MOSFET driver for controlling common cathode LEDs

[Remark]

Hello,

I have encountered a problem controlling 24V LEDs using a transistor. According to the common cathode LEDs configuration, I finished my design with N-channel and P-channel MOSFET drivers for high-side switching, where the MOSFET is turned on with a low signal and turned off with a high signal. However, when simulating the schematics, I noticed some voltage across the simulated 'LED' when the 'driver' is switched off. Could I ask where this voltage comes from and why it isn't zero or near zero when switched off?

[tooki]

Ummm… take a closer look at the “some voltage”: it’s 5.8ᴇ-5.

You do know how to read scientific notation, right?

That means 5.8 x 10^-5V = 5.8/100000 V = 0.000058 V = 58 microvolts.

That’s very much “near zero” in the context of 24V.

[PGPG]

I have encountered a problem controlling 24V LEDs using a transistor.

I think you have bigger problem in your design than 58μV at LED when switched off.
Have you read Absolute Maximum Ratings section of IRF9540NPBF datasheet?

[MrAl]

Hello,

I have encountered a problem controlling 24V LEDs using a transistor. According to the common cathode LEDs configuration, I finished my design with N-channel and P-channel MOSFET drivers for high-side switching, where the MOSFET is turned on with a low signal and turned off with a high signal. However, when simulating the schematics, I noticed some voltage across the simulated 'LED' when the 'driver' is switched off. Could I ask where this voltage comes from and why it isn't zero or near zero when switched off?

Hi,

What current does each LED draw and how are they wired, in series parallel, and if so how exactly?
The IRF9540 seems to have a pretty high current rating but you also have to watch out for power dissipation.

For your original question, LEDs do not draw much current when the voltage is below the normal operating voltage.  That means the resistance of the LEDs may be high until they start to light up.  The mosfet will have leakage current, and that will 'power' the LEDs, but they probably will not light up because this current would be far too small in most cases.  So nothing to worry about.  The leakage current could be from 2.5ua to 250ua, which is still not that much but could light up one or more LEDs slightly.  That's at max temperature though, which would only occur if you were using the transistor at the maximum current with minimal heat sink.

Is that voltage of 5e-6 volts correct?  If so, that's nothing to worry about.  It's from the leakage current of the mosfet and/or part of the internal voltage of either a diode or the LEDs (the models sometimes use small biases they may refer to as the band gap voltage although that may not really be the reason for every model).