66AK2G12: USB Eye pattern test is failing

Hello,

Thank you for the query. 

I am looking into the details of 66AK2G12ABYA100 USB Eye pattern in our app notes and revert back to you. 

Regards 

Rajashri

Hello, 

Is the eye diagram captured on the 66AK2G12ABYA100 evaluation board? If so, please mention which evaluation module is being used?

Regards

Rajashri

Hi Rajashri,

Customer uses their own boards.
All three captures are on their custom boards.

Thanks and regards,
Koichiro Tashiro

Hello,

Please share the custom board schematic to check on the USB section.

Thanks

Rajashri 

Hi Rajashri,

Customer's board is really simple around USB connection.
There are only a common mode filter and diodes.
Please see below diagram.
66AK2G12_USB_schematic.xlsx

All three devices (66AK2G12ABYA100, TMS320DM6446BZWT and 66AK2H06) use the same connections.

Thanks and regards,
Koichiro Tashiro

Hello, 

I am checking internally and will get back to you

Thanks 

Rajashri

Hello Koichiro Tashiro,

Can you please confirm if the same part number of common mode choke and ESDs are used in the previously used board of TMS320DM6446BZWT and 66AK2H06. Also, can you confirm if the connection around the USB section shown in the block diagram is same on the board of TMS320DM6446BZWT and 66AK2H06 . 

Also, please perform these two experiments on the  66AK2G12ABYA100 board.

1) Remove the ESD and capture the eye diagram.

2) Remove the ESD and short the common mode choke. Capture the eye diagram

Please share the results

Regards

Rajashri

Hi Rajashri,

66AK2G12 board design around USB was copied from TMS320DM6446BZWT. Part# are also the same.
66AK2H06 board is the same design, but part# are different as shown below.
Common mode filter
ACM2012-900-2P-T001(TDK) => DLW21SN900HQ2L　(MURATA)
ESD Diode
DT1446-04SO (diodes) -> TPD4E05U06DQAR(TI)

Rajashri Hazarika said:

Also, please perform these two experiments on the  66AK2G12ABYA100 board.

Please find below results. The situation is the same.
66AK2G12_Remove_choke.xlsx

Thanks and regards,
Koichiro Tashiro

Hello Koichiro Tashiro,

Can you please tell what is the length of the USB data lines USB0_Dp and USB0_Dm routed on the customer board.  Also, what is the length mismatch between Dp and Dm lines on the board?

Regards

Rajashri

Hi Rajashri,

Here are the length of each line.
USB0_DP: SoC to the filter is 108.162mm. The filter to USB connector is 5.508mm.
USB0_DM: SoC to the filter is 107.759mm. The filter to USB connector is 5.165mm.

Thanks and regards,
Koichiro Tashiro

Hello Koichiro Tashiro,

I need the following details on the schematic, layout and the board
Board: I don't see the experiment done (asked above) for ESD removed, keeping the choke connected.

While capturing the eye diagram, please use 3m cable and 5m cable.

Schematic:
1)is it in USB host mode or device mode?
2) is the USB connector type a or type b?
3) is the shield ground shorted to board ground via ferrite beat? please check below for your reference

It will be good, if you share the schematic design of the board or the USB section. Block diagram does not help to understand the design 

Layout:

1)How many layers of board it is? 
2)Does the layout of the board consider the following guidelines for USB2.0

Please share the board layout. It will help to understand the board.

Regards

Rajashri

Hi Rajashri,

Schematic:
1) USB is device mode.
2) USB connecter is type B.
3) USB connecter shield is connected to GND via 0-ohm resister.
USB connector GND is connected to GND via ferrite bead.
https://www.murata.com/en-us/products/productdetail?partno=BLM18BD252SN1%23

Layout:
1) 12 Layer PCB
2) These guidelines are met, excepts below as the top and the bottom layer are busy.
When possible, route high-speed differential pair signals on the top or bottom layer of the PCB with an adjacent GND layer.

Please see below figure.

USB_board layout.xlsx

Thanks and regards,
Koichiro Tashiro

Hello Koichiro Tashiro,

I have shared your inputs to the USB expert. Based on their response, I will update you

Regards

Rajashri