I have come across a gate driver circuit (proprietary) that measures open-collector current and sets a flag if the current exceeds a threshold (say 20 mA) to protect the NPN transistor.

This is the "schematic" from the datasheet. How would such circuit be implemented to be inexpensive and reliable?

Edit: My question is how can this open-collector circuit be modified to also measure the npn transistor current? The basic open-collector functionality is obvious. I'm interested in how to protect this open-collector circuit from excessive currents caused by too small value of R_pullup.

enter image description here

  • 1
    \$\begingroup\$ Seems like from your schematic above that the NPN transistor here is simply an open collector output that is used to report an error condition detected in another part of the gate driver chip. Using such output is easy because you connect the pullup resistor to the V+ rail of your microcontroller and the open collector error signal into an input of the MCU. \$\endgroup\$ – Michael Karas Dec 16 '15 at 17:00
  • 1
    \$\begingroup\$ This Data Sheet is what OP is referring to, page 23 . Semikron \$\endgroup\$ – Marla Dec 16 '15 at 17:02

The circuit you refer to Data Sheet (page 23, is merely an output signal to a user that an error has been detected. The detection circuitry is another matter.

The manufacturer has spent a lot of time and money in developing their detection circuit.

"How would such circuit be implemented to be inexpensive and reliable?"

Achieving Reliability is a broad subject involving both mechanical and electrical design.

The manufacturer states, "The error output of SKiiP® 3 V3 is short circuit proof.". Using an open collector means you can short the open collector to Ground with no ill effects.

  • \$\begingroup\$ Also the claim of short-circuit proofness is somewhat misleading (doesn't say short-circuit to what) but it cannot be shorted to Vs; needs an appropriate pull-up resistor to limit current to at most 15mA. \$\endgroup\$ – Fizz Dec 16 '15 at 18:45
  • \$\begingroup\$ @RespawnedFluff : Very true. Spehro's answer had that covered by the time I thought to edit it in to my answer. \$\endgroup\$ – Marla Dec 16 '15 at 18:48

Here is an example of a short-circuit proof open-collector output:


simulate this circuit – Schematic created using CircuitLab

The transistor types would have to be modified to suit the maximum voltage at the output and the current- since Q1 will dissipate maybe 0.75W with a short to +30V.

  • \$\begingroup\$ Perhaps it is obvious but could you point out how this circuit mod could be used to measure the open-collector current draw? \$\endgroup\$ – SunnyBoyNY Dec 16 '15 at 18:55
  • \$\begingroup\$ @SunnyBoyNY: As Q2 begins to open it will limit the current through R2 to about a Vbe/33ohms =~ 0.6V /33 ohms = 18mA. Q2 does does that by "stealing" Q1's base current. \$\endgroup\$ – Fizz Dec 16 '15 at 19:08
  • \$\begingroup\$ So are you saying that this circuit is not really suitable to measure the open collector current but to detect when the current goes above a certain threshold? \$\endgroup\$ – SunnyBoyNY Dec 16 '15 at 20:11
  • \$\begingroup\$ How could this circuit be modified to not dissipate that much power? Say to detect a 20 mA condition and based on that disable the output? Say the power rating is capped at 100 mW. \$\endgroup\$ – SunnyBoyNY Dec 16 '15 at 20:16
  • \$\begingroup\$ You could measure the emitter resistor voltage with an ADC input on your micro and go into a 1:10 duty cycle mode if an overcurrent condition is detected. That way the sc protect condition would be cleared in a definite time after removal of the short. I would not use a latching circuit in an industrial environment. \$\endgroup\$ – Spehro Pefhany Dec 16 '15 at 21:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.