Cutting VMODEM_P and VIO_P after hub.set mode:off — is it safe?

Hardware
1

We’re designing a custom carrier board for a Notecard + Swan product and need to clarify the power requirements for VMODEM_P and VIO_P. The Notecard Carrier Board Design Guide says to provide constant power to VMODEM_P and VIO_P to avoid power loss that could damage the modem, and that VMODEM_P should be directly connected to a battery or a supply capable of handling brief current spikes. Our product has a physical power button. When the user powers it down, our firmware performs an ordered shutdown: it sends a final status note, issues `hub.set mode:off`, and only then cuts the 3.3 V rail and enters STOP2. With the modem in airplane mode, we expect it to be idle at that point. We’d like to add a P-channel MOSFET to fully cut VMODEM_P and VIO_P during this shutdown, to eliminate the residual ~8 µA the Notecard draws in mode:off — relevant for units stored for weeks between jobs. Three questions: 1. Is it acceptable to cut VMODEM_P and VIO_P **after** `hub.set mode:off` has been acknowledged? Or does the “constant power” requirement apply unconditionally? 2. If it is acceptable, is there a recommended sequence between the two rails, and a minimum wait time after `mode:off` before cutting? 3. What maximum RDS(on) would you recommend for a series switch on VMODEM_P, given the 2 A peak current during transmission? For context: the carrier is powered from mains through a 5 V supply, with a single-cell LiPo backup. VIO_P would come from a separate 3.3 V rail. Thanks.

2

Hi @Antonio_Blanco and welcome to the Blues community!

Thanks for the detailed write-up. The short answer is that we don’t support this, and I’d recommend against designing it in.

Our Carrier Board Design Guide is unconditional on this point:

Notecard is designed to be powered continuously and should be left powered at all times. Provide constant power to VIO_P and VMODEM_P, and avoid power-cycling or resetting the device after boot.

That applies regardless of what the firmware has done beforehand.

Honestly, you’d save a little on idle current, and more importantly every power-cycle causes a full sync/new session with Notehub on the next connection (using more data and a longer session), so the net result is a power loss.

At this point I’d highly recommend you look at running your schematic through the Design Review Program to get another set of eyes on the problem!

Thanks,
Rob