The boundary scan test of AM4376 is error

Hi,

I will ask the factory team about this.

Hi,Biser

Thank you for your quick reply.

What do you think about the current situation regarding this matter?

Best Regards,

Shigehiro Tsuda

I have notified the factory team. Any feedback from them will be posted directly here.

Hi,

Updates to the BSDL file are only provided if something in the chain has changed other then the revision of the part. If only the revision has changed then the current BSDL can be modified by the customer with the new revision ID. The BSDL contains the following IDCODE definition.

attribute IDCODE_REGISTER of AM437x: entity is

"0000" & -- Version number
"1011100110001100" & -- Part number -- pin package ID
"00000010111" & -- Manufacturer ID -- Texas Instruments
"1"; -- Required by IEEE Std.

This is described in the technical reference manual for the part in section 7.3.1.16 CTRL_DEVICE_ID Register. Note that only the version number will ever change. The part number, manufacturers ID and the required trailing 1 should always remain the same between revisions. 

The errata document describes the differences between the revisions of the device. Table 2 provides the Silicon Revision Variables for each revision. Silicon revision 1.2 specifies a device revision of 0002b (sic) note that the formatting is incorrect. Since this is a binary number it should be 0010b but 2 is the correct code. '0010' should be substituted into your BSDL file for the Version number shown above. Once this is done your ID should match you scan.

Regards, 

Bill

Hi Bill,

Thank you for your reply.

Is it right by the following corrections?

[default　]

attribute IDCODE_REGISTER of AM437x: entity is
"0000" & -- Version number
"1011100110001100" & -- Part number -- pin package ID
"00000010111" & -- Manufacturer ID -- Texas Instruments
"1"; -- Required by IEEE Std.

[modification ]

attribute IDCODE_REGISTER of AM437x: entity is
"0010" & -- Version number
"1011100110001100" & -- Part number -- pin package ID
"00000010111" & -- Manufacturer ID -- Texas Instruments
"1"; -- Required by IEEE Std.

I can try this.

Best Regards,

Shigehiro Tsuda

Hi Shigehiro,

That should be correct. Let me know if you have any other problems.

Regards,

Bill

Hi Bill,

Thank you for your quick reply.

There is a request from our customer to want new BSDL file of am4378.
Can you have you update it?
In addition, the following problems occur newly, too.
There are SYNTAX_ERROR and the pin number, the disagreement of the pin name such as follows.

[default data]
reserved: linkage bit (18 downto 0),
reserved2: in bit;


[modified data]
reserved: linkage bit_vector(18 downto 0), ← error in writing
reserved1: in bit; ← reserved2 does not exist.

reserved is 19 pins, but is 28 pins with the data sheet of the device.
Only reserved _ 28 assigns VPP in BSDL.
There is no description of 8 remainder pins.
The pin number of reserved1 is AE10, but leads to GND in VSS_5 in EVM.

This BSDL file is one with many misprints, and an opinion whether is worth reliability appears in our customers.

Because there may be it other than the above, will you confirm it about the contents of the file again?

Best Regards,

Shigehiro Tsuda

Hi Shigehiro,

I can see that errors that you are referring to. I have posted a question to the product engineering team to see if I can get the latest version of the BSDL from them. Once I have that I will try and resolve the issues and provide an update.

Regards,

Bill

Hi Shigehiro,
Could you tell me which type test soft and emulator you used to carry out a boundary scan test?Can CCS implement a boundary scan test?Have you implement a boundary scan test with cjtag?

Best Regards.

Hi Bill,

Thank you for your quick reply.

 We wait for BSDL file to be updated newly.

Because our customer wants to complete the design of the product board,

please give a response immediate as possible.

Best Regards,

Shigehiro Tsuda

Hi Shigehiro,

I have a new version of the BSDL file which will soon be published on TI.com. This file should be correct.

Regards,

Bill

sprm635A.zip

Hi Bill,

Thank you for your reply.
I'll talk with our customer and confirm whether it works in this normally to our customer.

I thank for your correspondence.

Best Regards,

Shigehiro Tsuda

Hi Bill,

Thank you for your correspondence.
After confirming new BSDL file, the following does not seem to be revised.

the following problems occur newly, too.
There are SYNTAX_ERROR and the pin number, the disagreement of the pin name such as follows.

