HP-12c 'comma' variant (?brazil)

[dm319]

There looks to be a variant of the HP-12c being sold in Brazil which has markedly different algorithms to the regular HP-12c. I came across it on this reddit thread, and when tested with these calculations it exhibited very different abilities to the regular HP-12c or HP-12c platinum.

Here are the overall accuracy results:

Code:

|calculator                  |1    |1b   |2    |3    |4    |5A   |5B   |6A   |6B   |7    |8    |9    |10   |11   |12   | median|   mean|
|:---------------------------|:----|:----|:----|:----|:----|:----|:----|:----|:----|:----|:----|:----|:----|:----|:----|------:|------:|
|HP-12cp                     |12.4 |11   |11.7 |12.2 |11.1 |ERR  |ERR  |ERR  |ERR  |21.4 |12.2 |3.2  |3.5  |11   |2.8  | 10.972|  9.031|
|HP-12c                      |9.6  |8.6  |8.8  |10.6 |9.1  |ERR  |ERR  |ERR  |ERR  |19.7 |11.5 |1.7  |1.6  |4.6  |1    |  8.278|  7.337|
|HP-12c comma                |9.6  |9.5  |11.7 |7.2  |ERR  |8.7  |NA   |8.4  |NA   |ERR  |12.2 |3.2  |3.2  |NA   |NA   |  7.182|  7.027|

And here are some of the results obtained:

Code:

| calculator     | 1               | 1b            | 2            |
|----------------|-----------------|---------------|--------------|
| HP-12c         | -1368.148355    | 4.37321837    | 6803.092172  |
| HP-12cp        | -1368.14835535  | 4.3732183723  | 6803.092162  |
| HP-12c comma   | -1368.148355    | 4.373218372   | 6803.092162  |

| calculator     | 3              | 4                   |
|----------------|----------------|---------------------|
| HP-12c         | 331667.0067    | -7.983367984e-10    |
| HP-12cp        | 331667.006691  | -7.98336798336e-12  |
| HP-12c comma   | 331666.9849    | ERR                 |

| calculator   | 5           | 5A          | 5B  |
|--------------|-------------|-------------|-----|
| HP-12c       | ERR         | ERR         | ERR |
| HP-12cp      | ERR         | ERR         | ERR |
| HP-12c comma | 14.43587133 | 14.43587133 | ERR |

| calculator    | 7                  | 8               |
|---------------|--------------------|-----------------|
| HP-12c        | 2.181818182e-10    | 3.125004736e-6  |
| HP-12cp       | 2.18181818182e-10  | 3.125001e-6     |
| HP-12c comma  | ERR                | 3.125001e-6     |

You can see it is not coming up with answers for some of these problems, but it is also able return results for double root solutions that the regular HP-12c and HP-12c platinum can't.

Just wondering if anyone else has come across this version? It definitely seems to be a bit of a curiosity - I wonder how official it is.

[dm319]

I just had a quick look on amazon.com.br, and complaints are easily found:

Quote:Comprei essa calculadora e descobri que há um lote deste produto sendo vendido no mercado apresentando alguns defeitos, há vários relatos de outros compradores que também descobriram durante o uso da HP, a calculadora não muda de ponto para virgula e vice-versa e não executa taxa de retorno "tir" apresentando "error 7" ao invés de mostrar o resultado, ja efetuei a troca com o representante da marca e mesmo assim mandaram outra calculadora com defeito e pelo jeito a HP não está querendo retirar o lote do mercado e nem fazer um recall das calculadoras, quem comprar procure fazer alguns testes para ter certeza que a mesma não veio com defeito.

With google translate:

Quote:I bought this calculator and discovered that there is a batch of this product being sold on the market with some defects, there are several reports from other buyers who also discovered while using the HP, the calculator does not change from dot to comma and vice versa and does not perform rate return "tir" showing "error 7" instead of showing the result, I already exchanged it with the brand's representative and they still sent another defective calculator and apparently HP doesn't want to remove the batch from the market or do anything a recall of calculators, whoever buys it should try to do some tests to make sure it didn't come with a defect.

and another:

Quote:comprei uma hp 12 c com lote CIGP328 e não estava conseguindo calcular a Taxa Interna de retorno IRR e nem trocar ponto por virgula no separador de milhar, vi na internet que é um lote com defeito, gostaria de substituir.

