MOSFET for mciro Current Mirror

Discussion in 'The Projects Forum' started by shalimar, Mar 16, 2006.

  1. shalimar

    Thread Starter New Member

    Mar 16, 2006
    4
    0
    I have designed a circuit for generating micro current pulse. The basic structure is current mirror. When there is a current input, there is a current output, which is the exactly mirror or multiple of the input current. For I use both NMOS and PMOS to generate sink and source current.

    The amplitude of the current is micro A and the duration of the pulse is micro sec, so I think it might be some special requests for the discrete MOSFET. Could U give me some key parameters and types of discrete MOSFET for using in micro current design?

    Thanks and looking forward:)
     
  2. Papabravo

    Expert

    Feb 24, 2006
    9,900
    1,723
    I'm trying to visualize what you are describing, but I just can't quite get there.
    Could you post a schematic diagram?

    When you say micro current are we talking:

    microamps = 10^-6
    nanoamps = 10^-9
    picoamps = 10^-12
    femtoamps = 10^-15

    You might want to start by examining the datasheet for the 2N7002. In the absence of more details I don't know what else to tell you.
     
  3. windoze killa

    AAC Fanatic!

    Feb 23, 2006
    605
    24
    He did mention micro A. I assume he means uA.

    And for trivia..... what is 10^-26
     
  4. Papabravo

    Expert

    Feb 24, 2006
    9,900
    1,723
    "the amplitude of the current is micro A"
    If that is singular then we have one(1) microampere.
    If it is plural then we could have anything from 1 microampere to 1 milliampere.

    His question might be..."Can you build a current mirror that works in the the 1 microamp range with 1 microsecond pulses"

    But that is only rank speculation.

    atoamperes = 10^-18

    don't know about 10^-26 but I bet you do!
     
  5. bigbigblue

    Active Member

    Mar 15, 2006
    38
    0
  6. windoze killa

    AAC Fanatic!

    Feb 23, 2006
    605
    24
    oooppppssss meant 24 not 26. Fat fingers.
     
  7. shalimar

    Thread Starter New Member

    Mar 16, 2006
    4
    0
    I'm sorry that I have made some misunderstanding. And What I want to describe is that, I want to implement a current mirror. It could mirror the input current at the output stage. The basic current mirror consists of 2 matched MOSFETs.

    The requests of the current mirror is that the input current pulse will have an amplitude of 1 uA - 1 mA (1 uA = 10^-6A, 1 mA = 10^-3 A) and the pulse duration is 20 us - 400 us (1 us = 10^-6 s). So I think normal MOSFETs might not be able to generate this micro current and there should be some special requests for these MOSFETs.

    I have drawn the current mirror circuit in Cadence Capture. In order to stimulate the circuit in PSPICE, I selected the MOSFETs randomly (The attachment is the circuit I have designed).

    I used the enhanced-MOSFETs, PMOS for generating the source current and NMOS for the sink current. The first step is to use discrete MOSFETs for this circuit. Could you give me some info about what kind of MOSFET could be used in the micro current mirror and what the special requests they should be met. Thank you: )
     
  8. shalimar

    Thread Starter New Member

    Mar 16, 2006
    4
    0
    And I have forgotten the frequency of the input current, range from 0 - 350 Hz. Thank you again and looking forward.
     
  9. Papabravo

    Expert

    Feb 24, 2006
    9,900
    1,723
    A quick review of the 2N7000/2N7002 datasheet reveals that your drain current ranges of 1uA to 1mA occurs right in the the neighborhood of Vgs(th). The value of Vgs(th) is 0.8 Volts to 2.1 Volts. As I understand MOSFETS it will be difficult to get reliable and predictable operation for your range of currents in the neighborhood of the threshold.

    I've looked at a few other discrete enhancement mode MOSFETS and I don't think there is anything out there designed to operate at less than 250 uA in the neighborhood of Vgs(th) = 0.6 to 1.5 volts.

    I could be wrong since I'm no expert at IC design but as the geometries get smaller the ability to check circuits with discrete components vanishes in a puff of magic smoke. Sorry about that.
     
  10. shalimar

    Thread Starter New Member

    Mar 16, 2006
    4
    0
    Thank you. And now I know what direction I should go.

    I think I might try some MOSFETs with low current in the neighborhood of Vgs(th).

    If the inherent current is no less than 250 μA, I will test input current around 1 mA.

    And at same time, I will try to find the ideal MOSFET.

    Thank you : ) You really help me with the special requests of the MOSFET.
     
Loading...