7
\$\begingroup\$

Description

I am designing a system in which there's a circuit that controls the power applied to a micro SD card (enable/disable).

The circuit is the following:

enter image description here

The power control circuit is done by a P-MOSFET that is deactivated by default.

MICROSD_PWR_EN signal is connected to a pin of a microcontroller configured as open-drain.

Issue

The measured voltage on the pin VDD of the micro SD card should be 0V by default. However, this voltage is near +1V, which is neither a logic "0" nor a logic "1". The voltage measured on node "+3.3V" is +3.288V and the one measured on the gate of Q5 P-MOSFET is +3.285V.

Do you have any idea regarding this issue?

Could that be related to the 3mV difference between the source and the gate of the transistor?

Firmware solution

Firstly, thank you all for your answers.

It seems that I solved the problem by firmware: by configuring the SD card GPIOs as output open-drain and by setting them to logic "0", the voltage on the VDD pin of the SD card is now near 0V.

As everyone pointed, it is probably related to protection diodes of the SD card chip GPIOs.

\$\endgroup\$
3
  • \$\begingroup\$ I wonder if it's being self-powered to some degree by the pull-up resistors? Perhaps try removing them temporarily if that's easy enough and if that's the case connect them after the FET. \$\endgroup\$
    – PeterJ
    Sep 24, 2013 at 11:55
  • \$\begingroup\$ Google 'protection diodes'. But I think switching power like this can lead to data loss. What if you switch power while the card is writing data internally? Can you just manipulate the _DETECT line instead? \$\endgroup\$
    – markrages
    Sep 24, 2013 at 15:50
  • \$\begingroup\$ @markrages In my system, there's only read accesses to SD card. So, there's no case in which data can be lost. The reason to use this power control is to make sure at the start of the system that the card is in a correct state by forcing a power-off then power-on. I have seen SD cards that were in a unknown state after system startup and I want to avoid this. \$\endgroup\$
    – johsey
    Sep 25, 2013 at 7:17

2 Answers 2

Reset to default
6
\$\begingroup\$

The residual +1V that you are seeing is being caused by bias on the other signal pins to the uSD card. Current passes from either high levels on the microcontroller I/O pins connected at the SDIO interface or via the 47K resistors that you have on these lines into the controller chip in the uSD card. From there it passes through the input protection network on these pins to the uSD card VDD pin where you see it appear at the 1V level.

You can correct this situation by following the steps below:

1) Connect the supply line side of the pullup resistors to the switched VDD of the uSD card.

2) Whenever the microcontroller firmware goes to set the MICROSD_PWR_EN signal high to disable the card power set all output pins of the SDIO interface to a low level by outputting 0 bit values to their port register bits. Note that in some cases where the SDIO is enabled to an dedicated on-board peripheral on the microcontroller it may be necessary to set these output pins back to GPIO mode to allow the FW to gain control of the pins.

3) For any signals that are inputs to the microcontroller from the SDIO interface you need to arrange for these to go to a low level whenever the microcontroller sets the MICROSD_PWR_EN signal high. This can be done by one of two ways. You could change the 47K pullup resistor to a pulldown on these specific lines. Otherwise the input pins could be programmed back to GPIO mode and then set as outputs at a low level. This latter may be easier since then the I/O pins get handled the same as the output pins.

At the time the microcontroller goes to re-enable the uSD power by setting MICROSD_PWR_EN signal low the firmware would be written to re-configure all the SDIO interface pins back to their normal operating mode.

\$\endgroup\$
0
\$\begingroup\$

It's quite likely that what you are measuring is the leakage current from the pic lines and pull-up resistors producing a small voltage on the other pins and that due to internal diodes in the SD card this is leaking through to the power pin.

You are probably measuring that voltage with a high impedance voltmeter. Try putting 100k ohms from the Vdd pin to ground - does it reduce this voltage you measure?

Leakage thru the FET is also likely to be significant with these high impedances.

\$\endgroup\$
1
  • \$\begingroup\$ I've put on 100K resistor between the VDD pin of the SD card and the ground, the measured voltage is greater than before (+2.45V). \$\endgroup\$
    – johsey
    Sep 24, 2013 at 15:37

Your Answer

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

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