage MOSFET Switch

Discussion in 'The Projects Forum' started by JdogAwesome, Mar 24, 2016.

  1. JdogAwesome

    Thread Starter New Member

    Mar 24, 2016
    9
    0
    Hello! Ive been looking around on the internets trying to figure out how to build a switch to turn my Electric Longboard on and off. I have a build thread for my longboard here if youd like to see the entire build. Anyways im currently using this as the switch to turn the board on and off, and its such a piece of crap. Both the red buttons have broken off so I need to use pliers to turn the damn thing on and off, anyways enough complaining. So I clearly need to get a different switch so instead of buying another one of those I wanted to try making something like this. Its just a switch to turn the board on and off but using a pretty blue LED switch connect to a circuit that actually handles the high current. So I need your guys help to deign a circuit and what kind of parts I need to make this happen. The board runs at max 25.2V and around 10-20A crusing though burst current can draw up to, lets play it safe, 90A. I was thinking I would use a couple of mosfets in parrell to control the current and have the switch toggle the mosfets.
     
  2. crutschow

    Expert

    Mar 14, 2008
    13,033
    3,241
    You can use one large N-MOSFET if you have no problem with switching the negative side of the battery. (N-MOSFETs are cheaper than P-MOSFETs).
    Find one with a rating of at least 40V and an ON resistance of no more than a milliohm. That way you won't have to heatsink the MOSFET.

    Here's the circuit:
    The MOSFET is ON when the SW contacts are closed.
    The diode is to prevent any inductive spikes when the load is turned off. It should be connected close to the switch and MOSFET source connection.

    If it's more economical to use more than one MOSFET, just connect the sources, drains, and gates together of the added MOSFET(s). MOSFETs automatically share a load so no load sharing resistances are needed (as long as they are all the same part no.).

    upload_2016-3-24_11-45-47.png

    Zapper
     
    Last edited: Mar 24, 2016
  3. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    The MOSFETs on old PC motherboards usually weigh in at around 80A, but you have to watch the voltage, some are only 30V and some are even as low as 20V. On the bright side, VGSthr can be as low as 1.6V.

    MOSFETs are dead easy to parallel as long as all are identical.
     
  4. Roderick Young

    Member

    Feb 22, 2015
    408
    168
    Interesting. Is there a reason that R1 and D1 are on the load side, and not the battery side? It seems that the gate of the MOSFET might be exposed to a brief transient of more than 20V at initial turn-on?
     
  5. crutschow

    Expert

    Mar 14, 2008
    13,033
    3,241
    Good catch.
    Yes the resistor should be connected to the source, not the drain, as the corrected circuit shows below.

    But D1 must be on the load side to suppress any inductive spike from the load or wire inductance.

    upload_2016-3-25_1-54-41.png
     
  6. JdogAwesome

    Thread Starter New Member

    Mar 24, 2016
    9
    0
    Thank you for your circuit diagram! I uploaded a simplified version of your diagram just to make sure ive got everything right. I was planning on using a IRFP064N MOSFET though it has an RDS ON of 0.009Ohms which is obviously over a miliohm. If you think that mosfet would not be suited for the job do you have any that would be. Also could I use a RL207 diode(s) because I have some on hand. Also the longboard has regenerative braking which, obviously, sends power back to the battery, could that be a potential problem? Again thank you for your help!
    Switch Diagram V2 Simplified.jpg
     
  7. crutschow

    Expert

    Mar 14, 2008
    13,033
    3,241
    0.009Ω will cause a transistor dissipation of (20² * 0.009 = 3.6W) at 20A so the transistor will need a heatsink.
    That's why I suggested a milliohm or less since the transistor dissipation would be no more than 0.4W and wouldn't require a heatsink.

    The RL2007 diode is fine.

    Regenerative braking, which would generate a current in the opposite direction, is not a problem since MOSFETs conduct equally well in either direction when ON.
    Even if you happened to shut the transistor off when regeneration braking was occurring, the MOSFET substrate diode would carry the current.
     
    Last edited: Mar 27, 2016
  8. JdogAwesome

    Thread Starter New Member

    Mar 24, 2016
    9
    0
    Because the switch is going to be inside of my fiberglass enclosure with minimal to no air flow I should probably look for a different mosfet. The IRLB3034PBF-ND looks like a good option, it has an RDSON of 1.7 mOhm which means I only have to dissipate 0.68W so I could use a small heatsink. Also thank you for straightening out my question about the regenerative braking!
     
  9. crutschow

    Expert

    Mar 14, 2008
    13,033
    3,241
    You may still not need a heatsink.
    If you look at the first page of the MOSFET data sheet you will see that the junction thermal resistance to ambient is 62°C/W and the maximum junction temperature is 175°C.
    So for 0.68W dissipation, the junction temperature will be 0.68 * 62 = 42°C above ambient.
    This means the ambient (still) air can be up to 175-42 = 133°C (271°F) before you reach the maximum junction temperature.
    So unless there're some heat producing parts inside the enclosure (motor?) that will bring the ambient air to that temperature, you should be okay with no heatsink.
     
  10. JdogAwesome

    Thread Starter New Member

    Mar 24, 2016
    9
    0
    I was actually just looking up how to figure out what size heatsink I needed though you made it far easier for me! The only heat producing component inside the enclosure is the ESC though it doesn't get very hot, definitely not 271°F hot so I should be fine without a heatsink. Alright ive decided to go with the IRLB3034PBF I will post a reply once ive built the switch and ive tested it thanks again!
     
  11. JdogAwesome

    Thread Starter New Member

    Mar 24, 2016
    9
    0
    So after ordering the MOSFET, building the switch and installing it in the Longboard, it turned out great! Ive tested it a couple of times and so far it has worked very well without any problems. Ive attached some pictures below if you'd like to see it. Once again thank you crutschow for all your help!

    P_20160406_113057.jpg P_20160406_113123.jpg P_20160406_114627.jpg
     
Loading...