Hi,
We are seeing some Sata issue in our MX6Q public transportation system. The problem is that when system is tested at 80C , the Sata cannot be recognized most of the time, but ok at about 70C. Same failure can be seen for repeating reboot. Do you hear any similar problem from other customers or any hints would be very helpful.
Regards, Linda
What we see:
- Tested 4 boards and 2 have same problem.
- It happens when system REBOOT or recycle power under 80C, no problem below 70C or if system was started below 70C and cook up to
even 85C without restart/reboot.
- It is not related to ext HDD as we see error before checking HDD stage. putting HDD outside chamber also have same problem. See failure description below.
- It seems to us the Sata "RX_PLL" is not able to lock properly causing problem. Is there anyway/suggestion to get around the
problem by workaround Sata SW driver?
Failure description:
在SATA初始化时,在enable SATA clock(imx_sata_clk)之后,尝试去读rx_pll 状态时,发现rx_pll一直无法拉高,为不stable状态
/* waiting for the rx_pll is stable */
for (i = 0; i <= 5; i++) {
sata_phy_cr_addr(SATA_PHY_CR_LANE0_OUT_STAT, mmio);
sata_phy_cr_read(&ret, mmio);
if (ret & SATA_PHY_CR_LANE0_RX_STABLE) {
pr_info("sata phy rx_pll is stable!\n");
break;
} else if (i == 5)
pr_info("wating for sata rx_pll lock time out\n");
usleep_range(1000, 2000);
}
if you wish to start with sw, one can look at attached Linux Manual
Chapter 26 Thermal Driver and tweak passive point in imx_thermal.c
#define IMX_TEMP_PASSIVE | 85000 |
Sata "RX_PLL" is not able to lock properly due to weak signal,
one can tweak IOMUX GPR13 register, but you should use sata
analyzer to get best signals shape.
Thanks igor.
We will study but I am not sure if above is related as we now found 2 boards the SATA have same problem even at room temp of same failure behavior "wating for sata rx_pll lock time out". Unlike the one failed at 80C, these two boards can never pass SATA test at whatever temperature.
We have replaced the MX6Q in one of the above 2 failed SATA at room temp. and the board works fine now. It means the failure is not related to system HW design, but IC itself.
Please tell us under what conditions "wating for sata rx_pll lock time out" would happen? What we can try to workaround the problem? I have another board that failed SATA at room temp and 2 boards failed SATA at 80C.
Regards,
Allen
Hi Allen
this error may caused by many conditions, including chip
damage for example due to ESD shock or overheating during
soldering. In general you can send damaged chip for FA using
contacts in local marketing office.
Best regards
igor
This issue is solved by sata unstable patch, please close this issue
Hi Linda
for narrow down possible reasons one can try to run SDK sata test from OCRAM
iRAM (OCRAM) i.MX6 SDK Application
so one can diminish board noise, turning off DDR and other modules.
In general with temperature FR4 losses
increased, board currents increased, this may lead to
low levels of sata signals. Common procedure is to measure
sata eye diagram. For better signal shape one can try to tweak
IOMUX GPR13 register.
http://www.parkelectro.com/parkelectro/images/leysfinal.pdf
Best regards
igor
-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------