Quote:
Is it possible to order a motherboard, fully similar just like the motherboard of the original HP15C be put into place, just a faster CPU, etc...
Obviously to replace a legacy voyager PCB you'll need to address
severing the heat stake heads and securely re-cementing new ABS
heads. That's a substantial drawback for prospective users.
Next in line is the issue of reuse and swapping of
the display as an economic concession against requiring a
replacement display. Release of the stainless compression
frame and accurately reseating both it and the zebras on the
new PCB is a bit involved.
Aside from this either a CPU with integrated
LCD controller or addition of an external controller
is realistically required. Either is
preferable to attempting reuse of the existing R2D2 controller.
Then there is the issue of the nominal 4V5 battery rail.
The only SoCs of which I'm aware, able to
operate natively from 4V5 down to approx.
3V are lower functionality architectures. So to accommodate
say an ARM core will require at least an ultra low power linear
regulator to provide data retention or a systems approach where
all non-volatile data is saved to EEPROM/FRAM on power down.
In event the SoC itself is powered down, external circuitry
on the battery rail will be needed to effect a power on wakeup.
Most of the above have relatively straightforward solutions if
approached from a "labour of love" mindset. But otherwise there
are detractions to the effort: you probably won't create
enough overall interest to establish a self-supporting user
community and you'd be drawing from a diminishing and
unpredictable supply of repurposeable units.
The KINOMI module approach addresses the technical
issues above and as well your memory/performance goals, but
introduces its own requirement of a specialized desoldering
process to remove the original NUT cpu and swap in the upgrade
PCB without inflicting heat damage to the LCD zebras. That
process requires someone with specialized tooling and rework
skills to accomplish, who can justify providing the service
for a yet unknown number of potential customers.
Revisiting the prospect of a wholesale board swap, I came
to the conclusion it would only make sense if a display
upgrade was possible. As such
there was another variation in the KEMU/KINOMI project which
replaced the 7-segment display with a graphic alternative,
adding features intended as a partial solution for the
static keyboard layout. You can find details
here.
I don't intend to discourage your interest. But rather wanted to
share my observations after considering the above possibilities.
