I'm using resistors in a series circuit to vary the resistance between a constant 12V supply line and ground. enter image description here

Using transistors, the current either flow through all resistors, or is routed to ground after one of the resistors.

This creates a current signal in the tens of mA range, corresponding to which transistors are switched on or off.

Per specification, the slew rate of this signal should be between 6 mA / µs and 28 mA / µs.

To limit the slew rate, I previously used high value resistors on the base of the transistors, making them take longer to reach saturation. It has been brought to my attention that this might not be the cleanest way for the NPN transistors to operate; instead I should simply turn them on and off without such long delays.

Using an inductor in series with the resistors seems to work, but the inductors characteristic curve seems undesirable - the slew rate may exceed the maximum slew rate in the first part, and tapers off to a slew rate far below the minimum slew rate. enter image description here

Is there a way to low-pass filter a current signal which results in a more linear gradient between two current levels?

  • \$\begingroup\$ Why is the inductors characteristic curve undesirable? What is it that you are looking for specifically? \$\endgroup\$
    – Andy aka
    Commented Jun 4, 2019 at 8:06
  • \$\begingroup\$ @Andyaka Updated the question. Since the actual current steps may vary (between 7, 14 and 28 mA), sometimes the slew rate maximum is exceeded, and almost always falls short of the minimum slew rate in the later part. I'm open to completely different approaches of generating such a signal as well. I though of using PWM to step the signal, but that seems to require quite high PWM frequencies. \$\endgroup\$
    – towe
    Commented Jun 4, 2019 at 8:31


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.