NUC LED will not turn on after USB power event and computer shutdown

[Hermanoid]

This event has actually happened twice, to a NUC10FNH, and more recently, a NUC8i7BEH. 

We're deploying these NUC's in an embedded robotics application, and unfortunately, we're prototyping stuff so fast that we've fried a few USB peripherals. In particular, these USB-to-TTL converters (https://www.amazon.com/gp/product/B07D9R5JFK/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1) have gotten smashed or had their wires crossed twice in a pair of suspiciously recurring flukes. 

These converters fail pretty spectacularly in a show of melting wire and foul-smelling smoke. In the process, another USB peripheral (an ODrive) connected to the same USB splitter (4-way) stops responding to any communications. But, the USB functions of the NUC (and the NUC itself) continue unchanged (I can still use the Bluetooth dongle in the splitter for my keyboard and mouse) until the NUC powers off. After it is shut down, the NUC refuses to power up again - no light, no nothing, even after unplugging power before pressing "on" again.

I've gone through the online boot issue checklist to no avail.  "Bad power source" - I've tried multiple in multiple outlets. "Drive Problem" - I can take the same SSD and transplant it into another NUC, and it works perfectly. "Different Power Issue" - I've gone through the recommended procedure but nothing changed - the external LED remained unrelentingly dark.

I would think that these two NUCs are fried and totaled, except that they continued to work (through a full compilation of OpenCV in the case of the first NUC) with full USB functionality until turned off. This makes me think there may be some means of resetting these NUCs to save them from their (very expensive) funk.

Does this sound familiar?

[LeonWaksman]

NUCs having overcurrent protection on the front and back panel USB port. So, short in the converters connected to the USB port should not damage NUC. 

I would try to disconnect the Power Adapter and the CMOS Battery Connector and reconnect them after 15 minutes. In the NUC FN this connector is located on the board bootom side, therefor is relatively easy accessible after removing the NUC bottom panel. For the connector location see Item "C" on page 23 in Technical Product Specification.  Check on NUC FN if this works. In NUC BE in order to get access to the battery connector, the board should be removed from the case, since it is located on the board upper side.

Leon

[ClariceStarling]

These things seem easier to diagnose when you know your power adapter is working.  The only thing I don't like about my new NUC is that the transformer on the cord  doesn't have an LED light.  Not sure why Intel skimped on that because 99% of new computers have cords with LED lights! 

[ClariceStarling]

Let me guess ... you don't have an LED light on your power adapter to tell you it's working?

[Hermanoid]

No - but I do have multiple power supplies. I used on power supply for a functional nuc (and it worked), then moved that same power supply to one of the dysfunctional ones with no luck.

[Hermanoid]

@LeonWaksman - I tried your idea, disconnecting the battery on the underside for fully 15 minutes, and after reconnecting power the power button still refuses to illuminate and the NUC refuses to turn on. Do you have any other ideas?

[LeonWaksman]

I don't really have other better ideas. You may try  momentary short between pins 6-8 on the Front panel Header (see page 28-30 and explanation in para. 2.2.4.5.4 in the Technical Product Specification). This will replace the function of the Power Switch.

When the CMOS battery was disconnected, did you also disconnected the Power Cord?

Leon

[Hermanoid]

Yeah, I disconnected it. I fully removed the board from the casing, unplugged everything, unplugged the CMOS, waited 15 (well, closer to 30) minutes, then plugged the CMOS back in (I also tried it with it unplugged), plugged in power (and HDMI), shorted the two "reset" pins (because "resetting" seems like a good thing to do) with a female-to-female jumper, then shorted the two power pins. No response. I also tried the power button, of course.

I wouldn't be so concerned (we already replaced these NUCs and continued with development) except that we're all very concerned that this is going to continue happening.

[n_scott_pearson]

When the power supply is connected, does the green LED light up inside the chassis? If you cannot see this LED light up, either the power supply is bad or the NUC has suffered a failure in its power circuitry.

...S

[ClariceStarling]

Wondered why the power adapter didn't have a light ... but it could be that since they have the green LED on the motherboard, they figured they don't need a light on the power adapter also.