Is the 12C SDK still available?

[Katie Wasserman]

(11-22-2015 05:47 PM)rprosperi Wrote:  1. If the case (and presumably the 6-pin connector) are the same, is there any danger of harming the new units if I use the older (serial) cable? Or I suppose the opposite as well (though less likely with lower voltages?)

I don't think that harm will come to the calculator either way. I've made both mistakes in my testing and nothing bad happened. I'm bothered by the fact that HP used the same -- impossible to buy -- connector, they had the chance to install a micro USB connector. My guess is that HP wanted to do this as a silent roll-out so kept the case exactly the same. They also don't want to encourage hacking of the 12C, I think.

Quote:2. Can you post a pic of the new back-plate design (unless it's pre-production, then it's kinda meaningless)?

Yeah, it might be misleading. I need to look at it the next time I'm in Staples to see what the latest production run looks like (if they have it on display).

[rprosperi]

(11-22-2015 06:10 PM)Katie Wasserman Wrote:  Yeah, it might be misleading. I need to look at it the next time I'm in Staples to see what the latest production run looks like (if they have it on display).

Thanks for confirming that. Have you found a source of the USB version of the cable with the (JTAG?) connector?

Regarding the label, it's as I guessed. Unfortunately, in the retail store, there are no pkgs to open (not that one should do so anyhow...) and the blivet (that's what we always called the non-functioning mock-up samples used in some retail shops, like Staples) is very likely not updated for new production; as you said, it was intended to silently refill shelves with no fanfare. Even the HP Website has not been updated to show the "new" design with 2 batteries - and they've had many years to do that. Makes sense I guess, it's likely best overall to provide the impression that the venerable and trusted 12C design is still just-exactly-as-it-was all along.

It was noted at HHC that there is some small detail on the retail pkg which identifies the new models, I will follow-up to find out what that is, to be sure when checking a new one.

Thanks.

[Dwight Sturrock]

(11-22-2015 05:03 PM)Katie Wasserman Wrote:  

(11-22-2015 05:47 AM)Dwight Sturrock Wrote:  I recently purchased a 12C with a 2015 serial code date, and it definitely has (2) batteries.

You can confirm if you have the latest Atmel chip by checking the firmware date as follows:

Hold down <g> + <ENTER> when turning on the calculator.
You'll see the "1.L 2.C, 3.H" menu
press <2>
You'll see the checksum
press <ENTER>
You'll see the firmware date as YYYY-MM-DD

If it's in 2015, you have the new Atmel version. My sample is 2015-01-30.

For my 2015, 22nd week 12C, I get a firmware date of 2012-04-03. 


So it sounds like the last of the old batch

[matthiaspaul]

(11-22-2015 05:47 PM)rprosperi Wrote:  1. If the case (and presumably the 6-pin connector) are the same, is there any danger of harming the new units if I use the older (serial) cable? Or I suppose the opposite as well (though less likely with lower voltages?)

(11-22-2015 06:31 AM)Katie Wasserman Wrote:  

Code:


||
||   1    2    3    ||
||   4    5    6    ||
||

1 - reset button (connect to ground)
2 - ground
3 - erase button (connect to pin 4)
4 - erase button (connect to pin 3)
5 - USB green wire (data -)
6 - USB white wire (data +)

Assuming that this is meant to be the view onto the calculator and applying the standard pin naming scheme for dual-row pin headers suitable for flex-cables, the pin numbering would change as follows:

View onto 2x3 pin header on connector:

Code:


||  ||          ||
||   1    3    5   ||
||   2    4    6   ||
||

View onto 2x3 pad array on calculator:

Code:


||
||   2    4    6   ||
||   1    3    5   ||
||  ||          ||

Pinout of 2x3 flash connector on AT91SAM7L128 based calculators like
HP-12C (F2230A, 2008-2015) aka "HP-12C+", HP 12c 30th Anniversary Edition (NW258AA, 2011), HP 15c Limited Edition (NW250AA, 2011) aka "HP-15C+", HP 10bII+ (NW239AA, 2011-2015), HP 20b (F2219A, 2008-2012), HP 30b (NW238AA, 2010-2014):

1 VBAT (connect to pin 6 to erase)
2 NRST (connect to pin 4 to reset)
3 TXOUT (direction as per calculator view)
4 GND
5 RXIN (direction as per calculator view)
6 ERASE (connect to pin 1 to erase)

Pinout of 2x3 flash connector on SAM4LC2CA based HP-12C (2015+) and HP 10bII+ (2015+):

1 ERASE1 (connect to pin 6 to erase)
2 NRST (connect to pin 4 to reset)
3 USBD- (bidirectional)
4 GND
5 USBD+ (bidirectional)
6 ERASE6 (connect to pin 1 to erase)

So, the pinouts of the two connectors are for the most part compatible:

USB data lines have much lower operational voltages, but per the USB specification the circuitry must be able to withstand ca. 3.3V on them, so there isn't any risk of damage when using the old serial programming cable on the new USB calculator model - it just won't work. Well, in theory (depending on the boot loader code), it would be even possible to support both, USB and serial I/O over these pins.

