DIY Bare Bones Arduino Due trouble getting SAM3x8e to work!

I have created a system for a portable headphone amplifier with Bluetooth and I am incorporating an embedded Arduino to control display to a LCD screen for user input ext… I built the system shown below and everything works except for the SAM3X8E MCU chip. Basically when I connect the USB to a PC, the device is not recognized. If I hit the RESET button, the chip gets HOT. Uncomfortably hot to the touch. The erase button does not have this effect. I have done an electrical test with a multimeter and gone over nearly every part of the circuit and it all checks out. I have also referenced the Arduino Due eagle schematic many times. One thing I noticed that is most likely the issue is that VDDCORE is at 3.7V rather than 1.8!!! I cannot figure out why this is the case as I have tipple checked the schematics and actual board many times now. Can someone provide me with advice as to why this may be happening? VDDCORE is way too high and reset makes the chip overheat! If I unplug the usb then reconnect, the chip continues to overheat, (uncomfortable and “ouch!” type HOT to the touch). It is only after I power cycle the board a few times and press the erase button a few times that the overheat condition stops.

Also, I have a blue LED on my 3v3 output to the mcu. This LED turns off (indicating the regulator goes into some type of protection mode) during the overheat condition. Here are approx measured values:

VUSB: 4.78V

D+ : 1.23V

D- : 1.22V

VDDCORE: 3.7V !!! (should be around 1.6v ish, 1.95V max)

Here is the actual board along with schematic photos and the actual full eagle cad files:

Eagle Files (attached)

Portable Headphone (240 KB)

In the lower-left corner of your post is More. Click it then click Modify. Remove the two links; they do not work. Click Attachments and other options. Click Browse. Select one of the files. Click Open. Do the same for the other file. Click Save.

Sorry about that, I did not realize the links were broken. I uploaded the attachments with the correct upload method. :slight_smile:

What does your 3.3V rail measure?
I'd guess there is something wrong with the LP3872 circuitry. Perhaps beyond the thin, wandering traces and inadequate heatsinking...

3v3 measures between +2.952 V and +3.296 V. To explain this, I attached a picture of the board and it's jumper pins. Basically, the orange jumper disconnects the MCP73831 Li-Po charger IC from the rest of the circuit. The red jumper enables/disables 5v boost converter. I cannot explain why, but when I connect ONLY the orange circled MCP73831 jumper, the output voltage for the 3v3 regulator measures +2.952 V. Now when I connect the red circled 5V boost converter enable jumper AND the orange enable jumper, the voltage jumps to +3.296 V! Odd... I suspect that has something to do with the MCP73831 itself going in and out of "charge" mode or something?

I also measured the +-5v boost converter rails and they come out as +5.01 V and -5.04 V precisely. So when I have ONLY the orange jumper, there is not enough voltage to properly turn on the Bluetooth module, so it does not light up at all or boot. Only when voltage approaches 3.296 V will the Blue-Creation BC127 Bluetooth module turn on.

I 100% agree with you in regard to traces that are too thin and messy along with inadequate heat-sinking. Basically, I was under a huge time crunch to get something working (for a senior design project) so all of the analog audio components are hand routed, but for digital, the "auto" router was used. I am sure that a newer revision of the board with a hand routed digital side would help out a LOT. I know my power supply trace that leads from the MCP73831 to the 3v3 regulator then to the 5v boost converter is WAY too small.

I think … you have the SAM chip mounted wrong. Pin 1 is supposed to be in the upper left-hand corner, right? But on the more recent picture I can clearly see the pin-1 “divot” in the lower left corner instead. It looks like you assumed that “right-side up” lettering on the chip would result in pin1 on the upper left, but most of the atmel datasheets show pin 1 in that lower-left position WRT the drawings (package text is not included, though.) Also, see the Actual Due photos (Due has pin 1 in the upper right corner, and has different lettering/pin1 orientation than on your board. ( )

Misoriented chips are a common cause of high current consumption; I hadn’t considered this initially, but… Usually you get the input pin protection diodes across the power supply, in a conducting direction. In the days of UV-eraseable parts with transparent windows, I heard tales of glowing bond-wires!

I’m sorry. It will be annoying to fix :frowning: See if anyone around has a “chip quik smd removal kit” - they’re cool. (SMD Removal using a Chip Quik kit - YouTube )

(I would NOT count on the SAM chip still being “working”…)

westfw, you are the best… Just as per your assumption, I assumed the right side up lettering and the circle to the left represented pin 1 and proper orientation. I’m going to look into that quick chip stuff. Yes, soldering that chip in the first place was a bit nightmarish, so I am just going to buy a new one as this is more than likely fired. Although it would be cool to see if it still works!

You have some good eyes (must be the 3D? / sun glass)!

Edit: I am also curious to see which pin I attached RESET onto… Whatever pin it was, this caused a dead short to ground with no internal 15k pull-up as there is on the NRSTB pin which explains why the chip would overheat! I’ll write back when I either reorient the chip to see if it still functions or just flat out replace it with a new one (more likely).

BTW: W@W at that quick chip stuff, no doubt I will be using that!

UPDATE: I flipped the chip 90deg and to my surprise, my PC recognized the device INSTANTLY! I am ecstatic! Also, the weird power supply issues I was having have been completely eradicated. I am so surprised that the chip did not burn out. Soldering was difficult as I just used my hot air station rather than the quick chip. I even accidentally bent a pin removing it, but that was fixed. You can see the buggered pin on the bottom left. Might not look good, but no short circuits, so its fine. I have not tested full functionality of the chip yet, but once I get Arduino IDE to recognize the device as an Arduino Due (programing port) I will try it out on the ITDB02-3.2S TFT LCD screen.

:grin: :grin: :grin:

Now I just need to get this little guy to work:

And here it is!!! :grin: :grin: :grin:

I am using the UTFT library. Obviously, I still need to program my actual user interface as this is just the demo.

Thank you for the follow-up.

Very nice; I'm glad you got it to work.
It looks like the "runtime" numbers aren't resetting when it "restarts."

What a great project! And also thanks for posting your fix. I never knew the SAM3X8E was that abuse-resistant!