Quote:I bought an HP 12 c with batch CIGP328 and I was unable to calculate the Internal Rate of Return IRR or change the dot for a comma in the thousands separator, I saw on the internet that it was a defective batch, I would like to replace it.

I wonder what happened. Did someone somewhere just completely re-implement the HP-12c ROM, or accidently made a batch with an experimental ROM. The devices look well made, and it seems like HP are exchanging them for calculators with the same issue.

[Gerson W. Barbosa]

Isn’t the low price an indication it is probably fake? R$ 264,89 (~ US$ 48.00) compared to around R$ 400,00 at local stores.

[Eric Rechlin]

As I recall, counterfeit 12C calculators have been a problem in Brazil in the past, so that may well be what it is.

[SlideRule]

(06-28-2024 09:56 AM)dm319 Wrote:  … The devices look well made …

Anyone have a pic?

BEST!
SlideRule

[dm319]

My initial thought was this was a fake. Here are some photos. The comments suggest people are getting in contact with HP about this. Admittedly cheaper than HP-12c's in the UK, though prices here have gone up significantly in the last 5 years with inflation.

[Gerson W. Barbosa]

(06-28-2024 04:16 PM)dm319 Wrote:  My initial thought was this was a fake. Here are some photos.

Same ugly % key as in the one described in this old thread.

[dm319]

I've done a bit more digging.

So the main features of the HP-12c in question are:

1. Inability to switch from point to comma for radix
2. Error 7 on IRR and NPV
3. Inaccuracies and problems with TVM calculations
   

If you have a look at the official Brazil HP store, which sells the HP-12c, you can find reviews which suggest that people are obtaining these units through the official store.

Here is a photo of the 12c which exhibits unusual TVM behaviour:

serial number: C1QP328

Here is a product photo from the HP Brazil website:


They both have a tall '%' symbol, and they both do not have the 'Hewlett Packard' in the bezel.

This is the back of a unit as posted by a happy customer:


Serial number C1Q328, made in China, and uses 2 x CR2032 batteries.

Overall there are 166 ratings on the official Brazil HP store, of which 13 are 1 star. So the majority seem happy with their purchase. However, if you look at the low-rating reviews, they mention similar things, all by customers who have apparently a confirmed purchase.

Here are a couple of google-translations:

"I DID THE ORINALITY TESTS AND NONE OF THEM WORKED. I BOUGHT ON THE WEBSITE TO AVOID THIS DEFECT AND THE BRAND'S OWN WEBSITE DOESN'T SEND THE ORIGINAL PRODUCT SHAME"

and

"This is the fourth calculator I have purchased and it is defective. They all came with an "Error 7" defect in the IRR calculation and does not change the comma punctuation. Even the warranty couldn't resolve it. I would like HP to contact me, so that perhaps we can resolve the situation and I can have a calculator that actually works. I am deeply disappointed, unfortunately!"

[dm319]

Putting this altogether, my only conclusion is that HP Brazil is selling a version of the HP-12c which is a variant and has a completely re-implemented TVM solver and firmware.

I have no idea why. Have HP Brazil been sold a fake, or were they required to not use the regular HP-12c firmware? Did some experimental firmware accidentality get into the production line?

[AnnoyedOne]

(08-25-2024 06:39 PM)dm319 Wrote:  I've done a bit more digging.

FYI dm319 and I have exchanged many PM's about this.

They look "fake" to me.

The firmware info from one (serial C1GP328).

Code:


CHE--3198h
2023-03-19
(c)2008 HP

Note that these are labelled as a 12C and not a Platinum or Prestige! That said the firmware checksum above looks Platinum-ish. The CPU is unknown.

