UNO USB power switch, T1

Using Arduino IDE 1.x
1

While troubleshooting a problem that tracked the power source used, I had a look at the power switching circuit in the UNO Reference design diagram. The switch that turns off the USB power when Vin is detected, appears to be backward.

The switch, T1 is a P-channel MOSFET. To correctly bias this device, the drain should be more negative than the source. As shown in the reference diagram and as connected in the Elegoo UNO R3, the drain is connected to USBVCC and the load (the 5V power rail) to the source. Further, the diode from drain to source will always be forward biased by USB voltages greater than the 5V UNO rail. It is hard to see how T1 could turn off USB power.

The UNO design is mature and it is hard to believe that a backward T1 has escaped previous notice. However, I invite those with knowledge of MOSFET circuits to show me my error.

2

Hello,

Did you do exactly like the UNO scheme?

Did you use the LM358?

3

I didn't change any hardware. The backward T1 is backward on the UNO Reference Design Schematic and is connected backwards in the Elegoo UNO R3 I'm working with. The error, if it is an error, is likely to be on every UNO.

4

Hello,

Let me try to explain some considerations,

I think the Arduino UNO scheme is correct,

Look at IC U1A, it is a comparator, if input 3 (+) has a voltage value greater than the voltage value present at input 2 (-), then output 1 will have a low voltage value (next of 0V).

The opposite is also true if input 3 (+) has a voltage value less than the voltage value present at input 2 (-), then output 1 will have a high voltage value (close to 5V).

Note that input 3 (+) is receiving the voltage of a factor resistive divider of 2: 1 or is dividing in half. If input 2 (-) is 3.3V, this means that the VIN value must be above or below 6.6V so that input 3 (+) is above or below 3.3V.

Since the MOSFET is a P-channel, it must be with negative voltage in relation to the Source (Gate Threshold Voltage - VGS), so that it conducts the electric current between Drain and Source, then this situation occurs when the I.C. U1A is with the low output.

Note that when testing a MOSFET with a multimeter, when the gate is charged with polarizing loads above the recommended for VGS, the continuity test can show both directions from D to S as well as from S to D.

Then voltage may appear on the USBvcc line even when power is supplied by the VIN input (PWRIN - J4 connector).

Note that voltage regulators are usually used when the input voltage is greater than the output voltage, so if the input voltage is 5V, it does not make sense to use the regulator, could connect directly on the line of 5V (+5V).

When the line voltage of 5V (+5V) is 5V, and the USB port is 5V too, there will be no current circulation, as there can only be current flow if there is a potential difference.

5

Power P-channel MOSFET circuits when operating from a positive supply usually have the supply voltage on the source and the load on the drain. This makes the drain more negative with respect to the source and supplies the correct bias for a P-channel device. However, the Reference Schematic shows the supply, USBVCC, on the drain.

There is also the problem of the diode formed by the source connection to the device substrate. This places a diode across the drain and source with the anode connected to the drain. This is clearly shown in the T1 device symbol in the schematic. Imagine what happens when the MOSFET channel stops conducting because
the op-amp comparator senses Vin. Because USBVCC is on the drain, it is still connected through the forward biased diode.

If USBVCC was on the source and the drain was connected to the +5V rail, then the gate voltage would be 5 volts more negative than the source when the comparator was low and the MOSFET would conduct. When the comparator was high then both gate and source would be near 5 V and the voltage from gate to source would be near 0 V and T1 would be off. Also, the source to drain diode would be reverse biased. I believe that this is how the circuit should work.

6

Hello,

Well, instead of imagining, maybe you can use a variable source and test it in practice :wink:

7

Thanks to those who have responded to my post. The discussion has not gone as I expected but here is one last try. I believe that the UNO T1 was intended to be a high-side switch that disconnected USB power when Vin more than 6.6V was detected. However, T1 is not connected as I would expect a p-channel MOSFET to be. It looks to me like the source and drain are reversed. For clarity please see the attached drawing.

For more information on p-channel MOSFET high-side switches please see the following:

ArdPowSw.jpg
ArdPowSw.jpg1882×654 141 KB