[time-nuts] (OT) HP 16500B / 16500L - reprogramming MAC address

[Philip Pemberton]

On 18/02/11 21:50, Hal Murray wrote:
> My guess is something is broken rather than the chip has been wiped.  Have
> you checked the chip select or data path?

Seems unlikely. The board hasn't been touched. The chip seems to connect 
directly to the bus connector (some weird high-density IDC thing).

> If it's the MAC address, it's unique to each board.  Do you remember the old
> one?

It's printed on the bottom of the board.

> In the early Ethernet days, the MAC address was typically stored in a small
> ROM.  Since even the smallest ROM chips available had lots of room they added
> some other stuff.  A description of the format is probably available online
> and/or you can probably decode it from some driver software that reads the
> ROM.

Um, this thing runs custom software, a custom RTOS... custom everything. 
If it ran (say) uCOS or CP/M with a standard Ethernet driver and stack, 
then maybe it'd be doable.

> I doubt if there is a simple way.  Why would you want to do that?  The idea
> is that each board should have a unique MAC address.  You want to make it
> hard for idiots to do things like copy an address and destroy that uniqueness.

The 1670G (a later HP analyser) allows this -- if the EEPROM gets nuked, 
it'll set the MAC to all zeroes and disable the LAN adapter. You can 
then go into System Settings and a new Factory Options button appears. 
You push that, and it asks you for the MAC address from the LAN card 
sticker.

> The boards typically disabled the write-enable pin.  The chips were
> programmed before getting soldered down.

This one almost certainly wasn't!

Every other programmable device in the analyser has a thermal-transfer 
label (most of them seem to be Kapton-backed, a few are paper). This 
parallel EEPROM is the only one which doesn't have a label -- at all. 
Just the manufacturer branding...

-- 
Phil.
lists at philpem.me.uk
http://www.philpem.me.uk/