Turning off a LED driver (opamp based linear regulator)

[Flyback]

Hello,

We have a LED driver as in the attached schematic. We sometimes wish to shut it off by turning on the PNP and putting the opamp’s rail voltage on to its inverting input.

The opamp used is the MIC6211. The positive rail is obviously outside of its input common mode range. Some of the old opamps would malfunction if such a high voltage was placed on to one of their input pins. Do you think the MIC6211 will be OK with this?

MIC6211 datasheet:
https://ww1.microchip.com/downloads/en/DeviceDoc/mic6211.pdf

 

[OBW0549]

The opamp used is the MIC6211. The positive rail is obviously outside of its input common mode range. Some of the old opamps would malfunction if such a high voltage was placed on to one of their input pins. Do you think the MIC6211 will be OK with this?

There's really no way to predict what it will do. Why not just insert a 10 kΩ resistor in series with Q1's collector? That will shut off the LEDs by driving the op amp's output low, but not raise the op amp's inverting input out of the common mode range.

 

[AnalogKid]

Delete R6, R7, Q1.
Increase R2 to 1K.
Connect M2 drain to M1 gate (in parallel with R3).

Now, the shutdown command from the uC will shunt the gate voltage to GND and turn off M1.

Delete R9 - not needed.

ak

 

[ronsimpson]

 

[Flyback]

Thanks, the opamp rail may not be up sometimes, and i dont want to put a voltage on one of its pins when/if it has no supply, as this will mean current running through its ESD diode of its input pin. As such, we used the PNP which takes from the rail itself.

 

[rjenkinsgb]

How about just adding eg. a 10K resistor in series with the diode?

That prevents excess current while still overriding the feedback through R5 and shutting down the output.

If you want belt & braces protection, add a schottky diode from the opamp input to its positive power supply; that will prevent more than ~0.4V difference which should be safe for any normal silicon device, it's less that the standard conduction voltage.

 

[ChrisP58]

Or use a low Vgs N-channel mosfet as a source follower in place of the diode in ronsimpsons drawing. Tie the drain to the op-amp supply.

 

[AnalogKid]

I think that modifying the opamp inputs to cause a beneficial overdrive at the output is an unnecessary complexity. Better to leave the opamp's input characteristics completely out of it, and short the M1 gate to GND with M2.

ak

 

[ronsimpson]

and short the M1 gate to GND with M2.

That will leave the op-amp lifting up on a 10 ohm load. (max current on the amp)
You could increase the 10 ohms.

 

[AnalogKid]

That will leave the op-amp lifting up on a 10 ohm load. (max current on the amp)
You could increase the 10 ohms.

Increase R2 to 1K.