AWR2944: Outdated BSDL file?

Please give us until tomorrow to get back on this.

Regards,

Aydin 

Hello,

The version number is defined as 0001 in the file on the web. Are you sure you are using the latest version?

Regards,

Adrian

Hello,

I read 0010 for Version Number from the device.

The full response I receive on TDO from the device for ID Code readout is 11110100000011011110110111010100.

Thanks,

Balázs

I attached two screenshot. I marked the bits with red squares which are failing compared to the expected values generíted by the Keysight compiler based on the BSDL file.

IDCODE readout and PRELOAD:

EXTEST:

Can you please help me out as soon as possible?

Thank you very much!

Hello,

I am checking on this with our digital systems team. I will get back to you with an update as soon as possible.

Regards,

Adrian

Hello,

Thanks for the support in advance. I'll wait for your feedback.

Balazs

Hello,

Regarding version number, the correct version number is "0010" since this is ES2.0 silicon. This means that the BSDL model needs to be updated to accurately reflect the ES2.0 version number.

I am still awaiting feedback on the other errors you are facing.

Regards,

Adrian 

Hi,

Where can I find this other BSDL file? I'd like to try it out on monday morning.

Thanks,

Balazs

Hello,

Can you please give me an estimation about the new BSDL file? I need to report the progress to our customer.

Thanks,

Balazs

Hello,

Sorry for the delay. The team is working on correcting it. Let me check and see when it will be available.

Regards,

Adrian

Hello,

The team said they would have the updated model on Nov 13.

Regards,

Adrian

Hi,

Can you please give me a status update? Thank you.

Regards,

Balazs

Hello,

I just received the updated model, please see attached.

https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/1023/awr29xx_5F00_2p0.bsdl

Regards,

Adrian

Hi,

Only changing the Version number in the bsdl does not sole the issue with boundary scan. We tried this moths ago.

Lets focus on the IDcode first. Why do I get four 1 bits when they should be 0 according to bsdl?

attribute INSTRUCTION_CAPTURE of awr29xx : entity is "00000001";

My guess that the X bits are not handled correctly. Is this a design error?

Thanks,

Balazs

Hi,

I did some further analysis today. I checked the TDO responses for IDCODE instruction in a checkerboard pattern.
1101XXXX -> 11010101, 11011010
It looks like to me thet the INSTRUCTION_CAPTURE length is not correct in the device, only 4 bits instead of 8. 

Please check if my finding is a normal behaviour compared to jtag standard.

Regards,

Balazs

Hello,

Yes, I am frustrated because had highlighted this to the design team but it looks like they ignored this and only fixed the silicon revision issue.

I have requested them to look into this further. 

Regards,

Adrian

Hi,

I captured the communication on a bench programming station for this device in our lab and i am 100% sure that this device uses 4 bit instructions. As I see on that station every instruction is 4 bit wide. Can you please escalate this topic on a higher level? The SMH Technologies team who made the driver for this station figured it out somehow.