15c LE production



#2

How many of you think HP is pushing back the next batch of 15c LE production to address the faulty keyboard issue? (hpcalc.org says that their 15c LE has been backordered to 10/28)

On an unrelated note, does anybody else also think that another kind of old HP calculator revival like the 15c LE is simply not possible? 15c LE was only an incremental jump from the 12c, which is already in production. If you want to revive something like the 42s, they'll have to set up the production line to make that kind of device from scratch again. I just can't see HP engaging in that kind of business especially with the current business environment (new CEO, etc).


#3

Probably extremely unlikely that HP would re-introduce a Pioneer in the original form factor, for the reasons you cite. I had thought they *could* produce a 42s on the 17bii+ platform, But I've been informed that is impossible.

However, perhaps they could produce a "42s-like" calculator on a modern platform, much as they produced the 35s based in part on the 32sii/33s.

Whether there is or will be any incentive to do so is another matter.


#4

The 30b has the same number of keys as a 42S, and the same configuration if you look at them side by side. Even many of the labels are the same. Shift, up and down arrows, backspace/clear, big enter key, on/off.
The screen is the big difference. Swap out the 30b screen to something better and you would almost be there. If I was at HP I would use a 30b with a better screen to re-introduce the 42S. It wouldn't be that hard.
The 34S team has done a great job with the 30b considering the screen limitations. I can't even imagine what they could do if had a better screen.


#5

Paul, the 42s not only had a better screen, it had more memory and a printer interface. If you wanted to put this in the 30b package, a new board with a completely different chip would be needed. Given the late discussions about battery depletion, a new case design to support larger batteries would be needed, too.

So don't hold your breath!

#6

Well, a HP-designed 42S+ is possible if the current 17Bii+ is updated with the same ARM processor as the 12C+.

#7

Just put a bare-metal ARM port of Free42 into 30b hardware with a 17bii+ LCD, add some I/O and display driver code, and you'd have an awfully good start.


#8

The 30b hardware isn't capable of driving the 17bii+ LCD, or anywhere close to it. The Atmel chip can drive a maximum of 400 segments (or pixels), and the 17bii+ display has over 2100.

A "bare-metal ARM port of Free42" isn't trivial, either. I tried to do that a few years back, and it was taking a *lot* of work to rewrite all of the memory management. Obviously it can be done, but it isn't a small project.


#9

Connecting a different LCD isn't impossible but you can't just take a bare LCD, it needs it's own controller. There are enough pins on the Atmel chip to drive such a device. The internal LCD controller would then be disabled and the pins reused for I/O.

What you can't overcome are the memory restrictions. You would need some external memory chip connected to the device.


#10

Not going to happen in a low-cost HP product. In OEM quantities, LCD modules with controllers are *much* more expensive than bare LCDs.


#11

I agree. HP having chosen a system-on-a-chip design has been cost driven from the beginning and would be foiled by adding expensive external components to overcome its limitations.

#12

Quote:
The 30b hardware isn't capable of driving the 17bii+ LCD, or anywhere close to it. The Atmel chip can drive a maximum of 400 segments (or pixels), and the 17bii+ display has over 2100.

Bummer.


Quote:
A "bare-metal ARM port of Free42" isn't trivial, either. I tried to do that a few years back, and it was taking a *lot* of work to rewrite all of the memory management. Obviously it can be done, but it isn't a small project.


I don't doubt that. It would certainly be a better prospect than hacking new features into the existing 42s ROM binaries, though.

#13

Quote:
If you want to revive something like the 42s, they'll have to set up the production line to make that kind of device from scratch again.

Emulating old calculators like the 30-year-old 15C and the 23-year-old 42S is a really bad idea. Emulation on a fast processor winds up wasting the vast majority of available battery energy, compared to doing the same thing in some version of C+ that generates native code for the host processor. In that latter case, program run times are a small fraction of those in an emulated version.

The HP 30b has the same number of keys and key arrangement as the original 42S. There is no emulation layer on the 30b and it is the fastest handheld that HP has ever made by large margin (ten times faster than the HP 50G, eight times faster than the 15C-LE, both with emulation layers). The 30b is also a more attractive and sturdy package than the original 42S. The best solution would be an HP 42S equivalent in a HP 30b package, using something other than CR2032 cells. But something other than an AT91SAM7L128 would be required to handle the display requirements and other issues.


