Threaded Mode | Linear Mode

Michael Eckstein · 05-11-2014, 10:14 AM

It's well known, that early HP 35 had a.o. the famous 2.02 ln ex bug.
I have just recently discovered an another and extremely rare BUG in HP 35. That bug was very early fixed by HP company and therefore even vast majority of Red Dots don't have that bug already.
It might be in other HP 35 too, provided that that buggy ROM would be accidentaly placed into some HP 35. You can test your HP 35s by this sequence:

1 eex chs 10 ln

and if you'd be a lucky owner of such HP 35 :-) then you'd receive an erroneous answer differing from the correct answer by hunderts of percents!

Massimo Gnerucci · 05-11-2014, 11:25 AM

(05-11-2014 10:14 AM)Michael Eckstein Wrote: It's well known, that early HP 35 had a.o. the famous 2.02 ln ex bug.
I have just recently discovered an another and extremely rare BUG in HP 35. That bug was very early fixed by HP company and therefore even vast majority of Red Dots don't have that bug already.
It might be in other HP 35 too, provided that that buggy ROM would be accidentaly placed into some HP 35. You can test your HP 35s by this sequence:

1 eex chs 10 ln

and if you'd be a lucky owner of such HP 35 :-) then you'd receive an erroneous answer differing from the correct answer by hunderts of percents!

I get -23.02585093 the same output from a 41C... I guess mine doesn't have this bug.

What do you get instead?

Sylvain Cote · 05-11-2014, 11:49 AM

(05-11-2014 11:25 AM)Massimo Gnerucci Wrote: I get -23.02585093 the same output from a 41C... I guess mine doesn't have this bug.

Funny, I did the same test 15min after Michael post, compared it with the same calculator and got the same result! Smile

HP-35 v2 s/n:1143A57505 has ln/eˆx, arc sin, arc cos & arc tan bugs

Sylvain

Massimo Gnerucci · 05-11-2014, 12:12 PM

(05-11-2014 11:49 AM)Sylvain Cote Wrote:
(05-11-2014 11:25 AM)Massimo Gnerucci Wrote: I get -23.02585093 the same output from a 41C... I guess mine doesn't have this bug.

Funny, I did the same test 15min after Michael post, compared it with the same calculator and got the same result!

HP-35 v2 s/n:1143A57505 has ln/eˆx, arc sin, arc cos & arc tan bugs

Sylvain

However yours is not a Red Dot... I wonder what Michael's result is.

Michael Eckstein · 05-11-2014, 01:47 PM

(05-11-2014 12:12 PM)Massimo Gnerucci Wrote:
(05-11-2014 11:49 AM)Sylvain Cote Wrote: Funny, I did the same test 15min after Michael post, compared it with the same calculator and got the same result!

HP-35 v2 s/n:1143A57505 has ln/eˆx, arc sin, arc cos & arc tan bugs

Sylvain

However yours is not a Red Dot... I wonder what Michael's result is.

When that bug would be present, then the result should be -123.0258509.

Nicolas Recalde · 05-11-2014, 02:13 PM

(05-11-2014 12:12 PM)Massimo Gnerucci Wrote:
(05-11-2014 11:49 AM)Sylvain Cote Wrote: Funny, I did the same test 15min after Michael post, compared it with the same calculator and got the same result!

HP-35 v2 s/n:1143A57505 has ln/eˆx, arc sin, arc cos & arc tan bugs

Sylvain

However yours is not a Red Dot... I wonder what Michael's result is.

My red dot serial number 1143A 00051 does have this BIG bug.

I called BIG because the answer of 1 eex chs 10 ln

on my red dot is: -123.0258509 instead of -23.0258509

I understand that real bugs are of the order of 1% or less.

But this BIG bug has an error of more than 400%

I am wondering why we just came to know now about this error

if this bug was already know!!

Many thanks to Michael Eckstein for releasing it to general public.

aurelio · 05-11-2014, 03:36 PM

(05-11-2014 02:13 PM)Nicolas Recalde Wrote:
(05-11-2014 12:12 PM)Massimo Gnerucci Wrote:

However yours is not a Red Dot... I wonder what Michael's result is.

My red dot serial number 1143A 00051 does have this BIG bug.

I called BIG because the answer of 1 eex chs 10 ln

on my red dot is: -123.0258509 instead of -23.0258509

I understand that real bugs are of the order of 1% or less.

But this BIG bug has an error of more than 400%

I am wondering why we just came to know now about this error

if this bug was already know!!

Many thanks to Michael Eckstein for releasing it to general public.

Nicolas, you can send it to the HP, the will sure fix the bug, otherwise you can change it with mine, it works great Smile

Nicolas Recalde · 05-11-2014, 04:36 PM

(05-11-2014 03:36 PM)aurelio Wrote:
(05-11-2014 02:13 PM)Nicolas Recalde Wrote: My red dot serial number 1143A 00051 does have this BIG bug.

