HP-29C now has dots across the display...

[Bill Duncan]

Has anyone seen this before?

My beloved HP-29C that I've owned since the 70's now has a string of dots across the display, both in normal and programming mode. It's also acting strange. The dots are not full brightness, but appear to be flickering fast.

I've tried removing the batteries, shorting the terminals to get an error display back. That seems to have corrected it for now.

Is this a sign of impending failure? Perhaps some electrolytic capacitor failing? Maybe an overvoltage thing? (I'm using fresh alkaline batteries.)

[rprosperi]

All the dots on mean the batteries are low. Even though they may seem to be "new", more than likely they sat on the shelf a long time, or maybe one of them is failing, or something like that. Get some new, new batteries and I think your 29C will be OK.

Good luck and report back to confirm results.

[Geoff Quickfall]

Sage advise from a long time HP expert.

All dots is a low battery display.

Also, check connection at the base of the PCA that the battery pack engages. Use a pink eraser to clean up the connections.

[Bill Duncan]

Thanks all.

AFAIK, the low power display is a single blinking decimal point, where it should be. In the 45 years I've owned this baby, I'd never seen this behaviour for low battery power.

Yes, brand spanking new Duracell double A batteries. Tested. I'd cleaned the contacts with the pink eraser when I'd changed the batteries to be sure.

Puzzling, but it appears to be fine for now.

Thanks!

[rprosperi]

(09-02-2024 07:21 AM)Bill Duncan Wrote:  Yes, brand spanking new Duracell double A batteries. Tested. I'd cleaned the contacts with the pink eraser when I'd changed the batteries to be sure.

Good news that the problem appears to be gone. However I advise you to replace those Duracell batteries with another, more reliable brand. Duracells have long been known to be of poor quality and they often leak and damage units. As far back as the mid-80's, HP advised all calculator owners to NOT use this brand, and in the case of the 71B (which used 4 x AAA cells) even advised customers to get reimbursed by Duracell for any needed repairs, as they were known to leak and cause damage.

It sounds like you take good care of your toys, so please heed this advice and keep your 29C happy and safe for years to come.

[teenix]

The 19C and 29C had a blinking decimal point for low battery, while all the other Woodstocks had all the decimal points lit like the Classics, HP-65 and Spice. The other models, like HP-67 and Topcats had a separate LED for the indication.

I'm not sure why all the DP LEDs would light in the HP-29C. Maybe the offending voltage was messing with the ACT which gets its low power indication from the power monitor IC.

cheers

Tony

[AnnoyedOne]

(09-02-2024 01:32 PM)teenix Wrote:  Maybe the offending voltage was messing with the ACT which gets its low power indication from the power monitor IC.

Interesting. I don't own (I've seen the prices of working ones!) a HP-29C (but have a PC simulator), or any other Woodstock model, but I think my uncle did (a 25 or 29C). He was an engineer too but of a different type. Both are long gone.

If I had the 29C in question I'd hook up a variable power supply and slowly dial the voltage down and observe what happens.

Perhaps the 29C has an initial blinking DP but shifts to all DP's at a lower voltage level.

The engineer in me wants to experiment

A1

[Maximilian Hohmann]

Hello!

(09-02-2024 01:50 PM)AnnoyedOne Wrote:  The engineer in me wants to experiment

Not with my HP-29C! That was one of the most difficult to find (and costly) calculators in my collection and I would not make experiments with it, especially not experiments with the supply voltage... Normally undervoltage should not cause damage to the circuitry, but who knows with the flimsy power supply design of the Woodstock series.

In my experience, in most instances of intermittent malfunctions, the silver rivets on the PCBs which connect to the battery are the culprits. Silver oxidises quicky, especially when traces of leaked battery chemicals are present, and siver oxide is a poor electrical conductor. Sometimes it does not help to clean the contact faces because the connection between the rivet and the PCB trace is oxidised. It is rather difficult to re-solder but worth a try.

As an alternative to regularly clean the contacts they can also be gold plated. For that a galvanising pen, gold electrolyte and a power supply are required. The cost (without the power supply) is in the order of 50$/Euros and probably a little high for treating a single calcaulator. And one has to be very careful when applying the voltage for galvanising the contacts. 4V is required which is already too much for the calculator. So make sure where to connect the pen.

My pragmatic solution is to bypass the awful contacts altogeher. On some woodstocks (more will follow...) I simply glued and soldered a little 2-pin JST connector to the calculator PCB and a little cable with the corresponding plug to the battery pack (something like this: https://www.ebay.com/itm/253329082894). No problems ever since.

Regards
Max

[AnnoyedOne]

(09-02-2024 03:05 PM)Maximilian Hohmann Wrote:  Normally undervoltage should not cause damage to the circuitry, but who knows with the flimsy power supply design of the Woodstock series.

Yeah. I've heard about that. Look at it wrong and pzzt! My simulator was much cheaper (free) and nothing to fail.

From what I've seen most of these old Woodstock's have often been repaired in one way or another. A new (board to replace the ACT) CPU, batteries, power supply etc.

For my "experiment" I'd use a PSU with a current limit (set low) just in case.

A1

PS: I just reread the "manual" and my 29C "simulator" is actually an emulator (of original HP microcode). Even better!

[Maximilian Hohmann]

Hello!

(09-02-2024 03:29 PM)AnnoyedOne Wrote:  From what I've seen most of these old Woodstock's have often been repaired in one way or another. A new (board to replace the ACT) CPU, batteries, power supply etc.

