1
\$\begingroup\$

My team and I are creating a project where a sensor board is controlled by two Arduino Nano's, connected using a UART daisy chain. Whenever the board is turned on using a USB connection, the Arduino's and their sensor function normally. But when the 12v source is added, the programmer (or something else inside of the Arduino) of the Nano breaks. We are powering 16 leds and 16 tcrt5000's with these two Arduino's.

My questions are:

  1. What causes these Arduino Nano's to fry?
  2. If it's a short circuit, how do we keep it from frying again?

I've added a schematic below. The Schematic of the Sensor Board

*EDIT:

First of all, I'd like to add something: the two nano's have their own regulators. If I take out one of the two nano's and plug in the source, it runs fine. The LEDs and the Sensors work normally. If I add the second one, it breaks down.

Second, to clear up, I've added a picture of our setup by adding the design of the PCB. TOP LAYER BOTTOM LAYER

EDIT 2:

To be specific of the burned down Nano's: After plugging in both nano's and plugging in the source, the two nano's were not controlling the LEDs. After removing the power, getting burned by the hot regulator and hot pins from the Atmel, and plugging it into my laptop, I couldn't get the code to upload again because of a "Programmer not responding" Error. It seems like something inside the programmer got fried. I just don't understand how.

\$\endgroup\$
12
  • \$\begingroup\$ Difficult to tell from that schematic. Pictures of your setup would certainly help. \$\endgroup\$
    – Damien
    Mar 24, 2021 at 9:52
  • \$\begingroup\$ Can you link the part number of the actual voltage regulator on these things? When I google Arduino Nano I only get weird schematics where they specify a 2.5V output LDO ncp1117st25t3g from ON Semi, which they claim to give 5V out. RTFM disagrees. So either the design is completely nuts or they are actually using a different part. \$\endgroup\$
    – Lundin
    Mar 24, 2021 at 10:10
  • \$\begingroup\$ @Lundin It's the AMS 1117 5V regulator \$\endgroup\$
    – Laurens Verbruggen
    Mar 24, 2021 at 10:13
  • \$\begingroup\$ The SOT-223 one? \$\endgroup\$
    – Lundin
    Mar 24, 2021 at 10:19
  • \$\begingroup\$ @Lundin Yep, it's in a SOT-223 package \$\endgroup\$
    – Laurens Verbruggen
    Mar 24, 2021 at 10:21

2 Answers 2

Reset to default
2
\$\begingroup\$

My guess is the Nano's voltage regulator overheats as you are feeding it 12V. It has to drop that down to 5V and it does it by wasting the energy as heat. Get yourself a 12->5V buck converter power supply board and use this to generate the 5V from your batteries. Your Nanos will thank you. The power supplies are commonly called a LM2596 adjustable buck. stroomvoorziening

\$\endgroup\$
2
\$\begingroup\$

We are powering 16 leds and 16 tcrt5000's with these two Arduino's

If each LED and each Reflective Optical Sensor consumes 20 mA, that's a total current of 640 mA and the poor internal regulator on the nano is going to rapidly heat up and might shut down or, worst still, destruct.

It's not meant to be used with this amount of abuse because it is a linear regulator and will dissipate a lot of power (heat) in making the 5 volt rail from the 12 volt input supply: -

enter image description here

The NCP1117 is just not man-enough to supply this current and drop 7 volts across itself (about 4 watts dissipated).

\$\endgroup\$
7
  • 1
    \$\begingroup\$ So you found the same schematic as I did. That's a 2.5V output regulator, so... which one is actually used. The quality of these hobbyist boards overall is depressing. \$\endgroup\$
    – Lundin
    Mar 24, 2021 at 10:13
  • \$\begingroup\$ AMS1117-5.0, which is apparently the one actually used (hey why would anyone use the part specified in the schematic!) seems more rugged and is specified for 1100mA typical, 1500mA max. Available in different packages with slightly different thermal characteristics. So it should in theory work. \$\endgroup\$
    – Lundin
    Mar 24, 2021 at 10:23
  • \$\begingroup\$ @Lundin yes, it's meant to be an NCP1117ST50 device. But, also given the physical size of a nano, it ain't gonna handle much power before it overheats big time!! \$\endgroup\$
    – Andy aka
    Mar 24, 2021 at 10:25
  • \$\begingroup\$ Yep, the size of the heatsink makes all the difference for LDOs. So if they board manufacturer just smacked on AMS1117 as drop-in replacement, it might not fare much better if there's no cooling anyhow. Still, I kind of doubt that thermal shutdown would cause the part to burn up... unless it's the traces themselves frying. Starting to suspect that the problem is elsewhere. \$\endgroup\$
    – Lundin
    Mar 24, 2021 at 10:31
  • \$\begingroup\$ @Lundin possibly there are spikes on the 12 volt that exceed the 20 volt max rating. Or, if the voltage is slightly higher than 12 volts and the output current demand is slightly more than 1 amp, the regulator might fail: The regulator output current must not exceed 1.0 A with Vin greater than 12 V - quote from data sheet. \$\endgroup\$
    – Andy aka
    Mar 24, 2021 at 10:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.