#14

Quote:
Emulating old calculators like the 30-year-old 15C and the 23-year-old 42S is a really bad idea. Emulation on a fast processor winds up wasting the vast majority of available battery energy..

It need not unconditionally be so. The sam7l voyager product
was chosen to run at the maximum clock frequency for some
perceived (I expect primarily marketing) benefit. For a pocket
calculator otherwise wildly successful at a fraction of this
throughput, doing so was arguably a little heavy handed to
force as a non-negotiable default.

Quote:
..compared to doing the same thing in some version of C+ that generates native code for the host processor. In that latter case, program run times are a small fraction of those in an emulated version.

I don't think that is a foregone conclusion. In both cases
you are drawing upon some set of routines to implement the
abstract model functionality. In the pure c implementation
case, we define the model. In the emulation case the model
has been defined and we're making a best effort attempt to
map into an efficient implementation. An area in which emulation
falls short is where features exist in the abstract model which
are inefficient to support relative to the host architecture, or
features in the host architecture (eg, native multiply) which
aren't defined in the base model. I'd give emulation the upper
hand in terms of code density vs. the pure c implementation
equivalent where we're creating more memory consuming function
call invocations vs. encoding operation semantics in terse
instruction data. Doing so as a hybrid approach to reduce
memory footprint in a c implementation swings us back closer
to emulation where we're interpreting some form of optimized
intermediate code.

#15

10/28 is a guesstimate. I only put that so as not to disappoint people. If my supplier is right (though they have been known to be wrong in the past, hence the 10/28 guesstimate) I should get another shipment on 10/12.

From what has been posted here in the past, I would say you shouldn't expect anything different about these 15c calculators.

Eric

#16

Quote:

On an unrelated note, does anybody else also think that another kind of old HP calculator revival like the 15c LE is simply not possible?


I'd never say such a thing. A year ago, everyone who didn't go to HHC 2010 thought a new 15C was flatly impossible. I thought the 15C LE was a gift to the community, but looking at the reaction of buyers, as reflected in eBay auctions, it's clear they are selling very well. Obviously HP thought they could make a profit on this device. I doubt it was this simple, but maybe they looked at the robust sales of original 15Cs, did some math and went for it. It looks to me like they have a hit on their hands, despite all the defects Katie and others have turned up.

There's another calculator that sells very well on eBay, though it's in shorter supply than the 15C, so sales volume is necessarily low. I'm wondering if a reloaded 42S might not sell well. I was reminded in this thread that it would need a more expensive display. The engineering work to make that happen would boost their non-recurring expenses. But if they could sell enough machines it might cost out. If they went for a new machine in the spirit of the 42S instead, that would require a lot more engineering time and expense up front. I suspect one reason the 15C got a green light was the fact that very little time needed to be spent on the emulated 15C image, as opposed to the emulator itself.

I still think it's highly unlikely anything will emerge from the calculator division along those lines, but I've been wrong before. :)


#17

The 16C re-sells really well too. Once the bugs are worked out of the 15C LE with (hopefully) a new firmware release the 16C should be a whole lot easier than a 42s reissue.


#18

Quote:
a new firmware release the 16C should be a whole lot easier than a 42s reissue.

No doubt about that - however being a dedicated machine, its potential interest and appeal to the user community is substantially less.

I own two 16C and they're fast enough for what I used them. So speed increase alone isn't worth so much. My 2 cents.

#19

Quote:
A year ago, everyone who didn't go to HHC 2010 thought a new 15C was flatly impossible.

Not so. Research the archives. For the last couple of years there were several of us predicting it would happen who had no inside knowledge. In particular, DaveJ was outspoken in this opinion.

#20

Quote:

Not so. Research the archives. For the last couple of years there were several of us predicting it would happen who had no inside knowledge. In particular, DaveJ was outspoken in this opinion.


Clearly "everyone" was a gross generalization. :) My apologies for painting with too broad a brush.

#21

At HHC 2010, we were given a survey along the lines of "Which calculator would you like to see re-issued." (I'm pretty sure we took the survey before the pre-production 15c+ calculators were handed out.) I don't recall if there was a list of choices or if it was a write-in, but the 42S was certainly a candidate. I used to be a huge 42S fan, having replaced my 15C with one in 1988 and then using it for about 15 years on a daily basis, but, I voted for the 15C. The reason I gave was that the 15C could be brought back with no functional changes and serve an identifiable niche - a very powerful, small calculator that can be operated easily, programmable to do some complex things if you want, with a useful amount of memory but not so much that you cry if you lose it. The 42S, on the other hand, was a near miss at a truly fantastic calculator. If it was brought back as-is today, its shortcomings would really stand out. What we really want is the 43s, of course :-)

For the record, to make the 42s near-perfect in my opinion, it needs more memory (no problem today) some form of I/O, and complex number support as described here, for example. I'd also like to see a larger screen, to give better graphing support but mostly to display the stack, Last x, and the date and time (ala the one Eric showed for DIY5). But now I am heading into feature-creep territory.


