NorFlash unstable problem on P2020

取消
显示结果 
显示  仅  | 搜索替代 
您的意思是: 
已解决

NorFlash unstable problem on P2020

跳至解决方案
1,492 次查看
xiaocheng
Contributor II

We use the CPU model P2020NXE2KHC, the manufacturer with NorFlash is SPANSION, model: S29GL256P10TF101, configured on the CS0 of ELBC, efe05000: f0001001 fe000ff7, other CSs are not enabled;

At this time, we use in the firmware: mw.w f0000AAA aa 1;mw.w f0000554 55 1;mw.w f0000AAA 90 1;md.w f0000000, we find that the manufacturer ID read is wrong, normally it should be 0x0001, but there is When it is 0x0211; grabbing the waveform of the two bits of data 0x0200 and 0x0010 (the same) found to be very strange, it should be low normally, the details are as attach.

Have you encountered a similar situation? Or any suggestions?

标记 (1)
0 项奖励
回复
1 解答
1,434 次查看
xiaocheng
Contributor II

Thank you, everyone, the problem has been resolved, it is determined that the chip itself has a problem.

在原帖中查看解决方案

0 项奖励
回复
3 回复数
1,435 次查看
xiaocheng
Contributor II

Thank you, everyone, the problem has been resolved, it is determined that the chip itself has a problem.

0 项奖励
回复
1,485 次查看
ufedor
NXP Employee
NXP Employee

Data signal behaviour in the provided trace could be interpreted as contention.

Please provide the corresponding processor connection schematics as PDF for inspection.

Also please confirm that LGPL[04]/LGTA/LFRB/LUPWAIT is pulled up by an external resistor.

0 项奖励
回复
1,478 次查看
xiaocheng
Contributor II

We check that LGPL [04] / LGTA / LFRB / LUPWAIT has been pulled up, specifically it is pulled up to 3.3V through a 4.7K resistor.

One thing to add is that we actually mount 3 hardware devices on ELBC, CS0 (base address is 0xF0000000) and CS1 (base address is 0xF4000000) each mount 1 piece of NorFlash, the model is the same, the size is 32MB, in CS2 1 FPGA is mounted on it;
The following 2 phenomena were observed today:

First explain the current configuration of ELBC:
efe05000: f0001001 fc000e41 f4001001 fc000e41 .......A.......A
efe05010: ea001001 ffc00e31

[Phenomena 1] Both CS0 and CS1 can read and write 64MB, and the data will be there after power failure. The actual chip should be 32MB;
[Phenomena 2] When CS0 is normal and CS1 is abnormal (here means that the ID cannot be read correctly, the readout is 0xF400), write and read after CS0 is erased, and it is found that the data is normally written to CS0, but the program is read before erasing The original value of CS1;

0 项奖励
回复