cutoff voltage

Discussion in 'General Electronics Chat' started by barg, Aug 30, 2016.

  1. barg

    Thread Starter New Member

    Dec 23, 2015
    22
    0
  2. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    Try a LM339 comparator circuit
     
  3. barg

    Thread Starter New Member

    Dec 23, 2015
    22
    0
    Hi AAC,

    Can you advice why LM339 is better than tl431 in therm of reliability? and do you have any reference circuite for using with this lm339?

    Thanks,
     
  4. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    He is a little hard on the TL431 circuit.
    It needs a P channel FET.
    Since you have a 12 volt battery there is plenty of voltage to turn on the FET.
    Since the input is tied to the battery it is unlikely there will be a lot of spikes on it.
    Use a standard FET like 2N4905 and it will be well off.
    The hysterisis is needed so it doesn't chatter around the cut off voltage.
     
    cmartinez likes this.
  5. barg

    Thread Starter New Member

    Dec 23, 2015
    22
    0
    Hi Ron,

    Thanks for your suggestion! Reliability is a concern for this uv protection and i wonder if that circuit can be robust enough to be independent for long term?

    As for the Pchannel, if the vgs is max 20v and the battery is charged from a regulated but a not switched source (car alternator which are not alwaise covers all spikes that can reach up to 30v in extreme) can that affect / harm the mosfet / this tl431 circuit because it is connected to the same battery poles as the charge source by wires?

    In addition, i have a few clarification questions if possible:

    R6 (value 1 Mohm) is this one the hysterisis suppress resistor and located in parallel with the load which is R5 (15 ohm)? is R6 value will be change if the load other like 0.5A or 5A or remain 1Mohm?

    As to the value of R3 and R2 it seems that the cutoff will be at 11v, is it correct?
     
    Last edited: Aug 31, 2016
  6. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    Yes, the gate to source voltage is 20 volts. In most applications it is very difficult to make the battery voltage go very high, but maybe you should explain more about what you are doing, how big you expect the spikes to be and what produces them. Then we could design something around that.
    The circuit shown tuns off the load at about 10.5 volts and turns it back on at about 11 volts.
    The hysteresis only cares about voltage so it is the same no matter what the current.
     
  7. barg

    Thread Starter New Member

    Dec 23, 2015
    22
    0
    Hi Ron,

    its a regular charging from my jeep as last time i burned the circuite and i want this one to last. do you have any additional idea to protect the mosfet gate?

    i added a 10v zener to your design as in the attached, do you think it can protect the gate up to 30v? is how that i places it is correct?

    Thanks in advance!
     
    • ZEN.jpg
      ZEN.jpg
      File size:
      42.7 KB
      Views:
      9
  8. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    The zener and the TL431 will fight with each other when the voltage is high. You could make the zener 15 volts, but you will still need some resistance or the spikes may blow one or the other. You also probably need to worry about negative going spikes, so you need back to back zeners. But even with that the TL431 won't like minus 30 volts.
    Here is one that I think is pretty bullet proof, but of course it is more complex.
    U1 is a comparitor
    D1 clamps the maximum power supply voltage to 24 volts and minus .6 or so.
    D5 is the reference voltage for the comparator.
    D3, 4 clamp the gate to source voltage to about +/- 15 volts
    D6 blocks negative voltage from the transistor.
    Did you have your Jeep running without the battery when you blew up your last circuit?
    How sure are you about the 30 volt spikes? Could they be 50 volts?
    If you don't think they go negative we could take some parts out.
     
  9. barg

    Thread Starter New Member

    Dec 23, 2015
    22
    0
    Hi Ron,

    i have tested this circuit with LM2903M instead of LM339 as this is what i have for now, also i didnt have D1,D2,D3,D4 so they didnt participated also. the LM2903 output1 showed 0v, it doesnt sends any high to the mosfet in all situations... IN1 stayes in between 5.1v to 5.23v and IN1+ shows values from 4v to 20v (playing with the vin button of the lab power suply). I am trying to see where did i go wrong?? maybee because of missing D3 and D4 for clamping? or connection for R1? attched the exact circuite i made that doesnt works.

    Can you advice please for:

    Is D1 use for protection only? Is D2 is a regular shotky? if i think the voltage wont reach 30v, can i change D2 to 30v?

    if i think there will be no negative voltages from Vin, what needs to be eliminated?

    Is the cutoff in this circuite is at 10.5v and resume in 11v? and please advice if you had chance to test this specific circuit in the past?

    If adding a reverse voltage protection shotkky diode right after vin (as shown in the attched file), will it affect the reading of the battery by the compertor?

    Regards,
     
  10. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    It seems like you may be missing R3. The output should rise to about 0.6 volts and the transistor should turn on.

    If it never goes negitive or over 30 volts (counting the battery voltage) you can eliminate D1, D2, D6, D3 and make R1 zero ohms. Or you can use the TL431 with a couple of changes, as shown.
    The diode will make the battery voltage look about 0.6 volts lower than is really is.
     
  11. barg

    Thread Starter New Member

    Dec 23, 2015
    22
    0
    Thanks Ron,

    So can I dismiss also R1?

    As for TL431, Can you advice for R1 purpose? and if D1 is a 15v zener?
     
  12. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    Yes, replace it with a wire.

    If the voltage goes to 30 volts the 15 volt zener will turn on. Without R1 the current thru the zener and the tl431 may be very high. R1 limits this current to a safe value.
     
Loading...