Atmega256RFR2 voltage

[asmunderik]

I am having some trouble with my Atmega256RFR2.

Some background information:
I have created a custom PCB, this pcb contains the atmega256rfr2 with necessary circuitry.
It also contains a power circuit, an Attiny and a USB connector.
It works, i get wireless connection with the zigbee radio, i have connection and have tested the atmega mcu.

My problem:
The Zigbee radio is not able to recieve anything after the voltage rises abow 3.25.
Everything else like I/O pins work fine with voltages abow 3.25.

What i have done:
I am testing the wireless connection in Atmel Studio with the wireless composer 7.0 and i am able to hold a steady connection.
I have looked closely on the power signals while tuning the voltage up by small increments, nothing noticable is happening there.
I have bypassed the power circuit in case anything was wrong there.
I have ruled out that the ATtiny can be the problem, because i have another PCB without the ATtiny with the exact same case.
I am currently only powering the mcu and its external components with a lab bench power supply.

In the datasheet it is specified maximum voltage 3.6V. It is also stated in the datasheet about internal regulators.

I can not find the reason why the zibgee radio terminated its transmission when the voltage rises abow 3.25 volt.

Thanks in advance!

[Psi]

what is "abow" its seems like a combination of above and below, which seems a little contradictory.

Can you post a link to the datasheet for your zigbee module?

[asmunderik]

Well ofcourse i mean above! hehe sorry about that!

Here is a link:

http://www.atmel.com/Images/Atmel-8393-MCU_Wireless-ATmega256RFR2-ATmega128RFR2-ATmega64RFR2_Datasheet.pdf

[asmunderik]

Here is a picture i took of the Atmel Studio software when i tested different voltages. At first i start up with 3 volt, slowly incrementing to 3.25, the drop is when i am at 3.26 volt. I go up to 3.3 volt, before i go down again to 3.25.

[Psi]

are you clocking the atmega from a xtal or from the RC osc?

If its from the RC osc maybe you are having a baudrate issue as the voltage increases.

But we really need to see a schematic to know whats going on

[asmunderik]

For the test you can see in the picture i posted, the internal RC oscillator 16Mhz was used and prescaled down to 8Mhz.

Adding a picture of schematic for atmega256rfr2

[asmunderik]

I have now located the mistake! AREF should not be connected to VCC.. this made a short in the MCU (i think)... Problem solved..

Thanks for your help!

[senso]

And you should have 1uF in EVDD and AVDD and EVDDS should be feed via a ferrite, I'm also working with rfr2's and all my board work perfectly from 3.3v down to 1.9V.

[ataradov]

And another thing to keep in mind - you should not vary supply voltage a lot while MCU is running. Small variations like this are fine, but you should not switch between the extremes without completely disconnecting the power. This confuses some of the internal circuitry.

[senso]

I have around 10 running with 2 AAA batteries, and would like to add some sort of buck/boost dc-dc regulator to extract all the juice from the batts, but I need one with an ultra low quiescent current, because the micro's wake up once every 15 minutes for less than a second.