SAMA5D27C-D1G-CU USB Connection and Reconnection while flashing NAND SLC

[Naveen-Krishnan]

Hello friends,

I have a problem with one of my project. I need a help or some ideas to solve that issue, please help me out guys.

We use SAMA5D27C-D1G-CU as the main processor and MT29F2G08ABAEAWP NAND flash for storing the file system. For dev purposes we used SD-Card to boot.

When we use SD-CARD to boot our board we have no issues, everything works fine. But the problem is : when we flash the file system from SD-CARD to MT29F2G08ABAEAWP NAND flash through processor using bash command; any pheripheral which is connected to the USB bus of SAMA5D27C-D1G-CU is having CONNECTION & RECONNECTION issue periodically.

I don' know where it is coming from. Please help.

[robertferanec]

Please, could you clarify? Is the USB device re-connection happening during the process of flashing? Or it is happening after you flash the NAND and then you run your board from NAND?

[Naveen-Krishnan]

Hello Robert,

Thanks for your reply. The USB device re-connection is happening only during the NAND write. After flashing the file system into the NAND and turning on the board I have no problem.

Regards,
Naveen

[robertferanec]

This could be for example layout (crosstalk) or power problem (noise on power, power drop). But it is hard to say just like that - it could be also something else. How many layers has the PCB?

[Naveen-Krishnan]

Thanks Robert Feranec for your response. It is a 6 layer PCB

--TOP SIGNAL ( USB Bus, I2C UART signals)
--GND
--SIGNAL LAYER 1 - ( LCD Interfaces signals, NAND SLC interface signals)
--PWR
--SIGNAL LAYER 2 - ( Ethernet, SPI, SDIO signals)
--BOTTOM ( BUS USB )

I would like to add some details to the problem. We manufactured 5 PCBs and in 2 PCBs the problem does not exist, it works perfectly. The problem is present in the 3 other PCBs. Regarding the differences, there are none, all the boards have same design and same components.

I used an oscilloscope to check the Power supply of the NAND SLC during NAND writing operation, I saw peaks exactly at the instant when the USB pheripherals connection & re-connection happens. This noise in the power supply does not exist in the 2 PCBs which is working fine.

Really bizare. Please help me out guys !

Regards,
Naveen

[mohsin_qau]

Please mention your problems in the faulty PCBs.

[robertferanec]

We manufactured 5 PCBs and in 2 PCBs the problem does not exist, it works perfectly. The problem is present in the 3 other PCBs. Regarding the differences, there are none, all the boards have same design and same components.

- Means there is something running on the edge in the design or something is not correct.

I used an oscilloscope to check the Power supply of the NAND SLC during NAND writing operation, I saw peaks exactly at the instant when the USB pheripherals connection & re-connection happens. This noise in the power supply does not exist in the 2 PCBs which is working fine.

- Now the question is, if the peak is the cause for disconnecting USB or if it will just always appear when USB is disconnected.

[Paul van Avesaath]

I agree with robert on this, something is at the edge of its capabilities, but since there are 2 working PCB's and three failing it should be a matter of finding what is what.
it can be explained by component variation, but then you could measure it.also some soldering differences could cause it.
where all the boards bought at one supplier? at the same time? did you manually solder the boards?

some schematics of you design would be helpfull for us to understand what might be causing this.
maybe one of the boards is pulling a bit more power than the other,
you could try and add some decoupling?

you mention the failing is when using the USB.
can you share some details / schematics of this design?
how much power is being drawn by the USB? 5V/0.5A or 5V/1.0A, and how is power supplied by your board?

[Guest]

Much obliged for your answer, the USB gadget re-association is going on just amid the NAND compose, subsequent to blazing the record framework into the NAND and turning on the board i have no issue






Life is what happens when you’re busy making other plans

[Naveen-Krishnan]

Hi guys,
Till date we did not find a solution to the problem while writing to NAND from SD Card. But we found workover method, instead of using SD Card, we used USB interface to flash by using SAM BA application through our PC. We find it quite reliable.

I will you guys updated on the issue.

[robertferanec]

Thank you Naveen-Krishnan for sharing your solution.