Hello everyone.
I have a custom board based on i.MX6 ( Quad core version) and with linux running on it.
The board is bootingU-Boot from SPI. And then boots Linux from Network, SD-Card or CFast (on SATA). All works fine.
The configuration is done via GPIO overrides (BOOTMODE '10', "internal Boot"). The Fuses are not set.
I have an external reset-button which controlles the POR_B pin via a FPGA open drain output. The PF0100 also can generate a reset on POR_B pin with his open drain output RESETBMCU.
Now I have discovered that a reset propagated via FPGA doesn't work. The POR_B behaviour looks fine. So I had a deeper look into this issue.
What I discovered: The i.IMX6 seems to ignore the GPIO overrides (or wrong read) and is falling back to serialdownload mode
If USB cable with Manufacturing Tool is connected (SPI flash is empty), the i.MX6 resets correctly and is booting in serial download mode.
May I solve this issue with burning the FUSEs? Or do I have to solve this issue like it is done in i.MX6 POR_B pin behaviour by also resetting the PF0100?
KR
Ingo
Solved! Go to Solution.
Hi Ingo
yes, recommended to reset PMIC too. In your case
one can analyze boot pins with oscilloscope (check correct voltages)
then attach jtag and check SRC SBMRn registers, they should have
correct boot configuration. Also after POR i.MX6 works in LDO mode,
if PMIC was configured to LDO bypass before POR, this allso may cause problems.
Best regards
igor
Hi Ingo
yes, recommended to reset PMIC too. In your case
one can analyze boot pins with oscilloscope (check correct voltages)
then attach jtag and check SRC SBMRn registers, they should have
correct boot configuration. Also after POR i.MX6 works in LDO mode,
if PMIC was configured to LDO bypass before POR, this allso may cause problems.
Best regards
igor
Hi Igor,
that was a good idea with the SRC_SBMRn registers. I discovered that the strapping option have been overwritten by an EIM slave. Now, I've put the EIM bus slave into a High-Impedance mode while doing the POR_B (plus some time like mantioned in boot strapping timing) and it works fine.
Thanks a lot!
Regards,
Ingo
Ingo,
I am seeing the same problem you had where the i.MX6 does not reset after POR_B. Can you explain what you did to change the EIM bus slave into a high-impedance mode while doing the POR_B?