URGENT: problem with bare metal access to SE050 edge lock

[Sarasara]

We are writing a bare metal application, and we need to communicate with Edge Lock #SE050 from a processor #LS1043A.

Following the document UM1225 we send some supervisory command (prologue, S-block, epilogue) and the component responds as expected. When we send a Get ATR bytes request, we receive ATR bytes.

When we pass to send a command for the APDU layer the component responds with a very long sequence of zeros. We try to perform the command for get applet info bytes following the document AN12413 and using the same rules for prologue and epilogue (using an I-Block this time). I attach to this post: the bytes sent to the component, the electric scheme and the screenshot of the documentation. So, the component responds correctly at low level but does not respond at APDU level.

There is some hardware check to perform or other APDU commands to send or something at an intermediate level to understand the cause of the problem?

[rodolfoveltrigo]

@Sarasara @ginogatto @petervanackeren 

Hi Sarasara,

i have inspected your schematic and i made some comments:

1) First i don't see pull up resistors to I2C_SCL and I2C_SDA. Maybe they are already connected somewhere. Please check if you have 3.3kohm connected to VCC. 

2) I would add also a capacitor (100nF) between pin 18 (VCC) and pin 15 (VOUT)

3) I noticed pin 14 (ISO_7816_RST_N) is connected to Ground. From this document https://www.nxp.com/docs/en/application-note/AN13539.pdf (paragraph 4.4, page 12) it is explained: The EdgeLock SE05x reset pin does not apply for the I2C interface. Actually in table 7, it shows that this signal is not connected. So, i would try to disconnect it from ground and leave it NC.

4) consider using our SE05x Middleware which you can download from here (if not yet done): https://www.nxp.com/webapp/Download?colCode=SE05x-PLUG-TRUST-MW&appType=license 

5) My colleague Peter Van Ackeren (peter.vanackeren@nxp.com) in the past has successfully connected OM-SE050ARD demoboard to a Layerscape evaluation board (i don't recall which P/N). In case suggestions 1) , 2) and 3) are still not working, send a short email to him so that he may share a non-NDA bonded schematic.

Your comments are welcome

Cheers

Rodolfo