Signature Row Serial Number

Go To Last Post
9 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am using an AT90USB1286 and looking at the datasheet it says

Quote:
AT90USB64/128 includes a unique 10 bytes serial number located in the signature row

In the table (28-6 on page 361) it says:

Quote:
Unique Serial Number From 0x000E to 0x0018

I am a bit confused as I normally specify ranges inclusive of end points but 0x0E to 0x18 (inclusive) is 11 bytes not 10...

So does this mean the last byte is in 0x17?
Or are there really 11 bytes of serial number?

David

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Why not just try it for yourself?

The data sheet shows the SPM commands.
It does not seem to show how to read it via SPI or JTAG.

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Actually, I did read the bytes 0x0e through 0x18 and got reasonable values (i.e. appeared random) and then noticed I was only supposed to get 10 bytes.

I just read a 2nd device and found that the byte in 0x18 was 0x17 for both devices. So I expect that byte 0x18 is NOT serial number.

David

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Frog,

Where do you get the idea that a unique serial number must be a random number?

For a unique serial number identifier it makes no sense for this application.

Stan

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

Where do you get the idea that a unique serial number must be a random number?

For a unique serial number identifier it makes no sense for this application.


Presumably it's anything but random and in production Atmel just blow each successive chip to an incrementing number in sequence. Assuming chips from the same wafer make it all the way through production to adjacent devices in a tube or on a reel I think you might actually expect "siblings" to have very predictable numbers indeed.

Of course, rather tan just sequentially incrementing numbers it could be the date/time down to a fraction of a second of the test jig on the line that does the step where the serial number is programmed - that would make for something fairly unique and also easily traceable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As to my use of "random", I was not implying anything other that in my sample of 1 unit at that time I saw no definite pattern that would distinguish the byte at 0x18 from any of the others to discern if there was indeed an 11th byte of serial number.

My whole issue was with the wording in the table stating that the serial number was bytes "0x000E to 0x0018".

"A to Z" (to me) does not mean "A through Y"

David

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ah-ha. I haad never thought of reading 'other' Signature row bytes !

Sure enough. I tried some chips (sig #0x0E - #0x17):

"J93430" ff 16 03 09   mega328P
"G10235" ff 14 1F 17
"G19721" ff 19 1C 15
"T00000" ff 00 05 04   mega644
"F08302" ff 06 13 13   tiny4313

Note that the first six bytes are human-readable.

I have no idea what anything means. It looks as if the 'old' mega644 chip has an 'unused' serial.

I have often wondered what the meaning of undocumented bytes are. Like Cliff, I would guess that they will be wafer codes and date codes.

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

And I got:

"X24296" 15 02 09 19   90usb1286

David

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Rather than derail this thread with a seriously OT detour, I've started another thread.

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]