Protection diodes on NRF52832: max current

RE: Protection diodes on NRF52832: max current

Bwalker — Mon, 28 May 2018 14:16:41 GMT

Hi Hakon,

In this case, VCC is actually grounded, so the only real issue will be the diode current.

Thanks,

Ben

RE: Protection diodes on NRF52832: max current

haakonsh — Mon, 28 May 2018 10:45:33 GMT

Just to be clear, the maximum allowed GPIO voltage is VDD + 0.3 V. If VDD is at 0V the maximum allowed input voltage is 0.3V. If VDD is at 3.9V the maximum allowed input voltage is 3.9V. See Absolute maximum ratings.

The reason I cannot say that 10µA is okay is because I honestly do not know. This is outside our tested and verified specification.

If VDD is floating, a 5V input on a GPIO will result in the ESD diode conducting a current into the VDD net, thereby powering anything in that net at 5V - 0.3V = 4.7V, including the nRF52832 who will consequently break. You can try to limit the current going into the GPIO, but we cannot make any guarantees that the nRF52832 will perform to spec.

RE: Protection diodes on NRF52832: max current

Bwalker — Fri, 25 May 2018 15:15:29 GMT

Yes, the VDD is effectively grounded (low impedance) in this scenario.

RE: Protection diodes on NRF52832: max current

AmbystomaLabs — Fri, 25 May 2018 15:10:16 GMT

The ESD diodes really don't apply for this scenario. Unless you mean to imply in your schematic that nRF Vdd is actually grounded?

Normally people refer to a situation where nRF Vdd is unconnected or high impedance to the supply, then the only current flow will be due to the top FET in the cmos output stage being reverse biased due to no Vdd and will short your external voltage source through to nRF Vdd. The effect will be as if your external source is powering the nRF.

RE: Protection diodes on NRF52832: max current

Bwalker — Fri, 25 May 2018 14:55:53 GMT

Hi Hakon,

Do you think 10uA would be ok? 1uA is slightly too low for what I'm hoping to do.

Thanks,

Ben

RE: Protection diodes on NRF52832: max current

haakonsh — Fri, 25 May 2018 14:35:56 GMT

Hey Bwalker,

We don't have a defined spec, but 1mA on ESD diodes is way too much, 1µA should be fine.

Cheers,

Håkon.