mosfet load

Discussion in 'The Projects Forum' started by yourownfree, Aug 7, 2008.

  1. yourownfree

    Thread Starter Active Member

    Jul 16, 2008
    89
    0
    I have noticed that the load side for an N mosfet is connected to the drain side. Is there any reason why I cannot connect to the source side instead?
    Can I just use a diode for protection for spikes.
     
  2. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,347
  3. yourownfree

    Thread Starter Active Member

    Jul 16, 2008
    89
    0
    thanks, looked it over. it looks like it could be a possibility, as long as I'm not using an inductive load big enough to spike the the barrier. The mosfet i have is +/-20 volts gate to source. i will be using 15 volt max at the drain. The information said even a reference diode cant stop it from breaking down, but can help. Talking about gate to source. It would be ify i guess if it would work. I wonder if i could put a 12 volt regulator ic there to clamp it if it tries to go over 12 volts. I dont know. The idea is, to use one wire instead of two.
    If you have two wires, you have more inductance and resistance and a loss of power to the load, since the wire run could be several feet. That is my reason for trying to use the mosfets i already have the n type and hook the load to the source instead of drain. One side frame ground.
     
  4. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,699
    905
    Please draw the circuit you are suggesting. John
     
  5. yourownfree

    Thread Starter Active Member

    Jul 16, 2008
    89
    0
    Its just a basic switch
    here is the basic circuit with no extra goodies
    mosfet is HUF75639S3R4851
    I have three in parallel on a heat sink 2 x 3 x 1(inches). the 1 is the fin length.
    I do expect 20 amps current flow, but will limit it.
     
  6. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,699
    905
    OK, now look at your circuit with 20 A flowing across a 0.6 resistor. That's 12 V. So, the source will be at 12 V relative to ground. How are you going to get the gate to the 10V relative to source (i.e., 22V relative to ground) to fully turn on the mosfet?

    John
     
  7. yourownfree

    Thread Starter Active Member

    Jul 16, 2008
    89
    0
    Ok you have a point there. I am new to mosfets. But i can see now it is similar to transistors where the base voltage is relative to the emitter. Funny how i can look all day at the information and get nothing and i get one feedback in the forum and learn more.Your making me work for this one. I saw where 10 volts was the usual full turn on. Then is the threshold where it starts to turn on? Ok how about a voltage doubler at the input and a transistor as a switch frequency input 1 kHz. Maybe just using the drain side would be easier than all this. oh here is the new diagram i just drew
     
  8. nanovate

    Distinguished Member

    May 7, 2007
    665
    1
    Why not use a P-CH MOSFET?
     
  9. yourownfree

    Thread Starter Active Member

    Jul 16, 2008
    89
    0
    That would be too easy! I'd have to go and buy some. I already have N mosfets to work with, and was just curious if I could use them instead. I cant just go to Radio Shack, i have to buy online. Radio Shack is a joke. The people there dont know squat about electronics at least the ones I have seen. Besides they dont carry any parts a hobbyists can use. Jokers.
     
  10. thingmaker3

    Retired Moderator

    May 16, 2005
    5,072
    6
    What is needed here is called a "high-side driver."
     
  11. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,699
    905
    And by analogy to your suggestion in post #7, the charge pump in the high-side driver can be considered the voltage doubler. Is this going to be PWM or straight DC on/off with long (e.g., several second) cycle times?

    Most important, is your goal to make something that works or just to experiment with high-side mosfets? From your post #5, it sounds like the former. If that is the case, put the mosfet low side and be done with it.

    Alternatively, you might want to consider the International Rectifier IPS5551T, which is a three-terminal, N-channel mosfet device with integrated high-side driver. It looks just like a mosfet, has 6 mOhm Ron, but the "gate" pin is referenced to the high side instead of to ground.

    John
     
  12. yourownfree

    Thread Starter Active Member

    Jul 16, 2008
    89
    0
    Thanks, never heard of that phrase before now. Looked it up and checked it out. Made for floating source, interesting. Almost the same diagram i drew up.
    Thanks for all your replies I think i can make it now.
     
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    If you just got a VNH2SP30-E, you'd eliminate a lot of fiddling around with discrete components.
    http://parts.digikey.com/1/parts/577017-ic-driver-motor-h-bridge-vnh2sp30-e.html
    After all, what you really want to do is get one set of plates connected to 12v while the other set of plates gets connected to ground, and then swap back and fourth, right?