I called BIG because the answer of 1 eex chs 10 ln

on my red dot is: -123.0258509 instead of -23.0258509

I understand that real bugs are of the order of 1% or less.

But this BIG bug has an error of more than 400%

I am wondering why we just came to know now about this error

if this bug was already known!!

Many thanks to Michael Eckstein for releasing it to general public.
Nicolas, you can send it to the HP, the will sure fix the bug, otherwise you can change it with mine, it works great

I would love to know the history of this BIG bug in its early days when it was
discovered. Somehow it was kept quiet... until now!

Michael Eckstein · 05-11-2014, 05:58 PM

(05-11-2014 04:36 PM)Nicolas Recalde Wrote: I would love to know the history of this BIG bug in its early days when it was
discovered. Somehow it was kept quiet... until now!

This BIG bug is more than an unknown bug rather an ages ago forgotten bug.

A few words to this bug discovery:

I have recently seen a strange "HP 35 manual errata" page, where after the last correction of a typo in the manual was also mentioned this BIG bug, but no other bugs. Other now known famous bugs were still waiting for their discovery by HP. BIG bug was promptly fixed by HP and then forgotten :-) It might be present in a few tens calculators at that time and it could be in just a few ones remaining today in the hands of their lucky owners.

**Katie Wasserman** · 05-11-2014, 07:52 PM

Quote:My red dot serial number 1143A 00051 does have this BIG bug.

My red sot serial number 1143A 01660 does not have this BIG bug, but this isn't a useful data point because it had the 2.02 fixed ROM installed. Another HP-35 that I have with serial number 29273 that has the 2.02 bug still in it also does not have the BIG bug.

W_Max · 05-13-2014, 04:21 PM

Red dot PCB. Unknown S/N (non-original case) 2.02 present, BIG bug - not.

PCB photo: http://maxes.ho.ua/Pics/Calcs/hp351.jpg

HP-Collection · (This post was last modified: 05-13-2014 08:09 PM by HP-Collection.)

I copy the words of Nicolas Recalde:

My red dot serial number 1143A 00357 DOES HAVE this BIG bug.
I called BIG because the answer of 1 eex chs 10 ln
on my red dot is: -123.0258509 instead of -23.0258509

The very strange thing is: It does NOT have the "known bug"! What abour yours, Nicolas?

Seems to be a Bug in the Very early of the early serials....

HP-Collection · 05-14-2014, 11:09 AM

Correction: My bugy Red Dot is really bugy, it has BOTH bugs!
(I was off for too long and now I can not distinguish 2 from 2.02)

Massimo Gnerucci · 05-14-2014, 11:34 AM

(05-14-2014 11:09 AM)HP-Collection Wrote: Correction: My bugy Red Dot is really bugy, it has BOTH bugs!
(I was off for too long and now I can not distinguish 2 from 2.02)

Ah, OK.
'cause mine has the ln 2.02 bug but not the "BIG" one.

Nicolas Recalde · 05-14-2014, 02:55 PM

There is a very interesting web site about the HP-35 red dot:

http://hp35.wz.cz/HP-35RD.html

Sylvain Cote · 05-14-2014, 04:52 PM

(05-14-2014 02:55 PM)Nicolas Recalde Wrote: There is a very interesting web site about the HP-35 red dot:

http://hp35.wz.cz/HP-35RD.html

Yep! and managed by Michael Eckstein Big Grin

Michael de Estrada · 05-15-2014, 04:46 AM

My Red Dot also lacks this big bug, but that should come as no surprise considering its high s/n 1143A 04281.

ignazio · 05-15-2014, 07:44 AM

Hi,
even my 35 Red Dot lacks the known bugs, S/N 1143A02874.

HP-Collection · (This post was last modified: 05-23-2014 08:04 PM by HP-Collection.)

The serial must be lower than 1000 for that bug, I expect. All my higher Red Dots don't have the new bug.. ;(

Michael Eckstein · 05-22-2014, 04:48 PM

(05-15-2014 11:14 AM)HP-Collection Wrote: The serial must be lower than 1000 for that bug, I thnik. All my higher Red Dots don't have the new bug.. ;(

The threshold will be even lower, Red Dot 00840 doesn't have the BIG bug.
My Red Dot 02198 has old bugs only, too. In my collection there are another three HP 35 Versions 2 with bugs, but none of them has the BIG bug.

But I received today HP 35 Red Dot with super low and "jubilee" serial number 00100 from a previous owner, a former HP sales manager. This Red Dot has both known bugs and the new BIG bug, too! Not surprisingly the BIG bug appears also as a result of the calculation

10 eex chs 11 ln

But erroneous results give also slightly different calculations such as

123 eex chs 12 ln