SOLVED: Problems negotiating with xs1-u6a-64-fb96-c5

Technical questions regarding the XTC tools and programming with XMOS.
bafiogino
Member
Posts: 8
Joined: Fri Jan 03, 2020 3:26 pm

SOLVED: Problems negotiating with xs1-u6a-64-fb96-c5

Post by bafiogino »

Hello there,

I'm using USB Audio Software Version 6.15.2 with xTime Composer tool v14.1.2 with xs1-l8a and xs1-u6a.
Everything worked fine with my application when I used the xs1-l8a-64-tq48-c5 (TQFN package with external USB PHY).

In the last year, having difficulty recovering the xs1-l8a-64-tq48-c5, I redesigned the project using i xs1-u6a-64-fb96-c5 (BGA package with internal USB PHY).
The prototype with the new xs1-u6a-64-fb96-c5 works correctly without problems.
I produced my first 100 boards and I found 30 boards which were not working correctly!
When I connect the USB cable to the USB port of the board:
- sometimes the negotiation with the computer not work
- sometimes the device is not recognized or sometime the sample rate is not valid
- sometimes it works without problem
When it does not work I can just remove the USB cable, reconnect it and it all works.
It seems to be a problem with the USB PHY, but in the xs1-u6a-64-fb96 it is internal and I can't replace it so I tried to replace the xMOS but I still have the same problem.
Naturally I tested the board with different computers, but the result is the same.

The project is the app_usb_aud_xk_u8_2c, I just added my VID and PID.
The module_xud is the sc_xud v2.4.1.
I bought the chips from Digikey so I assume they are all good.

Can someone help me?

Regards,
Bafiogino
Last edited by bafiogino on Fri Dec 16, 2022 8:27 am, edited 1 time in total.


RitchRock
XCore Addict
Posts: 186
Joined: Tue Jan 17, 2017 9:25 pm

Post by RitchRock »

Are you sure it isn't a hardware problem with the USB jack and interface components? Would it be worth us looking at the schematic / PCB?
bafiogino
Member
Posts: 8
Joined: Fri Jan 03, 2020 3:26 pm

Post by bafiogino »

Thanks for the reply,

I followed the reference design of xMOS, as I wrote, the design works with TQFN version but with BGA version there are some problem with negotiation.
I enclose the schematic and the PCB of the usb interface

Regards,
Bafiogino
Attachments
usb_interface.pdf
(322 KiB) Downloaded 127 times
usb_interface.pdf
(322 KiB) Downloaded 127 times
pcb.pdf
(176.26 KiB) Downloaded 123 times
pcb.pdf
(176.26 KiB) Downloaded 123 times
RitchRock
XCore Addict
Posts: 186
Joined: Tue Jan 17, 2017 9:25 pm

Post by RitchRock »

What is the P/N of the USB jack at J4? Does this jack connect via wires soldered to the board or is it on the bottom out of view in the PDF?
bafiogino
Member
Posts: 8
Joined: Fri Jan 03, 2020 3:26 pm

Post by bafiogino »

J4 is a DIL VERTICAL PIN HEADER (Mouser code 855-M20-9723645), to which a cable with USB type A port is connected.
I have already used this solution for the production of 500 cards (TQFN version) without having any problems.
RitchRock
XCore Addict
Posts: 186
Joined: Tue Jan 17, 2017 9:25 pm

Post by RitchRock »

What is your P/N for Y1?
bafiogino
Member
Posts: 8
Joined: Fri Jan 03, 2020 3:26 pm

Post by bafiogino »

Y1 is a quartz with 24MHz of frequency, tolerance of 50ppm, Capacitance 12pF, Operating temperature -40+85°C
RitchRock
XCore Addict
Posts: 186
Joined: Tue Jan 17, 2017 9:25 pm

Post by RitchRock »

I notice the Mode 1 pin is not connected. In the data sheet on page 16 it looks like this should be connected to logic high for an oscillator frequency of 24 MHz. Further, it states that the mode pins must be held at a static value during and after desertion of system reset.
bafiogino
Member
Posts: 8
Joined: Fri Jan 03, 2020 3:26 pm

Post by bafiogino »

Of course, but at page 8 in the Signal Description section, it is indicates that the signals MODE[3..0] have an internal Pull-Up, so to configure the 24MHz I just connect MODE 0 pin to GND and leave MODE 1 not connected because there is a internal pull-up.
Is this wrong?
RitchRock
XCore Addict
Posts: 186
Joined: Tue Jan 17, 2017 9:25 pm

Post by RitchRock »

I read that also. I think someone from XMOS should chime in here. I don't know why they would tell us the mode pins must be held static during and after a system reset - maybe the internal pull-up is too weak? If you have access to that port you could try set_port_pull_up(). Your schematic does match XMOS on this as I'm sure you are aware.

The other thing I'm noticing is that you have M1, M2 and H1 tied to 3.3V. XMOS has these tied to 5V for USB powered designs. Can you modify these to attach to Vbus? The only schematic XMOS offers where these pins are tied to 3.3V is where the entire system is powered by 3.3V and Vbus is not present. On page 35 of the datasheet it states that in 5V Mode (perhaps this is when Vbus is connected), Vsup needs to be between 4.5v and 5.5V.
Post Reply