nRF52832: Power OFF/Reset with UART connected

RE: nRF52832: Power OFF/Reset with UART connected

Ole Bauck — Fri, 03 Feb 2017 15:06:41 GMT

You will probably not damage the chip, but as you are out of spec we cannot promise anything.

Input buffer is, well the input buffer, used to read the state of the pin. If the input buffer is disconnected you cannot read the state of the pin from the IN register. This has nothing to do with the protection diodes, those are on all the gpio pins regardless of their configuration. The gpio is configured as input with input buffer disconnected which means that they do nothing (don't have any function). Configure them as output by default might give unwanted behavior (like shorts). Pull up is normally not used, so this should be off, it might also lead to leakage current (pin is low, pull up resistor is on).

RE: nRF52832: Power OFF/Reset with UART connected

Marco — Wed, 18 Jan 2017 21:18:24 GMT

Another question, by default at reset GPIO pins are configured to be INPUT with NOPULL and DISCONNECT INPUT BUFFER. What does actually "disconnect input buffer" mean? Wouldn't it be safer to have the GPIO pins configured as output or input with pull-up by default?

RE: nRF52832: Power OFF/Reset with UART connected

Marco — Wed, 18 Jan 2017 21:09:48 GMT

Thank you Ole for the suggestion, however I cannot modify my actual board (just used for prototyping). Could my actual setting (UART Rx/Tx pins held at 3.3V leaking current throughout the nRF chip) cause any HW damages to the nRF itself? For instance, the nRF specification states that Pin in/out voltage should fall in the range -0.3V to Vdd+0.3V=1.87V. But this specification is not met when the main power source is removed. In such case pin connected to UART lines get 3.3V and Vdd happens to be 1.57V, thus (1.57+0.3)V=1.87V < 3.3V! Isn't it causing a dangerous latch-up condition?

RE: nRF52832: Power OFF/Reset with UART connected

Ole Bauck — Wed, 18 Jan 2017 12:09:22 GMT

There is no way to completely disconnect the pins. The only solution is to connect a series resistance. This will limit the amount of current that can flow (the voltage drop will be proportional to the current).

RE: nRF52832: Power OFF/Reset with UART connected

Marco — Tue, 17 Jan 2017 20:30:00 GMT

Yes, in my previous message I was stressing on the pins (rather than on the UART). As you said the problem is not the UART itself, but the rx/tx pins connected to the nRF and providing 3.3V even when the nRF is unpowered. I was wondering if there is any way (e.g. a nRF internal switch) that would allow me to force the nRF to keep the 2 GPIO pins (connected to the external UART RX and TX lines) "disconnected"/floating from the nRF internal circuitry. Thank you.

RE: nRF52832: Power OFF/Reset with UART connected

Ole Bauck — Tue, 17 Jan 2017 11:13:40 GMT

You can add a series resistance of about 10K to limit the current.

RE: nRF52832: Power OFF/Reset with UART connected

Ole Bauck — Tue, 17 Jan 2017 11:06:38 GMT

The problem is not with the UART, the problem is that there is voltage on one of the pins of the nRF when it is powered off. You will have to remove the voltage source when the nRF is powered off to remove the problem.

RE: nRF52832: Power OFF/Reset with UART connected

Marco — Tue, 17 Jan 2017 04:04:05 GMT

Thank you Ole for your support. Your help together with this link are the answer: devzone.nordicsemi.com/.../ It turned out that my sensor was not resetting according to its specifications. This was happening because the sensor did not get powered off properly (1.57V in due to UART pins). Is there any way I could force the UART pins to "disconnect" so that any voltage on UART won't affect the system?

RE: nRF52832: Power OFF/Reset with UART connected

Marco — Mon, 16 Jan 2017 21:21:01 GMT

As a test I'm just running an UART printing loop. When I power off the chip, messages stop been printed out via UART. When I power up nRF again, it starts again from the bootloader. UART Tx and Rx are connected to pins P0.30 and P0.31, and they feed the nRF with 3.3V. The nRF itself then keeps Vdd to 1.57V. 1.57V is within the absolute maximum voltage supply rating for nRF52 (-0.3V to 3.3V), however it is smaller than the minimum recommended supply voltage (1.7V). This might explain why the firmware stops running. In addition the nRF specification states as well that Pin in/out voltage should fall in the range -0.3V to Vdd+0.3V=1.87V, and this is not the case since nRF gets 3.3V in. I know I am looking at it in the opposite way, since power supply is not coming from Vdd directly, so I am not sure how such specification applies in this case.

RE: nRF52832: Power OFF/Reset with UART connected

Ole Bauck — Mon, 16 Jan 2017 14:22:57 GMT

The whole chip will stay powered if the voltage on a pin is supplying the chip. When you say that your firmware is not running, may it be because the code has entered an error condition and is stuck in the error handler?

You can of course measure Vdd when the chip is powered off and the UART connector is still connected, to see if what I described in my last comment is the case.

RE: nRF52832: Power OFF/Reset with UART connected

Marco — Fri, 13 Jan 2017 20:11:56 GMT

Do you mean that some power-blocks of the nRF stay powered up, while others switch off? Because for example my firmware stops running (even if UART is still connected) and starts again only when I power the nRF up again.

RE: nRF52832: Power OFF/Reset with UART connected

Ole Bauck — Fri, 13 Jan 2017 11:51:47 GMT

If the RX/TX pins are 3.3V while powering off the nRF, the nRF will actually not be powered off because of ESD protection diodes on the pins, see for example this picture. When the voltage on the pin is higher than the Vdd plus the diode forward voltage drop, the diode will start to conduct current and the new Vdd will be the voltage on the input pin minus the forward voltage drop over the diode.

I don't know why this is messing up your sensor SPI readings. You should connect a logic analyzer to see if the data transferred is correct.