开云体育

Dear all,

While building my 3rd hp70k-mms system, I want to assess the interest in
a project idea I've had for the past two years or so: A
*sniffer/injector* for the hp70k module-interconnection protocol
(*msib*). All (hardware/software) open source; this is not a commercial
proposal (apart from e.g. cost-price PCB delivery).

The very rough basic idea is to take a single-width hp70k module
(likely, for-parts-only), strip everything inside apart from the msib
interface and mechanical interfaces, and put inside one or more PCBs that:

* Interface with the msib bus such that 'basic bus transactions' can
be captured;
* Store these 'transactions' into (e.g.) a pcap file on an ssd
connected to a (say) an embedded Raspberry Pi;
* Transfer these 'transactions' over 1 Gbps Ethernet (copper/Fiber
Optic) from the Raspberry Pi;
* Interface over Ethernet both from the front and the rear;
* Import the pcap 'transaction' into Wireshark (on a remote machine);
* Interpret the 'transactions' using an hp70k-msib dissector;
* Perhaps add power monitoring, etc.;
* ...


This would be the basic first-order requirements of the project. In
subsequent phases, injection of messages/transactions on the bus and
running HP calibration (HP-BASIC) software from virtualized containers
could be implemented.

*My questions:*

* Is any project already working on something like this that I can
join (I could not find any)?
* Would this be interesting enough to the amateur hp70k community? Is
there sufficient added value to a tool like this over merely
swapping modules in order to get things to work?
* Does a tool like this have the potential to increase the public
knowledge on the hp70k mms? Could it reveal calibration data? Is it
worth it?
* *Would anyone be willing to help?* I figure that even a first
attempt at a working prototype would take over a year, but various
components of the project (like the pcap format definition,
wireshark dissectors, web-based interfacing, RPi firmware, KiCAD PCB
design of bus interfacing, ...) could be isolated and
designed/implemented in a team-effort... Needless to say this
project would required >=4 committed enthusiast aficionados...


Thanks in advance for your thoughts/vision on this, BR,
Jan - pa3gyf