Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Switching power circuits with small control circuits

  1. Feb 3, 2010 #1

    I would imagine this is a quite a common concept in electronics. I am currently trying to make a small circuit which will switch a voltage supply to a small 12V motor on and off. I've had some help on here which has enabled me to pin down the type of circuit I need to get the output I want.

    I do not know how to get this output to switch a circuit with much larger currents circulating however. I had a look at some Power transistors on RS. They were only sold in batches which made the minimum cost £150!!! Is there any way you can get a small output of say a few hundred millimaps to switch a circuit which is moving say 20 amps or more at 12V? I looked at some small relays which would have done the job but their oscillation time would have disturbed my intended frequency requirement and also their operational lifespan is too small for something that switches so frequently.

    Does anyone have any ideas what I can do here? Thanks>
  2. jcsd
  3. Feb 3, 2010 #2


    User Avatar
    Science Advisor

    Hi Adder. At the power and voltage range you're looking at you should be able to find a power mosfet to fit the bill. Look for low voltage n-channel power mosfets, they're reasonably cheap and readily available in that voltage/current range.
  4. Feb 3, 2010 #3
    Sounds great, cheers!
  5. Feb 3, 2010 #4
    I found this one. Is it any good? I see the Voltage rating looks high and so does the current but the power rating is only around 208 watts is that enough? Does this mean I have to design my circuit so not more than 208 watts is spent in the transistor?


    I'm not really too sure how to interpret what some of those ratings factors mean. I don't know what the following things stand for:

    ID Max
    PD Max
    Vds Max
    rDS Max

    Can anyone explain what they mean? I can figure it out myself how to make any small design changes once I know what limitations I'm really dealing with. Thanks!
    Last edited by a moderator: Apr 24, 2017
  6. Feb 3, 2010 #5
    Hi this looks like a good link, it specifies most of the stuff I needed to know right at the end. Might help someone else passing through this thread in the future:

    http://en.wikipedia.org/wiki/Power_MOSFET" [Broken]

    *ID means Drain Current I believe. Which is basically the amount of current flowing through the transistor's 'main pathway'.

    *Vds means Drain to Source Voltage. The voltage trying to push current through it from the power circuit.

    *PD Means drain power. Although niether ID or Vds can go above their maximums, neither can their product go above the PD value or too much work is done and it damages the transistor.

    *rDS means drain to source resistance. Has some relevant significance which I can't really remember.

    So to conclude:

    I lost my link before it disappeared. In case it disappears again here's the transistor's details:

    ID MAx = 20.7A

    Vds MAX = 600V

    PD Max = 208W

    Rds Max = 0.19 ohms

    So I can be sure my 12V dc input won't hurt the transistor and will give it a long lifetime. Taking the resistance to be 0.19 ohms max I can assume that resistance will be in series with my motor windings, so I can plan to keep the max current beneath 20.7A. At 12V 20A will give me 240W of power which should be plenty for what I need.

    For that obviously I need to work out roughly what the average resistance of the motor windings will be and maybe put an extra resistor in to reduce the current through the transistor to keep it at a safe value.

    BUT!!!!!! I am thinking of putting an inductor in parallel with the motor windings. Idea is rather like a car's ignition coil, when the magnetic field collapses it will give one huge but very short blast into the motor. Sounds great but I am very worried it risks wrecking the transistor. If I have my feed into the power transistor run through an inductor rather than direct through the model's motor windings is there a risk my transistor will have a large reverse bias voltage try to push current through it for a short time, or will this energy all get spent through the motor windings which will be in parallel with the inductor?

    HELP! Sorry not sure this one it could wreck my plans. Should I leave the inductor out of it? Thanks!
    Last edited by a moderator: May 4, 2017
  7. Feb 3, 2010 #6


    User Avatar
    Science Advisor

    You have those ratings correct as far as I can see.

    The 208 watt rating is one to be careful of. It is the rating with the transistor on a infinite heatsink. In reality, you should not design to have anything like this being dissipated in your transistor unless you have an excellent heatsink.

    The good news, though, is that the power is the current times the ON voltage of the transistor. This may be only a volt or two, but you need to check the ratings.

    If it was 2 volts and 20 amps were flowing, the power would be 40 watts dissipated in the transistor.

    For large currents, it is probably better and cheaper to place several transistors in parallel to make good use of the heatsink.

    You need to measure the actual current the motor draws on 12 volts when being used with a load. If this is more than the transistor rating, then you can place more transistors in parallel. Putting resistors in series is not a good idea as this will reduce the power of the motor.

    You will need diode protection to avoid transistor damage from switching spikes. This is a diode placed across the motor with its cathode connected to the + supply line of the motor.
  8. Feb 4, 2010 #7
    Your ideas are sound. Thank you very much for replying. I espcially like the concept of making use of two transistors. I think I will do this anyway as I do suspect some of the currents drawn might fly a bit close to the wire.

    I'll produce a drawing later of the final design, I am most grateful for your input.

    There is one final question though. Will I need quite a beefy diode to handle the death throes of the collapsing field induced voltage in the inductor and motor windings?