A1

[dm319]

(08-25-2024 06:55 PM)AnnoyedOne Wrote:  Note that these are labelled as a 12C and not a Platinum or Prestige! That said the firmware checksum above looks Platinum-ish. The CPU is unknown.

I don't have a regular HP-12c, but looking at the back of my Chinese-made HP-12c Platinum, there are many similarities. The size and location of the battery cover (mine doesn't have a screw though), and the info sheet on the back is the exact same shape and size. Mine is silver rather than metallic brown though, and has some differences (i.e. mine has change mode to RPN/ALG which is absent).

My serial is CNA3380P89.

The front is totally different and the font is not the same at all.

[AnnoyedOne]

(08-25-2024 07:46 PM)dm319 Wrote:  ...but looking at the back of my Chinese-made HP-12c Platinum, there are many similarities.

Some Chinese factories can make very good fakes. The details are often copied from known genuine products. One has to play "detective". And/or open a unit.

BTW these units might have (hidden) Platinum/Prestige functionality. You'd have to try it.

A1

[rprosperi]

One thing to keep in mind about customer complaints that a unit is bad, is that the Owner's Handbook still includes the original Voyager Self-Test instructions in Appendix E, and they do not work correctly on a 12C with Nut emulation. Why HP chose to leave these now incorrect instructions in the manual is not clear (whoever was in charge of the manual was probably clueless as he/she was told that it 'works just like the original'). My point being that customer complaints that a unit is bad simply because of failing these self-test steps are invalid.

Did anyone post which steps/sequence leads to the Error 7 on IRR and NPV ?

Without a specific test problem that leads to this, a series of customer complaints that they're all having this issue sounds non-legit to me, sounds more like people jumping on a bandwagon trying to get a replacement unit, for whatever reasons.

Also HP in Brazil no longer manufactures these, all mfg. is coordinated by Moravia, and then sold to Brazil dealers or distributors by Royal (HP's Licensee for the Americas sales markets) so any recent units legit dealers are selling are those w/o the "Hewlett Packard" legend, and made in Philippines.

I seriously doubt these units are Platinums with non-Platinum markings as the Platinum units use an LCD with notably different shapes and style.

For current units the rear plate on the 12C is brown and the 12CP is silver (which btw, is opposite of the early Platinums which had a silver face and black plate).

All that said, I have a 12C from recent production and it does not have the tall percent sign, it has a percent symbol that is similar to the one on the keys to its left, so perhaps there is an imposter source out there, pushing 12C work-alikes into the Brazil market.

[dm319]

(08-25-2024 10:31 PM)rprosperi Wrote:  Why HP chose to leave these now incorrect instructions in the manual is not clear (whoever was in charge of the manual was probably clueless as he/she was told that it 'works just like the original').

That's interesting to know. I suspect people will carry out this test when they find a problem, and so this would add to a confirmation bias that the device isn't genuine.

(08-25-2024 10:31 PM)rprosperi Wrote:  Did anyone post which steps/sequence leads to the Error 7 on IRR and NPV ?

From the reddit page:
"The calculator also has a bug when calculating the IRR of a cash flow with less than 4 CFn"

(08-25-2024 10:31 PM)rprosperi Wrote:  Without a specific test problem that leads to this, a series of customer complaints that they're all having this issue sounds non-legit to me, sounds more like people jumping on a bandwagon trying to get a replacement unit, for whatever reasons.

I can think of a few plausible scenarios:

1. The complaints are user error
2. These are fake devices not sold by HP
3. These are fake devices inadvertently sold by HP
4. These are devices sold by HP with firmware that was never meant for it
   

(08-25-2024 10:31 PM)rprosperi Wrote:  Also HP in Brazil no longer manufactures these, all mfg. is coordinated by Moravia, and then sold to Brazil dealers or distributors by Royal (HP's Licensee for the Americas sales markets) so any recent units legit dealers are selling are those w/o the "Hewlett Packard" legend, and made in Philippines.

Looks like some of the devices in the reviews have the bezel without Hewlett Packard, but some do. They seem to be made in China.

