RT1064: Problems with board bring up

[MSchulz]

Hello NXP,

I made a custom board with an RT1064 and I am now in the bring-up phase.

After applying and proofing all the voltages and power-on sequences I tried to connect to the chip via "MCU-Link Pro" in J-Link mode and with "MCU-Boot-Utility".

For the use of the "MCU-Boot-Utility" I applied the "Serial Downloader" mode by setting the boot_mode[1:0] bits to "01".

Both methods failed - I just cant connect to the device.

The device gets quite warm (appr. 60°C) and does nothing. (Connections of the Power Pins already double-checked)

I have no glue how to approach this problem.

So it would be nice, if someone could explain, what is crucial to bring the RT1064 into operation.

Many thanks in advance and best regards,

Marco Schulz.

[davenadler]

Congratulations! Hope you don't find as many problems with FSL drivers as I have ;-(

[MSchulz · Accepted solution]

Done!

The solution was the wrong usage of the NVCC_PLL pin!

Not to forget: as the PLL is in the 1.1V domain, the output amplitude of an external clock source in bypass mode which drives the PLL (connected to XTALI) must also not exeed 1.1V!

Now my board is brought up and I can start developing software

[davenadler]

@MSchulzwrote: The device gets quite warm (appr. 60°C) and does nothing.

That certainly should not happen - something is seriously wrong!
I hope you have not made the same mistake I did and crossed up 1.1v supply (output and input to CPU) and 3.3v net. Do double-check the 1.1v net. For your sake I hope this is the wrong answer!

Otherwise, use an IR thermometer gun (well, or IR camera, or your finger) to find out what else is hot on your board.

Best Regards, Dave

[MSchulz]

Hi Dave,

thanks for your reply.

What did you mean with crossed 1.1V and 3.3V net? Did you connect 3.3V to the VDD_SOC pins?

I added my schematic page for the RT1064 power supply:

I think, I made the core-supply the same like shown in the "Hardware Development Guide for the MIMXRT1050/MIMXRT1060 Processor" on page 7:

The measured voltage at DCDC_SENSE/VDD_SOC_IN is 1.15V which is in the typical range shown in table 1 of the before mentioned guide.

All 3.3V-voltages are in the range of 3.27V to 3.29V.

Do you have any other suggestions?

Best regards,

Marco.

[MSchulz]

SORRY, I think I found my mistake!

I feed 3.3V into the pin "NVCC_PLL" which is obviously ment to be a cap-bypass pin for the internal PLL supply.

**bleep** happens...

[davenadler]

Glad you found the problem!
At least this one, you can solder down a new RT with the pin bent up and fly-wire the PLL cap.
Then you can find the next bug before respinning the board
Good luck!
Best Regards, Dave

[MSchulz]

yes, its not that easy because the RT is a BGA but we are lucky that the wrong connection is right at the edge of the package. so we can de-solder the old one cut the fanout to 3.3V and try to apply a thin wire to connect to a cap before we solder a new RT in again...

[davenadler]

Sorry I forgot about BGA, I'm using RT1024 in easier package.
But, that RT is probably toast if it got that hot.
Good luck!