The Things Uno external power polarity reversed - Fix?

While trying to install a switch to use with a 9-volt battery for power, it looks like I managed to reverse the polarity and now The Things Uno will not power up from the external power jack. The device does still work from a USB power supply so whatever I fried must be isolated to the external power supply section. Looking at the schematic I see several components but I’m not sure which are most likely damaged or how to test. Any suggestions would be greatly appreciated.


Most likely, the Voltage regulator, VR1, unless its one of the rare ones that is reverse polarity protected.

How to test ? Well the output is labeled +5v, so is it ?

If you connected through the input jack nothing would have happened at all, given the diode that is immediately there.

The exception would be if you also had something else connected (like a serial cable) that had a common ground with the power supply, in that case you could potentially see far reaching damage.

Thank you for the responses. It appears that the diode (D3) may have been damaged. When an external 9v power supply is connected, there is 9V at the diode anode (input), but nothing at the other side or at the input terminal of VR1 (also nothing read at Vin). There is continuity from diode cathode to VR1 input. I’m not sure what the value should be for the diode.

That is very surprising and should not have happened unless a completely unsuitable part was chosen or external conditions beyond simple reverse polarity applied, or unless damage to subsequent components cause the diode to fail only when power was connected the correct way.

If you look at the design documentation for an actual Arduino Uno (and not this misnamed thing that is no sort of “Uno” at all) or for a Leonardo which this board more closely resembles, you might find a suggestion of a perhaps more suitable diode.

Is the input of the regulator shorted to ground? That could kill even a properly selected diode…

I had assumed, from your description, that you were adding the switch on the cathode (UNO ) side of the diode. Reversing the polarity on that side could cause significant damage to the board, but often the regulators go first.

Only a large reverse voltage would damage the diode, so looks like you shorted it to ground ?

Sorry - I was attempting to add a switch on one of the leads coming from a 9v battery in a prototype instrument enclosure that houses The Things Uno. We were using several different types of battery connector - some with red and black leads and some with two black leads (that’s my excuse anyway). So I just assumed that I had somehow reversed the polarity entering the external power jack, but there should never have been any more than ~9 volts from the battery.

I’m no expert with the electronics (obviously), but a reverse voltage of ~ 9 volts shouldn’t have destroyed the diode, should it?

There are diodes with a max reverse voltage around that level, but it would be daft to use one in that location on the things UNO, its meant to survive that sort of reverse voltage.

That’s what I thought - the diode must be there to protect the rest of the device from reverse polarity - and yet this one appears to be fried. Actually I discovered that we have multiple prototype nodes for which I managed to fry the diode after making this battery switch modification. We also have several nodes with the battery switch installed that work fine - I still have to examine the actual wiring on the nodes to see what is different - some may have the switch installed on the battery’s ground lead while others might be on the positive lead.

Is it possible that powering from USB at the same time that a 9 volt battery is attached (and maybe connected in reverse) could destroy the diode? I can’t look at a schematic and recognize whether that might be a problem.

Odd thisngs can somtimes happen when grounds get confused, such as when a project is plugged into a PC via USB, and a project is also powered from a seperate mains supply.

But a project powered by battery ought to be safe, I do that all the time.

I thought so too. In any case, I have ordered a number of diodes (1N5817) with 20V reverse voltage that I should be able to use to replace the damaged original diodes - the originals are surface mount but relatively large and I think I can remove and replace - it won’t be pretty but hopefully will allow for recovery of the functionality of the external battery power connection with reverse polarity protection.

Known issue guys. I have posted before about this. There is a small series of this board where the blocking diode was soldered in the wrong way. As it seems you have one of those. So no need to buy new ones, just solder it out and and back in with polarity reversed.

Thank you for the response. If I search on blocking diode I find your post from a year ago - but obviously the issue was not known to me or I would have returned the units immediately rather than wasting a day in the field and now several hours removing and soldering surface mount components in 10 nodes - components are cheap - time is not. These TTN “Unos” cost approximately $60-70 ea in the US and I think it’s reasonable to expect the boards to be fully functional when received. This was 10 boards out of the 14 that I have purchased so far that apparently had this problem.

I only had two fortunately. But finding the issue indeed took some time. Perhaps you can complain to your US supplier.

I have notified the supplier - it’s unfortunate that the things network did not do the same.