USB superspeed peripherals Forum Discussions

Hello, I'm designing a custom PCB with FX3 to convert 8/14 bit DVP to UVC. Here I'm attaching the schematics for the same.
With the help " Cypress FX3 CYUSB3KIT-003" I was able to verify the working of 8 bit. But the VIO voltage was 3.3V on CYUSB3KIT-003 and my monochrome sensor works at 2.5V, so i have to use a level shifter.

But the issue is, i want to get rid of the level shifter and want to directly interface the sensors 2.5V IO. I modified the power supply on evk kit to make 2.5V for FX3 VIO. After that, I'm not able to get streaming. Is there anything else I need to do on the Software side/Hardware side to get FX3 to work on the 2.5V level?

In my GPIF state machine, I want to compare the 2bit address bus to a constant value and then transition state based on that comparison. Reading the docs, it sounds like I need to use CMP_ADDR however I am unsure how this command, and the CMP_DATA and CMP_CTRL, families work. I have a few questions.

1. Are ADDR, DATA and CTRL in these commands simply counter registers or are they connected to the address, data and control busses respectively? If the ADDR commands are connected to the address bus, DATA to the data bus and CTRL to the control bus, how are these commands affected when I am actively using these components?

2. What does LD_ADDR_COUNT, LD_DATA_COUNT and LD_CTRL_COUNT do exactly? If they are simple counters, I imagine they would simply be configuring those counters but if they are connected to the respective components (as suggested by CMP_ADDR), what are they loading because I am actively using the address bus.

3. What are the CMP commands comparing against? I know the command allows me to compare ADDR, DATA, and CTRL against a constant but where are the values of ADDR, DATA, and CTRL taken from? If I don't initialize the ADDR counter, will CMP_ADDR compare the GPIF address bus against the constant? Is there any way to make a transition based on the current value of the GPIF address bus?

4. I am trying to use one of these counters to keep track of which address was most recently used. Is this possible using any of the ADDR, DATA or CTRL family of commands? I dont want any other commands to overwrite the counter I am using.

Thanks,

I've posted the below query on getting the Monochrome Camera working with FX3 and which is got resolved also.

How-to-get-Monochrome-Camera-stream-working-on-FX3-using-AN75779 

We are able to get streaming working on Linux Machine using both FFMPEG and GStreamer Pipelines.

Now, we wanted to get the camera stream working on Windows Laptop.

We first tried to read the frame using the gstreamer from a CLI as follows: gst-launch-1.0 ksvideosrc device-name=FX3 ! video/x-raw, width=640, height=480, format=GRAY8 ! videoconvert ! autovideosink.

This resulted in the following error:

We then tried to stream using an opencv Videocapture application. In this case, it was not ready to open the device itself. At this point I don't know what else I can do. Could anyone please help.??

Looking forward to a positive response.

Hi Support,

I have a technical question regarding the USB controller (CYUSB3014).
A loopback function is required for USB communication evaluation,
Is there any firmware or setting method for loopback?

The application note states:
9.3.1 USB BULK DATA LOOPBACK EXAMPLE
https://www.infineon.com/dgdl/Infineon-AN75705_Getting_Started_with_EZ-USB_FX3-ApplicationNotes-v11_00-EN.pdf?fileId=8ac78c8c7cdc391c017d073986235e0f

Best regards,

Koki

Our device uses the FX3 controller chip, running in Windows 10.

The device always comes up with "Cypress FX3 USB BootLoader Device" in Device Manger after system boot up.

Lately, many installations reported seeing "DTU device with FX3" in Device Manager after Windows update.

How do we stop this from happening?

Hello everyone,

I am developing a USB3.0 peripheral using FX3. I am currently facing an issue where sometimes USB3.0 is recognized as USB2.0. Is there any way to debug where the problem is? I am using Ubuntu system, but I cannot find any relevant logs in the system log and USB firmware.

Hi,

I have trouble using the GPIF machine of my . Over the endpoint, some packets are lost, and some packets are not sent chronologically

My system uses 3 endpoints, 2 for command exchange with the host, and 1 for data streaming. I also use actively the available UART, using a manual DMA IN and manual DMA OUT.

I have been using my command endpoints fine, and have no known problem using them.
The data streaming endpoint is used with the GPIF machine.

I have a FPGA that sends data over two ports of the GPIF machine, that the GPIF machine send over the USB endpoint.
Data is synchronized using 2 signal : IP, that indicates a packet start, and EP, that indicates a packet end.
See below the GPIF machine I use.

About the configuration part of the Fx3 firmware, I use a multiDmaChannel and a callback to handle data. See below the code snippet that describes both functions.

DMA configuration :

dmaMultiConfig.size = 16384;
dmaMultiConfig.count = 4;
dmaMultiConfig.validSckCount = 2;
dmaMultiConfig.prodSckId [0] = (CyU3PDmaSocketId_t)CY_U3P_PIB_SOCKET_0;
dmaMultiConfig.prodSckId [1] = (CyU3PDmaSocketId_t)CY_U3P_PIB_SOCKET_1;
dmaMultiConfig.consSckId [0] = (CyU3PDmaSocketId_t)(CY_U3P_UIB_SOCKET_CONS_0 | (CY_FX_EP_DSI_STREAM & 0x0F));
dmaMultiConfig.prodAvailCount = 0;
dmaMultiConfig.prodHeader = 0;
dmaMultiConfig.prodFooter = 0;
dmaMultiConfig.consHeader = 0;
dmaMultiConfig.dmaMode = CY_U3P_DMA_MODE_BYTE;
dmaMultiConfig.notification = CY_U3P_DMA_CB_PROD_EVENT;
dmaMultiConfig.cb = CyFxUvcApplnDmaCallback;
apiRetStatus = CyU3PDmaMultiChannelCreate (&glChHandleUVCStream, CY_U3P_DMA_TYPE_AUTO_MANY_TO_ONE, &dmaMultiConfig);

DMA callback :

void
CyFxUvcApplnDmaCallback (
CyU3PDmaMultiChannel *chHandle,
CyU3PDmaCbType_t type,
CyU3PDmaCBInput_t *input
)
{
CyU3PDmaBuffer_t dmaBuffer;
CyU3PReturnStatus_t status = CY_U3P_SUCCESS;

if (type == CY_U3P_DMA_CB_PROD_EVENT)
{
/* This is a produce event notification to the CPU. This notification is received upon reception of
* every buffer. The buffer will not be sent out unless it is explicitly committed. The call shall fail
* if there is a bus reset / usb disconnect or if there is any application error.
*/


glUVcProdEvent++;


/* There is a possibility that CyU3PDmaMultiChannelGetBuffer will return CY_U3P_ERROR_INVALID_SEQUENCE here.
* In such a case, do nothing. We make up for this missed produce event by making repeated commit actions
* in subsequent produce event callbacks.
*/
status = CyU3PDmaMultiChannelGetBuffer (chHandle, &dmaBuffer, CYU3P_NO_WAIT);

if (status != CY_U3P_SUCCESS)
{
glUVcGetBufferKo++;

nbSuccErrors++;
}
else
{
nbSuccErrors = 0;
}

if (nbSuccErrors == 2)
{
CyU3PEventSet(&glEvent, FX3_DMA_RESET_EVENT,CYU3P_EVENT_OR);
nbSuccErrors = 0;
}

while (status == CY_U3P_SUCCESS)
{

glUVcGetBufferOk++;

status = CyU3PDmaMultiChannelCommitBuffer (chHandle, dmaBuffer.count, 0);

if (status == CY_U3P_SUCCESS)
{
glUvcCommitBufferOk++;
}
else
{
CyU3PEventSet(&glEvent, FX3_DMA_RESET_EVENT,CYU3P_EVENT_OR);

glUvcCommitBufferKo++;
break;
}

/* Check if any more buffers are ready to go, and commit them here. */
status = CyU3PDmaMultiChannelGetBuffer (chHandle, &dmaBuffer, CYU3P_NO_WAIT);
}
}

}

Using this configuration I try streaming with the host, at low speed, low quantity of data ( < 10 KB/s)

Typically, my FPGA send a Leader packet, then a Payload Packet, then a Trailer Packet. I'm 100% sure packets are sent from FPGA in the right order, and that they are sent.

Using wireshark to capture packet, see the results I have :

Correct packets:

Too fast for GPIF packets:

So here comes my question : Do you have any clue as to why my packets are being scrambled and lost,  and how can I solve this issue ?

Thanks !
Bambo

Is there any reference code  about UAS?
Currently I'm studying for USB msc with EZ USB FX3 board, but it's not easy to find any information about UAS.
I saw a comment from 2 years ago that there is no example code for UAS.
But in the meantime, if someone has experience implementing it, please help me.