0
\$\begingroup\$

I need to control the 'IN' (enable) pin of this LDO Voltage regulator (ADP162) with a microcontroller (MSP430G2553). In the datasheet of the LDO, I couldn't find whether or not it has an internal pull-down resistance.

Is it necessary to put an external one or can you directly connect the GPIO pin of the MSP430 and the EN pin? I really use two LDOs: one to power one LCD and the other LDO to power one RF module. In the data sheet, in figure 39 you can see that the start-up time is almost 1 second, so I think it would give plenty of time for the microcontroller to set the appropriate value on the GPIO pin that controls the EN pin of the LDO.

Therefore, I think it is not necessary to use a pull-down. It is right?

\$\endgroup\$

2 Answers 2

1
\$\begingroup\$

If cost is not an issue, and power consumption not critical, I would put a 10K or 100K pull-down on that line. It will enforce state before the MSP VCC is high enough to guarantee state, and while / if the MSP is in reset, it will guarantee state as well. Invert this to a pull-up if the desired state during reset / boot-up is for the device to be "on".

Without this pull-down (or pull-up), you might see odd 'Heisenbug' behavior in the lab -- sometimes the local humidity and temperature conditions can cause unexplained inconsistent behaviors. This can also commonly happen on SPI or JTAG lines where a pull-up is omitted.

If you aren't building a million of these, the cost of adding the resistor is low (layout impact, assembly cost impact, the actual cost of resistor is negligible). If it's not a low-power wearable type application, a 10K resistor at 3.3V will cause you to burn an extra 330uA of current; 33uA if you use 100K.

In general, I would suggest that for I/O lines that control "actuators" / state / other items, something should enforce the state on that line prior to the system coming online and taking active control. With a properly partitioned system design, you could spend the pull-up/pull-down only where needed (i.e., a LDO control line), and leave other I/O floating, since those I/O have no effect until the LDO is powered on.

\$\endgroup\$
3
  • \$\begingroup\$ Consumption is critical in this design. I think the pull-down causes some additional consumption when the MCU forces a high level on the pin, is it correct? But this would be once a day for about 30 seconds. In this case, is it better not to use the external pull-down? Thank you. \$\endgroup\$ Commented Aug 30, 2017 at 16:53
  • \$\begingroup\$ I think I edited my answer just as you asked that :) -- I think a 100K PD is likely enough, and that would burn an extra VCC/100K amperes while the pin is in the high state. You could play with even higher values of a PD to see what works if the consumption is very critical. \$\endgroup\$ Commented Aug 30, 2017 at 16:54
  • \$\begingroup\$ Yes, my comment and your message edition have been crossed :) I think that an additional consumption of 33 uA per LDO and day for 30 seconds is acceptable. So I think that a 100K can be a good choice. Thanks. \$\endgroup\$ Commented Aug 30, 2017 at 17:18
1
\$\begingroup\$

In the datasheet of the LDO, I couldn't find whether or not it has an internal pull-down resistance.

The datasheet says you must "drive" the EN pin low to turn off the regulator or "drive" the pin high to turn on the regulator.

enter image description here

The "EN input leakage current" spec at the bottom of page 3 of the datasheet also indicates equal leakage current limits for high or low inputs on EN.

These two specs indicate there is no internal pull up or pull down on the EN pin, and you must provide one if your circuit requires it.

In the data sheet, in figure 39 you can see that the start-up time is almost 1 second, so I think it would give plenty of time for the microcontroller to set the appropriate value on the GPIO pin that controls the EN pin of the LDO.

When I look at that figure, I see a turn-on time of 1000-2000 us, which would be one or two milliseconds, not one second.

But this figure shows the response when the EN pin is driven high, presumably assuming the input voltage is already stable. The timing if the input power ramps up with the EN pin left floating is not indicated.

If you just want to turn the regulator on soon after the input power becomes available, you can just connect EN to VIN and not worry about the microcontroller at all.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks for showing me how to tell if a pin has an internal pull-up or pull-down. You're right, it's 1000 or 2000 us;) In this design I need to activate the LDO in very few occasions and as the consumption is critical, I need to control the LDO with the microcontroller, so I can not connect the EN pin directly to VCC. \$\endgroup\$ Commented Aug 30, 2017 at 17:46

Your Answer

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

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