GEM stopped powering up

[shortwh]

No LEDs and no network connection. I checked the power supply as I have needed to change that about 3 times now; however, the board is getting power all the way to barrel connector.

I left unplugged for several hours and plugged in again and it worked. I was just about to reinstall it on the wall when I tried to cycle power again and once again it will not come up.

Later it did the same thing again and I was able to attach to the GEM's communication module and I performed a reset from the comm module and that also failed to reboot. Cycling power once again failed to get the board back. I feel that I will be able to try again later but what can I do? If power flickers here I will be without GEM for several hours?

Any ideas? Is my GEM bad?

[shortwh]

Continuing to debug. It appears to be a bad 3.3V regulator next to the 5V power barrel connector. Sometimes it powers all the way up and other times it only supplies some voltage to the board and you get excessive ripple on the output of the 3.3V regulator. If you remove the WIFI Enet module than the GEM will boot all the way up but the ripple is more than 0.5V and the LEDs are flashing dimly and different. Other times it will be solid 3.3V and LEDs work normal. I don't have a way to see if this is an over current condition but at $319 for a new board verses $2 for a new regulator I'll probably try the regulator first.

This board also has the WIFI Enet module that about 1yr into operation we had a bad storm and Enet never worked again. WIFI has been working fine. Maybe I'll buy another one of those modules also. Maybe the module really is damaged and pulling more current than specified.

At all times I see 5.05V on the board from the 5V supply (now using adjustable bench power supply). I feel certain that the 5v wall power supply is working fine but is no longer being used while on the workbench so not important for now. I checked with Oscope and found no ripple on 5v side and no power dip when powering the board up. It is only on the 3.3V side that has a problem. BTW on the bench there are no additional items hooked up like input coils or 1wire temperatur probes.

Ben,
Do you guys offer replacement motherboards for less than $319?

[ben]

Continuing to debug. It appears to be a bad 3.3V regulator next to the 5V power barrel connector. Sometimes it powers all the way up and other times it only supplies some voltage to the board and you get excessive ripple on the output of the 3.3V regulator. If you remove the WIFI Enet module than the GEM will boot all the way up but the ripple is more than 0.5V and the LEDs are flashing dimly and different. Other times it will be solid 3.3V and LEDs work normal. I don't have a way to see if this is an over current condition but at $319 for a new board verses $2 for a new regulator I'll probably try the regulator first.

This board also has the WIFI Enet module that about 1yr into operation we had a bad storm and Enet never worked again. WIFI has been working fine. Maybe I'll buy another one of those modules also. Maybe the module really is damaged and pulling more current than specified.

At all times I see 5.05V on the board from the 5V supply (now using adjustable bench power supply). I feel certain that the 5v wall power supply is working fine but is no longer being used while on the workbench so not important for now. I checked with Oscope and found no ripple on 5v side and no power dip when powering the board up. It is only on the 3.3V side that has a problem. BTW on the bench there are no additional items hooked up like input coils or 1wire temperatur probes.

Ben,
Do you guys offer replacement motherboards for less than $319?

I'll email you instructions to send the board in and Paul will take a look at it further to see if we can repair it. If it's not repairable we can definitely work something out for you.

[shortwh]

After trouble with the return bouncing back...I finally replaced the 3.3V regulator and it seems to have fixed the power up problem. However, I tried replacing the WIFI/Ethernet module and that didn't fix the RJ45 Ethernet. I will try replacing the RJ45 jack. It has magnetics built in and maybe they were damaged by the lightening strike. I'll let you know if it works as the jack is on order.

[ben]

After trouble with the return bouncing back...I finally replaced the 3.3V regulator and it seems to have fixed the power up problem. However, I tried replacing the WIFI/Ethernet module and that didn't fix the RJ45 Ethernet. I will try replacing the RJ45 jack. It has magnetics built in and maybe they were damaged by the lightening strike. I'll let you know if it works as the jack is on order.

Did you try some monitoring to verify the channels are reading correctly? Use a known load as reference.

Do you have a Serial-to-USB cable that can be used with the GEM? There's some PHY settings on the module that shouldn't become disabled but we can make sure they are still set.