    30A output in a full H-bridge. Designed for automotive motor control, but I don't see why it wouldn't work for what you're wanting to do. You'll have to make a PCB for it.

    You write a program for a PIC, a Basic Stamp, an Atmel AVR, or homebrew up something with 555 timers to drive it.
     
  14. yourownfree

    Thread Starter Active Member

    Jul 16, 2008
    89
    0
    I am behind times, thats a great idea Sgt, dont know if i can see good enough to solder all those leads, it would look like one big solder glob. Just kidding of course. Next time i design, this looks like a fun one to play with. The other pwm i was talking about, i just ended up doing it the old fashioned way. Now i have ideas to work with for another unit.
     
  15. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I'm not a disbeliever. However, I'm a skeptic.

    Show me hard proof; scientific evidence - and how you got from where you started to success.

    Let's cut through the BS. You're attempting to build a hydrolyzer. The IC I mentioned is an extremely viable candidate for hydrolyzer experiments in low frequencies.

    Be safe. If you don't include sufficient safeties in your design, you won't be able to post your results.

    Wook
     
  16. yourownfree

    Thread Starter Active Member

    Jul 16, 2008
    89
    0
    what would you like to Know or see Wook? I am building a pwm for an electrolyzer I have not completed my test yet on this unit. You of all people would easily figure out the rest of the circuit if I explained it for all to see. It is still a pwm circuit no matter how you look at it. The front end is a simple 555 astable osc. the output is using HUF75639S3R4851 mosfets. Its the stuff in between and added features I cannot share at this time. I can tell you the duty cycle will not be changed by adjusting the frequency on the 555, other than the time period of the frequency itself. therefore i can roam the frequencies and adjust duty cycle to whatever i want at that frequency. I am still adding on goodies to the unit. It has an LCD frequency readout, will have current control and limit, short circuit protection, and more. The more is for the person who wants to test their electrolyzer, but the pwm's they are able to buy are short of what they need. I am certainly not going to buy one. Spend $30 or whatever it is ,for cheapo 555 timer hooked to a mosfet. Or even an op amp lm324 style. For me i need to know what different frequencies do to the electrolyzer, voltage and duty cycle. So far I have not seen a clear answer from anyone, or any proof for that matter. So I have to do it myself to find out. As for the PWM there are only so many ways you can manipulate the waveform and output. I am just exploring one of them. One of them I have not seen anywhere on the Internet, the reason I dont care to share at the moment my thoughts on it. If it doesn't work as intended, oh well it was just an idea, i have more.
    If it does then I have an idea and a working unit to prove it, and I along with my partner in the HHO project would like to have that extra step above and beyond the rest of the hho designers and truly have a different system than the rest. I am not the genius you are i guess but I have ideas too. I hope you are not trying to belittle me. Encouragement seems more appropriate. I have read some of your posts, matter of fact you are everywhere, you live here it looks like.
    I am not scientifically right on ,when i build something that works. I do have an understanding of electronics to some degree limited as it is with most folks. I do appreciate the replies, knowledge and info I do receive.
    I hope you understand the reasons for not cutting the BS.
    Now you know the rest of the story.
     
  17. Nando37

    New Member

    Aug 15, 2008
    3
    0
    !!!!!!!!!!!!!!!!!!

    I see that NO one inform properly.

    To be able to have a load connected to the SOURCE and GROUND the circuit needs to have the GATE Voltage referenced NOT to Ground BUT to the DRAIN, this way the FET can Saturate completely.

    The problem here is to find the way to have the ckt to perform such arrangement,

    One of them is to have a square waveform oscillator and a rectifying voltage doubler (if the power supply is 12 volts), this way the FET DRAIN goes up toward positive supply and the final voltage on the gate still is about 12 volts keeping the FET fully saturate.

    Of course to turn the FET OFF fast, you need to discharge the storage capacitor fast toward ground.



    Good luck
     
  18. thingmaker3

    Retired Moderator

    May 16, 2005
    5,072
    6
    I suggest you go back and read the replies more carefully.

    In other words, a high-side driver is needed. This has already been addressed in this thread.
     
Loading...