Intermittent boot up on RT1176 custom board

キャンセル
次の結果を表示 
表示  限定  | 次の代わりに検索 
もしかして: 

Intermittent boot up on RT1176 custom board

ソリューションへジャンプ
478件の閲覧回数
Dave7
Contributor II

Hi,

 
I have intermittent boot up issues on my RT1176 custom board.
 
Initially, it was an intermittent power issue where the internal DCDC does not always start up and the PMIC_ON_REQ remains deasserted. Thereafter, I found out that the DCDC_PSWITCH delay is slightly less than 1ms. Hence, I increased the resistance of the RC delay from 30K to 100K and managed to get the board's DCDC to start up consistently where PMIC_ON_REQ got asserted without fail.
 
However, the board does not always boot up from the octal flash. If the board power cycle from OFF to ON within 2 seconds, there is no problem. But if the board power is OFF for more than 2 seconds, it may or may not boot up.
 
When the board boots up, the DCDC_ANA is 1.78V and DCDC_DIG is 1.09V.
When it is not booting up, the DCDC_ANA is 1.78V and DCDC_DIG is 0.98V.
 
Would greatly appreciate any guidance on what I should check.
 

Thank you,

Dave

0 件の賞賛
返信
1 解決策
130件の閲覧回数
diego_charles
NXP TechSupport
NXP TechSupport

Hi @Dave7 

Thank you for commenting the outcome for this! 

I am glad you find out the problem with your power supply. 

As a side note, per my experience, it is better to include a reset monitor IC, as it is more robust than the internal POR controlling circuit. Altougth it is difficult for me to provide an actual scenario for this. You can consider this for any future board spin. 

Diego. 

元の投稿で解決策を見る

0 件の賞賛
返信
5 返答(返信)
152件の閲覧回数
Dave7
Contributor II

Hi Diego,

Apologies for the delayed response.

Regarding the issue with my custom board, it relies on the internal reset control with the R/C network on the POR_B pin. It does not use a reset supervisory IC. I was able to confirm that the board was running XIP from the NOR flash by monitoring the LEDs.

After further investigation and some trial and error, I discovered that the root cause was my external power supply. Specifically, the supply didn't fully drop to 0V when powered off for less than two seconds, which was causing the problem. Once I replaced it with a new power supply, the issue disappeared.

Thank you again for your time and support in helping me troubleshoot this problem. Your assistance has been greatly appreciated!

Best regards,

Dave

131件の閲覧回数
diego_charles
NXP TechSupport
NXP TechSupport

Hi @Dave7 

Thank you for commenting the outcome for this! 

I am glad you find out the problem with your power supply. 

As a side note, per my experience, it is better to include a reset monitor IC, as it is more robust than the internal POR controlling circuit. Altougth it is difficult for me to provide an actual scenario for this. You can consider this for any future board spin. 

Diego. 

0 件の賞賛
返信
316件の閲覧回数
diego_charles
NXP TechSupport
NXP TechSupport

Hi @Dave7 

I am sorry for the delayed response, some time elapsed since you created this thread, and I would like to still provide my feedback.


I would recommend checking your design against the RT1170 hardware development guide if you have not done that already . More specifically the bulk capacitance for the on-chip DCDC. Are you using the DCDC_DIG only to supply power to the VDD_SOC_IN? Meaure  the current being drained from the DCDC_DIG.


If you were able to find out something else, and you can, please do not hesitate to share it, and any additional detail regarding your application , whether it is automotive or industrial temperature range. Else, we could look at your schematic.

Thank you for your patience.  

Diego

0 件の賞賛
返信
288件の閲覧回数
Dave7
Contributor II

Hi Diego,

Thank you for getting back to me, no worries about the delay. In the meantime, I’ve made some progress and have partially resolved the issue by fine-tuning the startup slew timing of the DCDC to match the datasheet specifications. The board is now able to boot up as long as the power cycle time exceeds 2 seconds.

I look forward to hearing your thoughts on this. 

Thank you,

Dave

0 件の賞賛
返信
240件の閲覧回数
diego_charles
NXP TechSupport
NXP TechSupport

Hi @Dave7 

Thank you for your patience and update. 

I understand that your RT1176 custom board can boot if power cycle is  longer than 2 seconds. I would suspect regarding any excess of bulk capacitance on power rails, this might make supply rails time to steady state longer. Also check the reset circuitry. On our RT1xxx EVK designs reset  is not de-asserted after the last  supply rail is powered up, but it is on the range of  less than 300mS ( according measurements I have done for RT1060 EVK) Are you using a reset supervisor IC, or relying on the internal reset control mechanism of the RT?  As well getting a measurement of the power up sequence with an scope can clarify this. 

By the way, how are you detecting  and measuring that the MCU is running?  Are you doing XIP from a NOR flash memory?

Best regards,

Diego

0 件の賞賛
返信