(08-25-2024 10:31 PM)rprosperi Wrote:  I seriously doubt these units are Platinums with non-Platinum markings as the Platinum units use an LCD with notably different shapes and style.

I agree, it isn't a Platinum, but has some similarities with my one - 2 x CR2032, made in China - but agree the LCD of the model in question is very much regular HP-12c. Looking around, I can see the HP-12c 30th anniversary has two CR2032 and is also made in China.

(08-25-2024 10:31 PM)rprosperi Wrote:  All that said, I have a 12C from recent production and it does not have the tall percent sign, it has a percent symbol that is similar to the one on the keys to its left, so perhaps there is an imposter source out there, pushing 12C work-alikes into the Brazil market.

Look at the product page of the USA or other official websites - it doesn't have that style of typeface for the keys. I've only seen the tall '%' with the chunkier lettering in the context of HP-12c's from Brazil.

If I'm able to get hold of one - what tests do people suggest I try on it?

EDIT: Edited for tone, apologies.

[dm319]

(06-28-2024 08:39 PM)Gerson W. Barbosa Wrote:  Same ugly % key as in the one described in this old thread.

I sadly can't see the images from this thread - does anyone know how to look at them?

[AnnoyedOne]

(08-25-2024 10:31 PM)rprosperi Wrote:  Why HP chose to leave these now incorrect instructions in the manual is not clear...

The January 1992, April 2008 HP-12C and HP-15C CE (rewritten) manuals do detail the procedure on p17/p61 (ON and '.'). Correct? I know the self-test sequence changed.

FYI the self-test instructions for HP Voyager models can be found (in English) here

https://hpofficesupply.com/wp-content/up...f-Test.pdf

HP office supply has some other FAQ's about the Voyager models for those that can read English. I've no idea if there's Portuguese (Brazil) versions available.

I can

1. Perform calculations in algebraic and RPN modes.
2. Drive on either side of the road.
3. Speak/read/write international (UK) and North American "English".

However languages aren't my strong point. I once lived in a non-English speaking country. I never did learn the language.

(08-26-2024 12:41 PM)dm319 Wrote:  ...what tests do people suggest I try on it?

1) Open the battery compartment and take/publish a photo. Is a POGO (programming) connector present?

2) Use g-ENTER-ON and record/publish the firmware version.

3) RESET the calculator if it has a button in the battery compartment or remove the batteries.

4) Run whatever TVM tests you wish and compare/publish the results.

5) If possible read the firmware out of the unit. If you don't have a cable etc try and find someone who can.

6) Open the unit and take a look inside. Take and publish photo's.

Be a detective. Document what you do. Photographic "evidence" can help others see details you might miss. Details details details!

A1

[AnnoyedOne]

(08-26-2024 12:43 PM)dm319 Wrote:  I sadly can't see the images from this thread...

Unfortunately many members post links to 3rd party image sites to bypass size/etc limits this one imposes. Those links often change or get deleted over time. You can look at the original URL (if possible) and manually edit it. That may, or may not, work.

A1

[EdS2]

(08-26-2024 12:43 PM)dm319 Wrote:  

(06-28-2024 08:39 PM)Gerson W. Barbosa Wrote:  Same ugly % key as in the one described in this old thread.

I sadly can't see the images from this thread - does anyone know how to look at them?

Try an old capture from the Wayback Machine

[AnnoyedOne]

(08-26-2024 04:39 PM)EdS2 Wrote:  Try an old capture from the Wayback Machine

Yeah, archive.org has lots of old stuff. Except when they don't. I experience about a 50% success rate with them. Worth trying but...

Unfortunately the image of a unit with an elongated % key was hosted on Dropbox (I looked at the URL). It wasn't cached so...

Quote:Message #11 Posted by gene wright on 3 Apr 2011, 2:17 p.m., in response to message #9 by Jeff Kearns

The serial number is: 4CY02805760. You can see that the % symbol font looks different from the other % on the calculator and looks different my my 12c+ % key.

A1

[dm319]

Yes, unfortunately the key image is not there. I tried putting the Dropbox link into the archive but at least on the first hit it was a 404.