This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
hello, ti expert:
The sdk version we use is 0703, and now we encounter a DDR problem. We use Micron’s LPDDR4 chip, the model is MT53D1024M32D4DT-046 AAT.The power-on and power-off test of the switch machine has always been very stable. Now there is a demand that a simple reset of the TDA4 chip can also achieve a stable restart. Now the problem arises here. On the MCU side, the reset is implemented through the MCU_PORz pin, TDA4 reset without close power, and it is found that lpddr4 will fail to initialize. Tested hundreds of times, occasionally a few times, the frequency is also very high. Stuck in the Board_DDRChangeFreqAck function, has been blocked in HW_RD_REG32(BOARD_DDR_FSP_CLKCHNG_REQ_ADDR) & 0x80 here, as shown in the picture below
We now implement reset by pulling down MCU_PORz, the datasheet is explained as follows
I saw that there is an api that can also implement reset. What is the difference between the following api and the above direct control pin?
Sciclient_pmDeviceReset(SCICLIENT_SERVICE_WAIT_FOREVER);
Regarding this problem, I found a lot of similar problems on E2E, the following link generally means that it can be solved by changing the configuration, and the configuration file has a newer version
The following problem is solved by frequency reduction. We require DDR4 to run at the highest bandwidth of 2133 clock, 4266Mb/s
e2e.ti.com/.../tda4vm-ddr-init-failed-in-sbl
This is 4266Mb/s E2E, but no one replied
e2e.ti.com/.../tda4vm-ddr-at-4266-init-failed
Hi,
Do the following two bullets describe the correct observation?
In the reset scenario, does the DDR initialization always fail, or just occasionally?
Regards,
Kevin
just occasionally
In the 500 tests, only 4 times, that is to say, not every time will be blocked in DDR init, so it is an unstable problem