USB Forum Discussions

Hello,

I tested the video output on the custom board with the firmware provided by the kba below.

In order to exclude and test the image sensor, we tested it with data from fpga, which outputs a virtual pseudo image, and it worked well on the existing AN75779 project.

https://community.infineon.com/t5/Knowledge-Base-Articles/Using-CyU3PDebugPrint-API-to-Send-Debug-Messages-over-USB-CDC-Interface-from-FX3/ta-p/251458

Problem:

In a project with that CDC interface, we found that the image did not come up until we opened the COM port on the CDC interface.

(Image1 : without open com port)

(image2 : video capture with opening com port)

Can you tell me why this happens, and how to fix it?
Below is a log that is output when I open a port without video output, as in Image 1.

Additionally) When the DebugInit() function is annotated, video is output immediately without the issue.

It seems to be related to the code below.

Thanks.

I am using the FX3 KIT to implement a camera interface using the code found in AN75779. Can someone explain what it means to be told that a "DMA Reset Event: Frame Timer Overflow" has occured? Does it mean I should increase the Frame rate or reduce the frame rate?

The BT601 format data of the f 5640 module has been successfully debugged according to the 《Infineon-AN75779 _How_to_Implement_an_Image_Sensor_Interface_with_EZ-USB_FX3 _in_a_USB_Video_Class_(UVC)_Framework-ApplicationNotes-v13_00-CN.pdf》document and the "cyfxuvc_an75779.zip" code. Is there any reference material for the f 5640 module BT656 format data to fx3 ?

Hello.

I'd like to know detail HPI register about BCR-Plus.

Below is HPI register in BCR HPI Utility user guide. 

I don't find detail document. Let me know if anyone know link or data.

Regards.

Robert.

Hi,

Hello Infineon!

I've implemented a FIDO CTAP 2.1 interface on an FX3s but I am getting very strange behaviour with it... I've posted in the FIDO community google group but I don't think I'll get a response there... Any thoughts/ideas would be most appreciated!

To summarize:

1. Driver and report descriptors work as expected
2. HID Communication works

The wierdness:
Situation 1 (working situation)
1. Load up a test page for FIDO (Chrome browser, webauthn.io or other tool, all have same behaviour)
2. Start a FIDO registration process, browser says "insert FIDO device"
3. Program FX3s with FIDO firmware, it registers and handles INIT and GETINFO steps correctly, and sends registration info

Situation 2 (not working situation)
1. Program FX3s with FIDO firmware
2. Load up test page
3. Start FIDO registration, gets stuck in a loop performing INIT and GETINFO until "device fail" message

The part that is really confusing me is during INIT stage, the FX3s must send a new CID value, which the webpage/host must use for future responses, which it DOES use for the getinfo request... This confirms that FX3s -> FIDO HID -> browser and back communication is working successfully in situation 2.....

Is this potentially an issue with chromes implementation of FIDO HID? Is there a way I can "trick" the host with the FX3s to do a reset on a specific USB interface? I do not want to disconnect USB via CyU3PConnectState since it is a composite device and other interfaces may be doing something.

I should also point out there's shouldn't be any real difference in timing between messages sent and received (approx 10~50ms), nor is the content of the replies different. I will be investigating this with wireshark shortly but I am doubtful that a message timeout is the problem since the host should be sending an error if there is one as well.

Regards!

Fujimi

Edit: This only happens on USB3, works as expected on USB2 (Both scenario 1 and 2 work USB2), further edit, responses are always 64 byte in size which is the buffer size so there should be no issues with set wrap up (I think)

I am using the GPIF2 function of FX3 to receive large amounts of data from an FPGA, and then I am transmitting this data to a PC via USB using the Bulk-in path.

During USB Bulk transfers, I am experiencing a delay between I2C packets when using FX3's I2C Master to communicate with the FPGA's I2C Slave.

When not performing data transmission via Bulk, the operation works correctly with very short delays.
However, when using Bulk transfers consecutively, there is a significant delay between I2C packets.

I used FX3's SDK version 1.3.4, and here are the settings for the I2C:

Are there any settings required to simultaneously use GPIF2 and I2C, or if not, please explain what is causing the delay?

Hi,

I am not able to use the CY4500 PD software on MacOS. The software launches but in some sort of corrupted state that leaves it non-functional. It does not detect the analyzer hardware and clicking "Start" from status bar does nothing. I have attached a screenshot.

I am on M1 MacBook Pro, MacOS 13.5.2.

I posted previously about the software not launching at all and was told to install Java 8. I edited the program's Info.plist to force it to use JDK 1.8.0_181 because it was trying to use JDK 10 before, despite using jenv to switch my default Java to 8. This enabled the software to launch but it is still non-functional as described above.

I have also attached log file.