The first sign of HP-35’s ln(2.02) bug

[Matt Agajanian]

Hi all. How long after the 35 was released did the discovery of the bug occur?

How long did it take after the bug was corrected were new 35s released?

[teenix]

I read the fascinating document "Remembering The HP-35A" by Richard J. Nelson and this is a bit of what he had to say...

"Exactly when the bug was found I don’t remember or have, at this time, documentation to determine the date."

There are a couple of scans of the HP explanation and voluntary recall documents included but no dates.

I found a mention of a Red Dot model being repaired on May 4 1973.

cheers

Tony

[RMollov]

why is 1 enter + = 2

[teenix]

(09-09-2022 06:25 AM)RMollov Wrote:  why is 1 enter + = 2

It's all to do with how the 4 level stack works.

All cleared

T 0
Z 0
Y 0
X 0 Displayed value


Key = 1, T is lost, Z -> T, Y -> Z, X -> Y, 1 -> X

T 0
Z 0
Y 0
X 1 Displayed value


Key = ENTER, T is lost, Z -> T, Y -> Z, X -> Y, X = No Change

T 0
Z 0
Y 1
X 1 Displayed value


Key = +, Y + X -> X, Z -> Y, T -> Z, T = No Change

T 0
Z 0
Y 0
X 2 Displayed value


Examples

T 12
Z 4
Y 9
X 2 Displayed value

Key = -, Y - X -> X, Z -> Y, T -> Z, T = No Change

T 12
Z 12
Y 4
X 7 Displayed value

Key = x, Y x X -> X, Z -> Y, T -> Z, T = No Change

T 12
Z 12
Y 12
X 28 Displayed value





Lots of info on the stack and RPN in the user manual and could be on rear calculator label.

cheers

Tony

[AndiGer]

(09-08-2022 11:36 PM)teenix Wrote:  I read the fascinating document "Remembering The HP-35A" by Richard J. Nelson and this is a bit of what he had to say...

"Exactly when the bug was found I don’t remember or have, at this time, documentation to determine the date."

There are a couple of scans of the HP explanation and voluntary recall documents included but no dates.

I found a mention of a Red Dot model being repaired on May 4 1973.

cheers

Tony

Michael read the date he is used to in European format, but in US format it is April 5 in 1973 ... arrived at HP at April 4.

Andi

EDIT: Interesting (to me) is that the so-called "big bug" was known at HP as one can see from the errata sheet.

[pauln]

Are HP-35's with the bug more rare and presumably more valuable?

[AndiGer]

(09-09-2022 05:00 PM)pauln Wrote:  Are HP-35's with the bug more rare and presumably more valuable?

Good question ... or is a repaired red dot the more seldom and therefore the more valuable version? I have no idea.

[Didier Lachieze]

(09-09-2022 05:00 PM)pauln Wrote:  Are HP-35's with the bug more rare and presumably more valuable?

They are only a small fraction of the overall HP-35’s: produced in 1972 and not sent back to HP for the recall or later on for repair. I should have some rough estimation of the % of the HP-35’s remaining with the bug but I don’t have access right now to my virtual collection.

[Kretsh]

I owned one of the first manufactured HP35 calculators and received a certified letter from HP, just 1 or 2 months after receiving the calculator, alerting me to the computational bug: if e (epsilon) is raised to the 0.000200 exactly (or something like that) it would give an incorrect answer.

The letter stated, that if I could not "live" with this error, I could take it to my local HP center and they could change out the ROM chip with the corrected algorithm, at no charge. That I did as the HP center was just up the road in Dallas, from where I worked in Dallas. The ROM was swapped out and HP also changed the slider/button on the On-Off switch to designate the update. This new sliding button was only fractionally higher.

Kretsh
Ex-Texas Instrument Employee
1971-1974

[CY-CL]

Actually 86 HP-35 RedDot's are known still to exist:

78 still working
30 still in original condition
49 have the 2.02 Bug
5 have the Big Bug

[EdS2]

The archived site of the late Jacques Laporte says this:

Quote:In 1972 (probably by the fall of 1972) , when the problem was discovered, HP had sold some 25000 units, but only 5000 machines were returned to have their 3 ROM chips exchanged.

Bug free units were shipped starting January 1973 :
- Serial Number 1302AXXXXX (second week of 1973) and higher are error free,
- before that all units needed modifications.
The recall process started in February 1973: "We have taken immediate steps to correct the problem, and will have replacement read-only-memories available sometime after the first of the year" wrote Ron Stevenson Customer Service Manager, to the "Dear HP-35 Owner"

[Didier Lachieze]

I'm back home and have checked my Virtual Collection.

(09-10-2022 07:43 AM)EdS2 Wrote:  The archived site of the late Jacques Laporte says this:

Quote:In 1972 (probably by the fall of 1972) , when the problem was discovered, HP had sold some 25000 units, but only 5000 machines were returned to have their 3 ROM chips exchanged.

Bug free units were shipped starting January 1973 :
- Serial Number 1302AXXXXX (second week of 1973) and higher are error free,
- before that all units needed modifications.
The recall process started in February 1973: "We have taken immediate steps to correct the problem, and will have replacement read-only-memories available sometime after the first of the year" wrote Ron Stevenson Customer Service Manager, to the "Dear HP-35 Owner"

This is confirmed by appendix D of Remembering the HP 35A, in particular Fig. D4 & D5.

So assuming that all units prior to the date code 1302A were produced with the bug, this represents at least 77194 units (39071 1143A + 24776 1230A + 13347 1249A). This assumption is confirmed by the highest serial number with the bug in my virtual collection: 1249A 11780.

If, when the bug was discovered, there was 25000 units already produced*, this should have been in late August 1972.

I have in my Virtual collection 568 units with date code 1143A, 1230A or 1249A.
120 confirmed with the bug, 53 confirmed without the bug and 395 unknown.
So based on this small sample 69% of the units with a known status have the bug and 31% don't have it. This would indicate that a lot of the units recalled were not sent back to HP, 31% of 77194 represents 23930 units returned to HP.

Now for the Red Dot units, merging my database and the one from http://hp35.wz.cz/ (which I didn't so far to ensure that my collection methodology would be the same for Red Dot and non Red Dot units and to avoid an over-sampling of Red Dots) I come to these numbers:

• 113 units with a Red Dots serial number
• 95 units with the Red Dot on the front
• 52 Red Dot units with the bug
• 27 Red Dot units without the bug
• 16 Red Dot units with unknown status
  

*this 25000 number is floating around on the internet, but I’ve never seen any historical evidence from HP internal documents. So if someone can provide such evidence I would be more than interested.