Input Power Conditioning (Clamp)

Discussion in 'The Projects Forum' started by ChateauduChillon, Dec 18, 2014.

  1. ChateauduChillon

    Thread Starter New Member

    Jan 28, 2014
    24
    0
    Hi all,

    I need system to condition an incoming Vcc line to deliver ~250mA current when Vin > 7V, with the following voltage transfer characteristics:

    [​IMG]

    for 0V < Vin < 7V : Vout < or = Vin

    for 7V < Vin < 8.5V : Vout = Vin (say within 20mV)

    for 8.5V < Vin < (10-14)V : Vin < Vout < 14V

    for (10-14) < Vin < 28 V : Vout = 12V



    First simple design I sketched of doesn't seem viable:

    [​IMG]

    Second design may work, but seems more hokey than necessary and I'm not sure how to look for a correct switch as I'm not too familiar with power switches:

    [​IMG]

    Here I have the Vcc incoming line switched vs the output of a 28V capable switching regulator set to output a fixed 12V, starting up via some startup circuitry e.g. res divider to EN pin.

    Switch control comes from a 28V capable comparator which should trip at some point (e.g. 10-14V is OK) which should be accomplished by choosing R1, R2, R3, and Vz appropriately.

    Thoughts on the second design? Switch recommendations? I am somewhat space constrained. Below are my thoughts. Am I missing a simple solution?

    [​IMG]

    Thanks,

    Chateau
     
  2. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    If I read it right you want Vout to equal Vin up to 12 volts and then Vout to remain at 12 volts. Is that correct? At what voltage do you need to deliver 250 ma,? All of them or just at 12 volts.
     
  3. ChateauduChillon

    Thread Starter New Member

    Jan 28, 2014
    24
    0
    That would be a sort of the ideal case.

    We would need to deliver 250mA anytime Vin > 7V.
     
  4. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    Usually voltage regulators are used for this. But Vout will be a few volts lower than Vin until it reaches 12 volts. Would that work?
    Maybe tell us what you are trying to do instead of specing it.
     
  5. ChateauduChillon

    Thread Starter New Member

    Jan 28, 2014
    24
    0
    Unfortunately it wouldn't :( I am just interfacing with an existing system and these are the specs I need to hit. Do you have any feedback on my proposed design?
     
  6. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    OK. Well that kind of forces you to the shunt regulator like your first picture. The problem is it wastes a lot of power so it is big.
    Your second picture Vout does not track Vin, it just turns on when Vin reaches 12 volts.
    Here is a sim of what I was thinking.
    Let me think if there is an easy way around the problem.
    Is there always a fixed load?
     
  7. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    The LT1084 will deliver almost a volt less than the input voltage. O.P. requested tracking to 0.02 volts.
    This circuit will do that job, but it also illustrates how difficult the request is.
     
  8. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    Yes, so it needs some sort of switching arrangement.
    It looks like below 7 volts is kind of a don't care. Above 7 volts Vout = Vin so a fet switch feeding a shunt regulator might get us up to say 12.5 volts, then switch to a very low drop out regulator and turn the shunt off. Pretty complicated, but maybe possible.
     
  9. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    Don't forget, post #1, "I am somewhat space constrained".
    I want 0.16% precision, multiple switch points, and it must be rather small.
     
  10. ChateauduChillon

    Thread Starter New Member

    Jan 28, 2014
    24
    0

    Sorry I meant 2nd design, 3rd picture. I am proposing switching between Vin and a regulated version of Vin, Vout will definitely track Vin until the switch connects Vout to the regulated output.
    Load is essentially static yes.


    Right, I did not suggest LDO as a design option because I know that an LDO is not going to work. Hence, my more complicated suggestion in the 3rd picture.

    This is similar to what my third picture is proposing.


    Why couldn't the circuit in my 3rd picture accomplish this is I guess what I'm asking.
     
  11. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    Ahh, I see there is a second supply.
    Something like this?
    It is off by about 50mv right around 12 volts because the knee of the reference is not perfect.
     
Loading...