#22

Quote:
For the record, to make the 42s near-perfect in my opinion, it needs more memory (no problem today) some form of I/O, ...

At this point, it becomes almost mandatory that the calculator becomes large enough to carry at least two AAA cells. Button and coin sized cells can not support much I/O, especially if one of the most desirable methods (micro SD) were employed. I would like to see a design that used coin or button cells for compactness, had micro SD and USB, but only allowed writes to the micro SD when external power through the USB connector was present.

Quote:
...and complex number support as described here, for example.

Your reference doesn't really describe complex number support, just complex number entry method, which IMHO just scratches the surface of complex support. Complex number support in depth and detail means any function that the machine performs which is defined for the complex domain as input or output is performed naturally by the machine. In that respect, the 42S overwhelms the capability of any other handheld made by anyone ever, except the RPL machines. There are a few omissions, such as lack of a complex gamma function. I'm quite happy with the current 42S complex number entry method, and I am unconvinced of any real advantage for other approaches.

Quote:
I'd also like to see a larger screen, to give better graphing support but mostly to display the stack, Last x, and the date and time (ala the one Eric showed for DIY5). But now I am heading into feature-creep territory.

I would like to see an HP 42S replacement that, in addition to the micro SD and USB external power source mentioned above, is the same size (accepting the limitation that places on power source), faster, and has real time clock, calendar, and financial programming of the 17bii. And...one that isn't "orange on brown/dark brown" ugly.


#23

Quote:
Your reference doesn't really describe complex number support, just complex number entry method...

Yes, I guess I should have said complex number “entry, display and manipulation" instead of “support.” I concur with your desire for full support for any function defined for the complex domain.

Quote:
I'm quite happy with the current 42S complex number entry method, and I am unconvinced of any real advantage for other approaches.
You are OK with:

key in real component
press ENTER
key in imaginary component
press shift
press STO (for shifted COMPLEX function)

vs.

key in real component
press "i"
key in imaginary component?

Can you cite any disadvantage to the entry method I proposed? (Other than needing an "i" key and so requiring some change to the keyboard.) In addition to the extra keystrokes required for entry, the 42S method has the disadvantage of pushing T off the top of the stack when you want to enter a complex number.

Quote:
I would like to see an HP 42S replacement that, in addition to...

Put an unshifted "i" key on it, and I'd buy it in a heartbeat. Otherwise, it might take a minute.

...

Edited: 4 Oct 2011, 12:22 p.m.


Possibly Related Threads...
Thread Author Replies Views Last Post
  App production on Prime Geoff Quickfall 2 416 10-11-2013, 04:51 PM
Last Post: Geoff Quickfall
  OT - A lucky find - Busicom Handy-LE LE-120A Cristian Arezzini 2 530 09-26-2013, 04:43 AM
Last Post: Cristian Arezzini
  Low power warning for HP-15C LE and batteries Nick_S 1 479 09-16-2013, 09:34 AM
Last Post: Hans Brueggemann
  HP 15C-LE replacement still available? Borja 16 1,662 08-22-2013, 11:16 AM
Last Post: Michael de Estrada
  JTAG on HP-12C and HP-15C LE Ingo 5 879 07-01-2013, 06:37 PM
Last Post: Paul Dale
  HP 15C LE extremums Richard Berler 29 2,912 05-20-2013, 03:26 PM
Last Post: Dieter
  New 15C LE bug? Paul Dale 3 663 02-05-2013, 09:27 PM
Last Post: Mike Morrow
  HP 15C LE, Program Display Format Control Uli 4 752 01-20-2013, 01:22 AM
Last Post: Ethan Conner
  hp 15c calculator production. William L. Drylie 10 993 11-19-2012, 12:48 AM
Last Post: jason foose
  HP-15C LE, wow! New owner euphoria. Sasu Mattila 16 1,496 09-07-2012, 06:55 PM
Last Post: lars Bergström

Forum Jump: