USB hosts hubs transceivers Forum Discussions

Hi 

our system is NVIDIA Jetson Xavier NX

 when USB3.0 SSD unplug form Jeston " native USB3 port", OS can show unplug device information.

but when USB3.0 SSD unplug form CYUSB3314-88LTXI USB3.0 Hub,  OS have to wait a long time and then show the unplug  device information.

can I fix the issue?, I want CYUSB3314 have the same behavior as native port.

您好，

目前使用CYUSB3304-68LTXI

进行设计，已经上板进行测试，在做高低温实验时，发现在-40度出现USB报错，当温度回到20度时，USB又恢复正常。请问这是由什么原因造成的，谢谢

Hello,

I have a question about the operation of DSx_PWREN ( or PWR_EN) in CYUSB33xx and CYUSB230x.

When is DSx_PWREN ( or PWR_EN) turned On/Off ? 

What is the trigger for DSx_PWREN ( or PWR_EN) to turn ON / OFF?

Is it related to the timing of connecting and disconnecting the USB host?

We are use HX2VL USB2.0 Hub on our system.

But our system has low power requirement for USB ports.

Does HX2VL meet below mention of LPM & BESL support?

Thanks,

Mitchell

I once asked the following question.
https://community.infineon.com/t5/USB-hosts-hubs-transceivers/Signal-quality-measurement-in-HX2LP/m-p/322491

After confirming the test procedure of HS HOST Signal Quality with Tektronix, we received a reply that it is necessary to prepare the DUT so that it can be set to Test Mode by itself.

The HX2VL has a TEST pin, but the HX2LP does not.
How can the CY7C65630 go into test mode by itself without the TEST pin?

Thanks,
Tetsuo

Hello,

I have essentially the same question that was addressed in this post: CY7C65631 linux compatiblity and few questions.

It has a disposition of solved, but I am not finding the answer in the link to the solution.  The message board would not let me reply directly to that post, so I am re-posting here.

Specifically, I need to know how to perform the in-circuit programming of the EEPROM on an embedded linux target.  I have the CY7C65631 device designed in, with the recommended 25LC040T/SN EEPROM, and I am working now to figure out how to get it programmed.

Here is the response from Sananya:

----------------------------------------------------

Hello,

HX2LP can work on a Linux host since it needs standard hub class drivers which are present with Linux.

The EEPROM doesnt have to be pre-programmed as HX2LP is capable of programming the EEPROM using Blaster utility. You could also eliminate the EEPROM since it has a ROM with default configuration that it can use for boot-up. However, please ensure that the SPI data line has the required pull-down resistor even when not connected. Please refer to 25LC040 which is used in the HX2LP CY4606 DVK.

Yes, an oscillator can also be used instead of the crystal. You can refer to the App Note for clock requirements-https://www.cypress.com/documentation/application-notes-obsolete/an49150-schematic-and-layout-review-checkli...

Best Regards,
Sananya

----------------------------------------------------

The Overcurrent detection in the CY7C65630 datasheet has the following description:
After detecting an overcurrent condition, hub reports overcurrent condition to the host and disables the PWR # output to the external power device.

Does this "disable" go from Low to High?
Or, go from Low to Hi-z?

Also, is a pull-up resistor required in addition to the inverter? If yes, what is the resistance value? I don't understand when this pin output Hi-z state.

Since the opposite connection destination of the PWR is active HIGH, I would like to know if it works properly by inserting the inverter.

Thanks,
Tetsuo

Hello,

We are currently in development of a product implementing HX3 CYUSB3304 USB hub. We configured the hub as I2C slave to be configured from an external I2C master according to this article: https://community.infineon.com/t5/Knowledge-Base-Articles/EZ-USB-HX3-I2C-Slave-Mode-Operation-KBA90943/ta-p/248113 We do not use any EEPROM.

We generated a suitable configuration file using HX3 Blaster configuration tool and are sending the data in 0xD4 format all in one write access.

As we write to the hub (starting with 0xC0 slave address, 0x0000 memory address, 0x4359... config data) the hub acknowledges every byte. However at the end of the transaction the hub is still not enumerating. We have to send further (between 5-10 individual write attempts) in format 0xC00000 (so no actual data being written) until at some point the hub does no longer acknowledge the byte and successfully enumerates. If we do not send these later bytes even after few minutes the hub still does not enumerate.

Is this intended behavior? Or is there anything wrong here?

Thanks!

Best Regards

Is it a problem if the absolute maximum rating of input voltage -0.5V is violated even for a short time?

The undershoot of the input signal of a pin is large, and the time is about 1 ns, which is about -0.8V. I would like to know if measures are needed.

Thanks,
Tetsuo