For the serial pinout there was a risk to permanently damage the calculator when not powering the calculator externally (through VBAT) without batteries, but through older batteries inserted. This risk does not exist for the USB pinout any more, as USB data lines aren't strong enough to reverse power the calculator through the clamping diodes.

The arrangement of the ERASE1 and ERASE6 pins certainly warrants some more investigation. It's possible that one of them is actually VBUS (or a 3.3V-compatible derivation of it).

Greetings,

Matthias

[Katie Wasserman]

(11-22-2015 08:04 PM)Dwight Sturrock Wrote:  For my 2015, 22nd week 12C, I get a firmware date of 2012-04-03. 

So it sounds like the last of the old batch

Yes.

[Katie Wasserman]

(11-22-2015 08:31 PM)matthiaspaul Wrote:  The arrangement of the ERASE1 and ERASE6 pins certainly warrants some more investigation. It's possible that one of them is actually VBUS (or a 3.3V-compatible derivation of it).

Yeah, one of them is almost certainly Vdd of the calculator, I didn't bother to check I was just tracing the pins on the cable.

[Katie Wasserman]

(11-22-2015 06:49 PM)rprosperi Wrote:  Thanks for confirming that. Have you found a source of the USB version of the cable with the (JTAG?) connector?

It's not JTAG just some crazy thing the HP came up with for these calculators.

Anyway, the cable I have for the USB version is just a hacked up one for the serial version. It should be possible to modify the old cable by adding a DPDT switch and a USB connector so that it would work for both calculator versions, as only 2 pins need to switch.

[Dwight Sturrock]

(11-22-2015 08:45 PM)Katie Wasserman Wrote:  

(11-22-2015 06:49 PM)rprosperi Wrote:  Thanks for confirming that. Have you found a source of the USB version of the cable with the (JTAG?) connector?

It's not JTAG just some crazy thing the HP came up with for these calculators.

I wonder if they might be persuaded to sell us some of these cables? I assume they use them for the calcs initial programming?

[rprosperi]

(11-22-2015 08:45 PM)Katie Wasserman Wrote:  It's not JTAG just some crazy thing the HP came up with for these calculators.

Anyway, the cable I have for the USB version is just a hacked up one for the serial version. It should be possible to modify the old cable by adding a DPDT switch and a USB connector so that it would work for both calculator versions, as only 2 pins need to switch.

We're talking about the same cable as one uses for updating the WP-34S (or 20b/30b/12c+/15c+) right?

DPDT switch?

I like the idea of a single cable for all machines. Far too many cables in my life already...

[matthiaspaul]

(11-22-2015 06:31 AM)Katie Wasserman Wrote:  Yeah, that's just the designation to use for SAM-BA, the chip is actually a ATSAM4LC2CA

In case you have opened the calculator already, is it the TQFP100 or the VFBGA100 package variant or a raw die version of the chip?

Greetings,

Matthias

[Katie Wasserman]

(11-22-2015 10:08 PM)rprosperi Wrote:  

(11-22-2015 08:45 PM)Katie Wasserman Wrote:  It's not JTAG just some crazy thing the HP came up with for these calculators.

Anyway, the cable I have for the USB version is just a hacked up one for the serial version. It should be possible to modify the old cable by adding a DPDT switch and a USB connector so that it would work for both calculator versions, as only 2 pins need to switch.

We're talking about the same cable as one uses for updating the WP-34S (or 20b/30b/12c+/15c+) right?

DPDT switch?

Right.

The DPDT (double-pole, double-throw) switch would be needed to swap the both data lines between the USB interface and the existing RS232 level converter. I don't think that there is enough room inside that little black shell for such a switch, so it might require some repackaging.

[Katie Wasserman]

(11-22-2015 10:13 PM)matthiaspaul Wrote:  

(11-22-2015 06:31 AM)Katie Wasserman Wrote:  Yeah, that's just the designation to use for SAM-BA, the chip is actually a ATSAM4LC2CA

In case you have opened the calculator already, is it the TQFP100 or the VFBGA100 package variant or a raw die version of the chip?

I just looked at it last night, but I think I remember it was the TQFP100 package, with all the proper labeling on it.

[matthiaspaul]

(11-22-2015 08:04 PM)Dwight Sturrock Wrote:  

(11-22-2015 05:03 PM)Katie Wasserman Wrote:  If it's in 2015, you have the new Atmel version. My sample is 2015-01-30.

