TI中文支持网Part Number:TUSB4041PAPEVM
The usb 2.0 hub design is follow TUSB4041PAPEVM Schematic.And use the external eeprom for PID/VID setting. The eeprom bin file is showed as picture.
In our application, the Host is the Mac mini (mac OS system). The hub is used to communicate with our product, and the product will power cycle which means the device will disappear and reconnect.
There exist a strange phenomenon: if the hub first power on ; or mac mini restart ; the hub will drop (the mac mini can't read hub device) still for about 6s when our product power cycle and reconnect. And It is particularly important to note that this phenomenon only occurs under the conditions mentioned above.
We capture the data of USB, and show error as below photo. Could you please help analyze it and provide some suggestions?
Links：
Hello! We have received your case and the investigation will take some time. Thank you for your patience.
,
Links：
Hi:
What customer for this issue?
Can you remove external EEPROM and test again? hub should work in its default mode.
Can you take waveform for 3.3v/1.1v and GRSTz signal?
Can you send hub schematic for review?
A block diagram will be easy to understand the setup.
,
Lingnan Zeng：
Hi. Thanks for your reply.
The issue is happened in our project. and the schematic and block diagram showed as below:
The mac mini try to install the system into the product through USB2.0.And the restore process have two steps. there exist a reboot between two steps.In the normal case, Mac mini only need to spend 3s to find the device (testing product). But in the abnormal case, Mac mini will spend at least 8s to find the device, and we check the system log of mac miin, find that there exist a reconnection of the hub.But it is strang that this phenomenon only occurs under the conditions (the first test after power on hub board or restart Mac mini). We suspect that there might be some abnormality in the configuration of TUSB4041?Could you help to review the eeprom setting as the last meassage showed?
For below two suggestion,I will upload the result once it done.
Can you remove external EEPROM and test again? hub should work in its default mode.
Can you take waveform for 3.3v/1.1v and GRSTz signal
Mac_TI_USB2_Controller_Board_Rev2.0.pdf
,
Lingnan Zeng：
Can you remove external EEPROM and test again? hub should work in its default mode.
Answer: there is nothing improve to use the default mode.the issue still exist.
Can you take waveform for 3.3v/1.1v and GRSTz signal
Answer:Capture the waveform of 3V3,1V1,GRSTz signal. And there have not voltage drop or shake.we can refer to the screenshot for detail. The GPIO means the time of fail trigger.
,
Lingnan Zeng：
Hello, I already update the files from my side for few days. could you help to give some suggestion. Thank a lot.
,
Links：
HI:
For Type-C port, there is only 5.1k PD on CC pin which is ok for UFP. But you can only get 500mA default for USB2 VBus.
With 500ma Vbus, it is used not only for 3.3v and 1.1v supply for hub. but also for all downstream port VBus. Based on USB2 spec, each downstream port Vbus should support 500ma . I don't think power is enough for your whole system.
Once mac mini restart, it will take time for CC communication then turn on Vbus and hub got reset due to Vbus reset. Can you check when Vbus gors high after mac mini restart?
,
Lingnan Zeng：
Hi, thanks for your reply. But the hub board have not use the 5V (vbus) of PC. there exist an external 5V for power supply showed as below, and you can refer to the attchment for detail schmatic. And we capture the 5V (PC side & fixture power supply side) during test, there is nothing abnormal.
8081.Mac_TI_USB2_Controller_Board_Rev2.0.pdf
,
Links：
Can customer connect R706 directly to PP5V_USB from tyep-C connector?
,
Lingnan Zeng：
Thank you for your assistance.The DOE have been done on site, but nothing improve. There still exist a disconnection for 6s. In addition, we have purchased the Demo Board (TUSB4041PAPEVM), and programmed the external eeprom to keep it in line with the designed board.The result is same. So I think the issue isn't cause by circuit designed. Is there exist any IC incompatibility problem with Mac mini? Just reupload the USB raw data, can we get any information from it?