[old BSDL data]
reserved: linkage bit (18 downto 0),
reserved 2:in bit

[new BSDL data]
reserved: linkage bit_vector(18 downto 0), ← error in writing
reserved 1:in bit

reserved is 19 pins, but is 28 pins with the data sheet of the device.
Only reserved _ 28 assigns VPP in BSDL.
There is no description of 8 remainder pins.
The pin number of reserved1 is AE10, but leads to GND in VSS_5 in EVM.

[Probably correct data]
reserved: linkage bit_vector(27 downto 0),
"AB6,AB7,AC5,AC7,AE9,H19,H21,P21" is not defined.

Would you confirm it?

Best Regards,
Shigehiro Tsuda

Hi Bill,

There is a request from our customer to want new BSDL file of am4378.
Can you have you update it?
In addition, the following problems occur newly, too.
There are SYNTAX_ERROR and the pin number, the disagreement of the pin name such as follows.

BSDL file:

"VDDSHV2           : (V16, V17, W17)       ,"&

The Datasheet of AM437x:

VDDSHV2 is not W17 pin but  is W16 pin.

Which is right with BSDL and a data sheet?
Probably I think that a data sheet is right.

Best Regards,

Shigehiro Tsuda

Hi Shigehiro,
You are correct that VDDSHV2 should be W16 as shown in the data sheet. There are also three other signals that have slightly different names then those that appear in the data manual but I don't see any syntax errors in the latest release. Can you tell me what the error was?
Regards,
Bill

Hi Bill,

Thank you for our quick reply.
I understand that VDDSHV2 is W16.

The following contents are the error.
reserved: linkage bit_vector(18 downto 0)
"reserved : (AA10, AA7, AA9, AB10, AB9, AC10, AC12, AC6, AC9, AD1, AD2, AD10, AD11, AD7, AE11, AE12, W10, Y10, Y7) ,"&

The reserved pin of the datasheet of AM4378 is the following.
(AA10, AA7, AA9,AB6,AB7, AB10, AB9, AC10, AC12,AC5, AC6, AC7,AC9, AD1, AD2, AD10, AD11, AD7, AE11, AE12,AE9, H19,H21,P21,W10, Y10, Y6,Y7)”

The BSDL file is not defined in pins "AB6,AB7,AC5,AC7,AE9,H19,H21"
and P21 pin is defined "VPP" .
"VPP : P21 ,"&

Would you understand the contents of the error?

Best Regards,
Shigehiro Tsuda

Hi Bill,

Thank you for quick reply.

We think that the BSDL file should define all pins for a boundary scan test, is it a mistake?

Will these pins be to have nothing to do with a boundary test?

Because we trouble our customer, will you reconfirm whether it must be a pin name and a description as well as our indication matter by a comparison between the data manual and the BSDL file?

And we want you to update the BSDL file like the last time.
Is it possible?

Best Regards,

Shigehiro Tsuda

Hi Bill,

Thank you for your correspondence.
Because there is a demand from our customer, I want updated new BSDL file.

Is it possible?

Best Regards,

Shigehiro Tsuda

Hi Bill,

Thank you for your correspondence.
I am sorry many times.
Because there is a demand from our customer, I want updated new BSDL file.

Is it possible?

Best Regards,

Shigehiro Tsuda

Hi Bill,

Thank you for attaching new BSDL file.
I confirm whether our customer does not have any problem with this file.

Thank you for your correspondence.

Best Regards,

Shigehiro Tsuda

Hi Bill,

There was the answer to an existing error having been broken off with a file of BSDL attached by our customer.
Thank you for correspondence in various ways.
But it does not seem to have been yet made file update when I confirm web of AM335x.
Can you have web of AM335x update it?

Best Regards,
Shigehiro Tsuda

Hi Shigehiro,

Can you confirm if the customer was able to use the attached version of the BSDL file? I wanted to be sure it was correct before releasing it on TI.com.

Regards,

Bill

Hi Bill,

Thank you for quick reply.

Yes,Our customer was able to confirm that a boundary scan worked using an attached version normally.

Please release it on AM437x web site.

Best Regards,

Shigehiro Tsuda