The reason for the repairs is usually battery leakage. The older calculators (like the HP-35) had a fully enclosed battery compartment inside wich an almost completely enclosed battery pack was placed. There is almost no chance for battery leakage to reach the calculator internals. The woodstocks still have that separate compartment but it has some openings. And the battery packs are not fully closed either. This results in an almost 100% chance of battery fluids reaching the PCB, componets and keyboard internals. Then there is this mediocre charging circuit which uses the battery itself for voltage regulation. A dead or poorly connected battery will result in overvoltage applied to the circuitry. No problem until the calculator is switched on. After that, death in an instant. And there is no proper charging current regulation with the power connector either. People who have the habit of leaving their devices plugged in all the time (no problem with modern devices apart from the unnecessary engery waste, but a cardinal sin when operating vintage machinery) will kill the batteries quickly. They keep charging and charging to the point when they start squirting their contents over those little microcontrollers...

(09-02-2024 03:29 PM)AnnoyedOne Wrote:  For my "experiment" I'd use a PSU with a current limit (set low) just in case.

That's not so easy because the LEDs require quite some current. Enough to fry the more delicate integrated circuits in case something is wrong.

Regards
Max

[Bill Duncan]

Interesting Bob, thanks.

Oddly, I've always had the opposite opinion of Duracell. Decades ago I had issues with Eveready batteries leaking, so I switched to Duracell and have never had problems with them. Odd.

These cells were new, and registering with 1.65 volts which is what I'd expect. If anything, I thought it might've been over voltage, as the original NiCD cells are around 1.25v.

[Geoff Quickfall]

I have only had Duraleaks destroy my electronics. Especially my mini mag lights, which were in my flight bag. I think multiple pressure changes during climb and descent weakened the seals causing them to leak. Switched to Eveready and never had a problem again.

Had to drill the durlaleaks out of the maglight battery tube to get them free!!!

Cheers.

I cannot test the low voltage, single period display on my bench as I only have 29C Panamatik versions with GPS, RTC and IR printing.

Can any one else do the test?

Cheers

[AnnoyedOne]

(09-02-2024 05:00 PM)Geoff Quickfall Wrote:  ..I only have 29C Panamatik versions...

For those that don't know the Panamatik board uses a Microchip PIC microcontroller to emulate the original HP microcode in Woodstock model HP calculators. It is used to replace a burnt out (destroyed) ACT CPU (from a power supply issue). It also adds additional functionality.

However as a modern design it doesn't behave exactly like an original ACT CPU hardwarewise. Performing my "low voltage" test on such a unit thus won't tell you much.

A1

[Geoff Quickfall]

I have only had Duraleaks destroy my electronics. Especially my mini mag lights, which were in my flight bag. I think multiple pressure changes during climb and descent weakened the seals causing them to leak. Switched to Eveready and never had a problem again.

Had to drill the durlaleaks out of the maglight battery tube to get them free!!!

Cheers.

I cannot test the low voltage, single period display on my bench as I only have 29C Panamatik versions with GPS, RTC and IR printing.

Can any one else do the test?

Cheers

[teenix]

(09-02-2024 04:12 PM)Bill Duncan Wrote:  Interesting Bob, thanks.

Oddly, I've always had the opposite opinion of Duracell. Decades ago I had issues with Eveready batteries leaking, so I switched to Duracell and have never had problems with them. Odd.

These cells were new, and registering with 1.65 volts which is what I'd expect. If anything, I thought it might've been over voltage, as the original NiCD cells are around 1.25v.

I did some voltage measurements on a HP-45 board with raised battery voltages and the VCC voltage was the one that rose, the others seemed stable. I haven't tested a Woodstock.

I don't know if this raised operating voltage caused issues with the circuitry.

cheers

Tony

[BobVA]

(09-02-2024 05:57 PM)Geoff Quickfall Wrote:  Can any one else do the test?

What won't I do for science? :-(

Blinking dot at 2.33v, then lights out shortly after, no other display artifacts. If I crank up the current limit it will stay on to lower voltage (less than 2.00) but the current draw goes up alarmingly (160 mA is where I quit, for fear of damaging the 29's power supply), so exactly how low it could go is up for speculation. Practically probably not of much interest, since I doubt AA's could produce that amount of current if they were dropping in voltage.

Keeping the voltage constant at 3.00 and lowering the current limit resulted in a hard cutoff at 123 mA. I might have seen a string of decimal points across the display once when that happened, for a split second. Perhaps some edge interaction between the multiplex circuitry and the low voltage warning?

[rprosperi]

(09-02-2024 01:32 PM)teenix Wrote:  The 19C and 29C had a blinking decimal point for low battery, while all the other Woodstocks had all the decimal points lit like the Classics, HP-65 and Spice. The other models, like HP-67 and Topcats had a separate LED for the indication.

I'm not sure why all the DP LEDs would light in the HP-29C. Maybe the offending voltage was messing with the ACT which gets its low power indication from the power monitor IC.

cheers

Tony

Thanks Tony! Sorry for my poor advice above, I clearly mix up my Woodstocks! I will let my 21 and 27 get low enough to see the dots, but my 29C always has either no batteries or full batteries installed. After surviving a heart (ACT) transplant, I treat my 29C very gently...

[AnnoyedOne]

(09-03-2024 11:59 AM)rprosperi Wrote:  ...I treat my 29C very gently...

Wise man! I recommend cotton wool or other padding during storage

A1

[AnnoyedOne]

(09-03-2024 02:07 AM)BobVA Wrote:  What won't I do for science? :-(

Thanks for putting your 29C in deadly peril and "taking one for the team" Interesting results.

So all dots may've resulted.

A1

[teenix]

(09-03-2024 11:59 AM)rprosperi Wrote:  Thanks Tony! Sorry for my poor advice above, I clearly mix up my Woodstocks!

Me too :-)
I had to look up the Woodstock info and I verified with the emulator.

I guess the engineers decided on a new approach for the 19C and 29C.

cheers

Tony