Variant board do not work :(

hello
I am trying to design a board with ATSAMD21G17 with QFN package... and after 4 months, I have received my board correctly assembled. Anyway When I try to use the SWD connector to burn it or just to supply 3.3V , My chip samd21G become really hot and I can not never obtain 3.3V ... then big problem!, I think I make an error in the design (because I have more than 50 boards with this problem!) but after reading manual of the chip, check my design I do not manage to find the solution... Maybe if there is an expert he will see the problem very fastly (I assume a supply problem ... I have also an other board with ATSAMD21G18 and no problem with them...

thanks for your help , I can send you also the layout

the die pad (pin 49) is to GND but I am not sure , maybe problem is from that?

I think the only way you are going to find the error is to go through your board and check the physical connections pin for pin using an Ohm meter. Specifically, check that all ground pins read near zero Ohms to ground and high resistance to 3.3V and that all 3.3V pins read near zero to 3.3V and high resistance to ground. Your schematic seems to be correct but we can't compare the board layout because we only see one side. In fact, I can't see how the processor is connected to ground.
Good luck!

Are you sure?

Unless you've figured it out by now, better of to dissasembly one board instead of guessing. Try to keep at mininum the components in the board so you can investigate the issue.

Remove everything, including the chip and start off from the power supply; no serial comms components; no pullup resistors, and so on. If the power supply is functioning properly, then move on and place the chip again.

It seems to be time consuming, but it's not.

@edouard56899, your topic has been moved to a more suitable location on the forum. Installation and Troubleshooting is not for problems with (nor for advise on) your project :wink: See About the Installation & Troubleshooting category.

Hello
Yes good Idea
I have already removed just the Chip and everything seems to be correct. I have a DC DC concerter and when the Chip IS not present ,i have 3.3v ,but when chip is present it is no more the case...i have just 0.9v and the ohmeter indicates 4ohm between 3.3v and ground.

I check again and again the pins of the Chip without understanding the problem...

1 Like

Hi,
Did you make and prove a prototype before going to PCB?

Is this your first PCB?
Can you please post a complete export jpg of your PCB, rather than a part of it?
Did you place test points on your PCB to measure power supplies and input/output signals?

Can you please tell us your electronics, programming, arduino, hardware experience?

Thanks.. Tom... :smiley: :+1: :coffee: :australia:

1 Like

Hi @edouard56899

In the layout image that you posted, I noticed that the GND pads aren't connected to anything. Did you fill the unused areas with a ground plane prior to manufacture?

1 Like

Hello
Thanks again !!!
Then ,no i did not make a prototype before and i have a batch with 50.boards with this problem... I was sure but now i am sure that I miss something in my design. If you have eagle i will.send a link ,in this matter you will have schematic and layout.

Yes i have a ground plane in the two side and gnd in the Chip IS linked with via. It IS not visible in the picture sorry.

In fact i have already made a board with samd21g18 but in august this package was unavailable...then i make antoher layout with samd21g17. But i did not think there was a Big compatibility between them.

I am completely sure that problem Come from thé chip because if i remove it ,everything IS ok and i have 3.3v After my DC DC concerter...

It IS like vddcore connected to 3.3v or gnd but it.is not the case... Really incomprehensible

1 Like

Hi,

No, please post EXPORTED image(s) of your PCB.

Is that your complete diagram, I don't see any I/O connections?

Thanks.. Tom... :smiley: :+1: :coffee: :australia:

1 Like

@pascalou3131 VDDCORE should be around 1.8V.

1 Like

Hi,
What is going on here?

@pascalou3131 or @edouard56899

Tom.... :smiley: :+1: :coffee: :australia:

1 Like

I would breadboard the circuit myself, to check you can make it work.

Just add the basics first, enought to upload a blink sketch etc.

Normally you do that before designing a PCB.

1 Like

Hi, @pascalou3131 or @edouard56899

Can you please post an EXPORTED jpg image of your PCB?

Tom.... :smiley: :+1: :coffee: :australia:

1 Like

Yes for this project I change just the chip SAMD21G18 by SAMD21G17 .. with the G18 this board was working perfectly, I do not know why it is not working with this chip G17 (in QFN package)
the die pin is normally connected to the GND no? I am not sure bacause in the TQFP48 you do not have this die pin...

Not sure my problem is coming from that? but it is the only difference...

1 Like

1 Like

Hi,
You should be able to produce an image with MUCH better resolution than that.
It is totally unreadable at individual component level.

Are you using the EXPORT in Eagle?

That image is only 27K png, where as your schematic is 187K png.

Tom... :smiley: :+1: :coffee: :australia:
PS. What is the application of this project?

1 Like

Try downloading the image you posted above.

Can you read it ?

1 Like

Hi,

Yes, but you should have posted it as a new post.
What is the application of the project?
Have you designed PCBs before?
Did you apply a silkscreen to your PCB to identify your major components?

You will probably have to get the DMM out and check continuity of connections on your PCB with those in your schematic.

Tom.... :smiley: :+1: :coffee: :australia:

1 Like