    Thank you sir :wink:
  9. Feb 4, 2010 #8


    User Avatar
    Science Advisor

    Adder Noir, don't use a 600v rated mosfet in a 12V circuit unless you really need it. In the design power mosfets they have to trade-off a LOT of current carrying capability in order to get a high breakdown voltage. Look at how much higher the current rating is for a similarly priced mosfet but rated at only 40 volts instead of 600V (202 Amps!)

  10. Feb 4, 2010 #9
    You are my hero! Nice spot pal cheers! I'm looking at the ID is ir rated to 202A and the Vds is rated to 40V.

    I see the PD rating is around 200W though which sounds quite low for this transistor. Is this PD rating meaning the power exhibited by the gate input to the source or is it the total power it can stand from the power circuit?

    I'm thinking that at only 12V and if its rDS is only max 0.004 ohms then I can get away with a very large current as the heating effect will be very low. Also obviously the voltage at the Source will be lower than 12V as the transistor is wired in series with the load.

    Maybe I will thus only need this 1 very good transistor which you've found! Can someone clear up the rather low sounding PD concept though before I go and finalise a design? Cheers guys :wink:
  11. Feb 4, 2010 #10
    Hi this is it all done. Pretty sure I've missed a coupling capacitor off the input side but not to worry I'll add it later.

    http://img213.imageshack.us/img213/7808/motorcontcircuitscan100.jpg [Broken]

    The motor is drawn rather simplified as if it were just direct connected to all this control circuitry, infact naturally it will have a larger cable and all that sort of thing so don't worry I know how to wire this so it wont cause the cable to catch fire. I don't understand much of the 555 circuitry I've just lifted that off another site and rigged the input into my motor set-up. I've also made R2 variable as this is what I want the user to control with a control knob.

    I am a little worried about the HIGH and LOW cycles. I assume the transistor is only turned on during the HIGH cycles. I hear that by altering the value of R2 too much and by affecting the ratio of its size in relation to R1, I can unbalance the HIGH and LOW cycle times so the are not 50%/50% . This is'nt a big deal if the transistor is only on during the HIGH cycle. The kind of control I'm looking for here is not mega precise and will be done by eye. I would like to ascertain if this is true though before going to a build.

    I'd appreciate some input if possible thanks :wink:
    Last edited by a moderator: May 4, 2017
  12. Feb 4, 2010 #11


    User Avatar
    Science Advisor

    You might like this circuit:

    http://www.dprg.org/tutorials/2005-11a/index.html [Broken]

    It gives pulse width modulation and a circuit for controlling a motor.

    A large FET will get too hot to touch if it is dissipating 10 watts without a heatsink. So, you need to understand that dissipating 200 watts requires excellent heat conduction and a very good, very large heat sink.

    Your diagram shows a bipolar transistor, but I assume you are still thinking of using a FET.

    The other thing to watch is that FETs have a lot of input capacitance. The effect of this is to make the voltage of the square wave rise slowly rather than quickly. This will make the FET run hotter than if it just switched quickly.
    So, you may find you need a driver between the 555 and the FET.
    like this:
    http://dl.dropbox.com/u/4222062/FET%20driver.JPG [Broken]

    The output of the 555 would feed in at the left side of this diagram.
    Last edited by a moderator: May 4, 2017
  13. Feb 5, 2010 #12

    Oh my that's some seriously good information alright! I think I will indeed swop my design for the one shown in that link, but I'll wire an inductor across the motor in parallel and include a drain diode for when the inductor switches off and see if that helps it out. I'll prob include a bypass switch on prototype versions to switch it out of the circuit and back in for testing reasons.

    I'm pretty much sold on the concept that 10W is even enough to get really hot (too hot). I think using the rDS value I can work out how much heating will occur within the transistor by working out how much current will be flowing through it but for that I need to get off my butt and get some values for the likely resistances of the motor windings I intend to power. I can do that tonight when I go to the model railway club I'll take my electric testing meter, should give an excellent accuracy value.

    It's all good so far! Only one problem remains. I really don't understand what a driver circuit is? How does this negate issues of capacitance in MOSFETS? One other thing is how come that while a current is rising, capacitance is causing real power loss as heating? I thought a capacitor could only demonstrate magnetic based reactive resistance which caused no heating effect?

    Seems I have so much still to learn! :wink:
  14. Feb 5, 2010 #13


    User Avatar
    Science Advisor

    Don't use the parallel inductor, it will do nothing but waste power (and build up a large free-wheeling current which will be detrimental to your circuit).
  15. Feb 5, 2010 #14
    The gate of a MOSFET has a certain amount of capacitance between it and the source terminal. If you tried to instantly switch the gate voltage from zero to say 5V, the capacitance would prevent the gate to source voltage from instantly reaching its final value (somewhere above the Gate Threshold Voltage). In other words, you need to charge this equivalent capacitor first, before your gate voltage reaches the desired voltage. The driver, being a low impedance source, deals with this issue by sourcing and sinking sufficient current to the gate. This decreases the time it takes to charge the capacitance thus increasing the switching speed.
  16. Feb 5, 2010 #15
    More good stuff guys thanks I'll reply in more detail tomorrow it's 1200am GMT here, cheers!
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook