FS85 SBC Power Interrupt Issue

[asaunder]

Hello, I am using an FS85 SBC on my PCB and have loaded my own custom OTP on the device. It works well the vast majority of the time, but there is one repeatable failure I am encountering. If the supply to the SBC (28V in my case) is quickly cycled on and off (within 50ms for example), a state is entered in which the VPRE seems to get stuck an infinite loop of turning on and off every second or so until I disconnect power and reconnect. This can happen by accident if I am connecting my PCB to a power supply and accidentally bump the 28V slightly before making full contact, and it also happens in a controlled test setup where I send a momentary power interruption using a waveform generator controlling a MOSFET.

Does anyone have experience with this or have an idea of why the SBC would enter such a state when power is momentarily disconnected? 

Note: It is running in debug mode so I don't think it has anything to do with the watchdog.

[guoweisun]

How many boards do you find this kind of issue?

Can you repeat this issue?

[asaunder]

I've seen it happen on two different boards that implement the FS85 SBC. Yes it is an issue I can repeat by inducing a very short power interrupt.

[guoweisun]

FS85 can fully support 28V no problem, so in your failure boards if set input voltage to 24V,does it works well?

[asaunder]

Yes under normal circumstances the board can work fine at 28V or 24V. This is a consistent issue related to a momentary power disconnect followed by full power again (around 50ms time disconnected or less).

Based on the datasheet, the SBC seems to enter the DEEP-FS state after the power interrupt. Then the infinite loop begins. VPRE and all of the regulators go off for approximately 4s and then come back on for about 1s over and over. It appears that the infinite autoretry OTP feature is occurring while in the DEEP-FS. WAKE1 and VBOS are both on the whole time in the error state, even when VPRE and the regulators turn off.

What I am not sure of is why, once power is steady, VPRE and the regulators continue to turn off. Remember, this issue only occurs when power is supplied, removed, and quickly reapplied (within far less than a second). If power is turned on normally, my design works fine.

In case it helps, I've linked an image of my schematic.

[asaunder · Accepted solution]

I believe I have discovered what is happening. It says in the datasheet that DBG must come high with VSUP before WAKE1. If a short power interrupt occurs (50-100ms) then it is enough time for DBG to go below its threshold but WAKE1 to stay above its threshold. Thus when DBG comes back up, WAKE1 is already high and so the device does not enter DBG mode. I do not have a watchdog running in my SW yet, which explains the infinite resets every 4s following the entry into the error state.

Thanks for the replies!