For my 2015, 22nd week 12C, I get a firmware date of 2012-04-03. 

That's interesting given that this variant of the 12C should have been introduced in 2008. What are the differences of the 2008 version compared to the 2012 version?

Does someone have a (complete) list of known firmware dates?

Greetings,

Matthias

[Katie Wasserman]

(11-23-2015 02:22 AM)matthiaspaul Wrote:  

(11-22-2015 08:04 PM)Dwight Sturrock Wrote:  For my 2015, 22nd week 12C, I get a firmware date of 2012-04-03. 

That's interesting given that this variant of the 12C should have been introduced in 2008. What are the differences of the 2008 version compared to the 2012 version?

Does someone have a (complete) list of known firmware dates?

That would be nice to know. I've got a couple of different firmware versions but not the 2012-04-03 one. I would doubt that anyone has this info, I'm not even sure that HP tracks that. One would assume that all of the releases for the same version of the hardware are bug fixes and they probably only keep track of the latest version.

[Tim Wessman]

Hello,

There are two ways to tell the new version of the 12C or 10bII+.

Plate on bottom of calculator:
1. There is a "Rev 1" on the old atmel and a "Rev 2" on the new one.

By serial number:
1. NEW 10BII+ starts with “PHA”
2. NEW 12C starts with “PHA” or “9CJ” or “7CD” or “3CD”


We could not touch the ID at all in this revision. That is why there is not any change to the plug. In addition, adding any component was not allowed.

Note that we never tested plugging the original cable into the port. I honestly don't know what would happen if you plug a non-modified serial cable into what now has USB connections. I don't think it would do any damage, but I never wanted to find out since we had very limited numbers of the new boards at that point.

However, due to a serious oversight in our electrical design, someone accidentally left a surface mount on an extreme edge of the boards that happens to perfectly fit a surface mount micro USB. These pads were also accidentally connected to the internal USB circuitry of the chip. Should you accidentally cut a small hole in the plastic case on the edge that accidentally lines up with the accidental surface mount, you might be able to accidentally use USB to do reprogramming. Happy accidents! (obviously voids your warranty)

[rprosperi]

Thanks Tim!

And its good to know that accidents can still occur from time to time...

[Steve Simpkin]

(11-23-2015 05:08 PM)Tim Wessman Wrote:  However, due to a serious oversight in our electrical design, someone accidentally left a surface mount on an extreme edge of the boards that happens to perfectly fit a surface mount micro USB. These pads were also accidentally connected to the internal USB circuitry of the chip. Should you accidentally cut a small hole in the plastic case on the edge that accidentally lines up with the accidental surface mount, you might be able to accidentally use USB to do reprogramming. Happy accidents! (obviously voids your warranty)

LOL
Well, accidents happen

[matthiaspaul]

(11-23-2015 05:08 PM)Tim Wessman Wrote:  We could not touch the ID at all in this revision.

Thanks for the information, Tim!

Is there any more or less complete cross reference list of product IDs with calculator names/revisions HP could share with the community?

I understand that some MPNs are used for more than one revision for various reasons, but still it would help to more reliably identify a particular model/variant.

I don't know when HP started using these codes, but the earliest I am aware of is F1000A for the 512 KB variant of the HP 95LX in 1991. I also have some higher codes which have indirectly been associated with calculators introduced in 1988, but they may correspond with unknown later revisions. But stumbling upon "new" codes by accident and collecting them is one thing, having an authorative reference would be another...

Quote:you might be able to accidentally use USB to do reprogramming. Happy accidents! (obviously voids your warranty)

Sweet! This is one of those rare cases, where accidents in the design process actually increase the value of and demand for a product, I guess. ;-)

If just, by some further and completely unexplainable accident, the internal OS/emulator would support the transfer of programs, data, and settings between the calculator and an external application at runtime (similar to http://hp-15c.homepage.t-online.de/dm15.htm )... ;-)

Greetings,

Matthias

[Dwight Sturrock]

(11-23-2015 05:08 PM)Tim Wessman Wrote:  However, due to a serious oversight in our electrical design, someone accidentally left a surface mount on an extreme edge of the boards that happens to perfectly fit a surface mount micro USB. These pads were also accidentally connected to the internal USB circuitry of the chip. Should you accidentally cut a small hole in the plastic case on the edge that accidentally lines up with the accidental surface mount, you might be able to accidentally use USB to do reprogramming. Happy accidents! (obviously voids your warranty)

Awesome! Even better than HP selling us the "not JTAG just some crazy thing the HP came up with for these calculators" cables.

The small hole in the case could be cut with a router (Porter-Cable, not Cisco) and a small straight bit using a template that has the pattern to be cut and also holds the case rigid.

[rprosperi]

I wonder if anyone accidentally drew a diagram of where the accidental connector was placed? You know, just to document the accident accurately...