Hi,

I've got some questions about the HP 15C LE.

Would the HP 15C LE work exactly like a HP 12C LE (except for the labels on the buttons) if you put the firmware of the HP 12C LE on it? Or are there some internal differences?

Since it's an ARM CPU emulating a NUT CPU (afaik), which firmware can you change: only the NUT firmware, or can you change the whole firmware including the NUT emulator part, and e.g. create a firmware that just uses the ARM CPU directly to do stuff?

Does it really only have as little memory as the HP 15C had, or does it in fact have a lot more memory in it, for example to run the emulator, but it's just not accessible since the emulated HP 15C firmware isn't aware of its exixtance? Does a chip of 500 bytes of memory (or how much was it?) even exist today? How much physical memory is in there?

How large is the chance that there'll be released a bugfixed version of the HP 15C LE firmware, e.g. fixing the weird screen flashes that happen if you hold down "SST", the high battery drain when holding down a button, etc...?

Thanks a lot!

Last year one of the forum members purchased a HP-15C LE which was incorrectly loaded with HP-12C firmware at the factory although working fine (as a HP-12C). The person noticed the problem when he tried to take a square root of a number, instead he received the same number back (n button on the 12C). Here is the LINK.

Go to youtube and search uhmgawa. Hp voyager KINOMI and Kinomi Voyager Development may interest you.

HP hasn't split the firmware in separate files so you could decide which ROM image you want to run on the emulator kernel. Instead you just get an image to download to the Atmel processor that is fully integrated. In fact, HP has adapted the emulator to the ROM it emulates for various (technical) regions. One result of this approach is that an older (power consuming) bug in the keyboard handler which had already been fixed for the 12C has crept into the 15C LE firmware.

The Chip, an AT91SAM7L128, has 128KB of flash memory, 2KB of battery backed RAM and 4KB of working memory that gets lost in certain power saving situations, i. e. most of the time. This has a few implications:

- You can only emulate systems with at most 2KB or RAM. All the voyagers fall in this category.

- Flash ROM is sufficient for all needs. I've not seen an image file but I assume, they are much shorter than 128KB.

The ROM of the 15C has not been patched to support more than the original 64 registers (448 bytes). It would have been possible but the emulation isn't designed for it. Some RAM is required to save the emulated CPU state and registers but the 2KB aren't nearly exhausted.

What a shame this person sent it back! If HP never releases the firmware, it's rare and special since nobody can create it by flashing. If HP does release the firmware for flashing some day, then it's not special anymore but then it could be flashed to the correct one and nothing of value was lost (and it could still be rare due to the firmware version).

I really hope a bugfix version will be available some day!

Quote:

---

I really hope a bugfix version will be available some day!

---

Quote:

---

I've got some questions about the HP 15C LE.

Would the HP 15C LE work exactly like a HP 12C LE (except for the labels on the buttons) if you put the firmware of the HP 12C LE on it? Or are there some internal differences?

---

Fundamentally there should be no differences as AFAIK the SoC is
generic Atmel issue. But this does raise an interesting point
as it is possible a board layout difference exists between
the 12c+ and 15c le, where a gpio etc.. is bonded to vdd/gnd
differently between the two board builds. The firmware could
then verify the model specific configuraton before agreeing to
operate. However while I didn't find evidence of that for the
15c le voyager, it is possible to
exist on a COB sam7 only detectable from a practical perspective
by crawling into the unit with diagnostic firmware to poke
around a bit. However given the reported case of the 15c le
mis-flashed with 12c+ firmware, I'd hazard such a contingency
doesn't exist.

HP must have done something to this extent to distinguish between the 20b and 30b. If you load the 30b firmware (which is available through Tim) into a 20b it behaves exactly like a 20b without programmability.

Quote:

---

HP must have done something to this extent to distinguish between the 20b and 30b. If you load the 30b firmware (which is available through Tim) into a 20b it behaves exactly like a 20b without programmability.

---

So perhaps we have a race between last buy for the obsoleted die
form of the at91sam7l128 and the product life for the 30b/20b.
While I fully expect HP had the wherewithal to engage Atmel
contractually for a supply of the die through the product life,
the lqfp package in the 15c le sort of sticks out like a sore
thumb. I suppose such a supply contract could have been limited
to programs in place at the time it was executed. Just wondering
whether we'll see a pcb layout change to accommodate a lqfp
sam7 in a 20b/30b at some point.

One question I've had which I hope someone can answer is whether
the 12c+ 30th Anniversary Edition uses the lqfp package or rather
the considerably less inviting COB die.

Quote:

---

What a shame this person sent it back! If HP never releases the firmware, it's rare and special since nobody can create it by flashing. If HP does release the firmware for flashing some day, then it's not special anymore but then it could be flashed to the correct one and nothing of value was lost (and it could still be rare due to the firmware version).

---

Actually, I still have it, tucked in a drawer.