Opening and trying to fix a leaked HP 48S

[ytterligare]

Long story short : after 6 years and a divorce, finally I got again ownership of some boxes of "toys" I left in the old home where ex-wife reside. Don't ask, I know.
BTW, a box contained my collections of HP Calculators...I'd say a pretty good collection af almost all HP models.
Now, call me fool, or maybe it was just the hurry to quit, during the inspection I noticed that two calculators, a 48sx and a 48s still had the battery in...and they leaked. The horror.

Well, I estimated on a scale of 1 to 10, the damage was 4 on the SX and 5 on the S.
After scratching and polishing the battery terminal and the plastic with Isopropylic alcohool, the sx came back to life...the joy !
But...the 48s did not : no matter how well I did polish it, the damn thing won't turn on.
I have tried any known combination of paper clip and key pressed, plus tried alkaline and eneloop batteries, no joy.
I also watched a couple of video claiming that some old 48's refuse to turn on unless you press a zone inbetween the "A" and "B" keys....tried several times with no luck.

Decided then to open the calc, and check if the leakage did something to the PCB.
Came across this article which (almost) carefully explained how-to. :

https://documents.epfl.ch/users/f/fr/fro...20HP48.pdf

Well, I generally I'm afraid to open old calcs, but what did I have to lose..? So....






As you can see, I made through step 1 and 2 somehow...but...step 3 seems impossible to me..: how do I "First, unhook the positive battery terminal from the battery compartment. " ???
I'm stuck. I cannot seems to be able to pop the remaining 6 rivets...partially because I believe the battery terminals are preventing it..





Anyone can explain how to "unhook the positive battery terminal" ? Any other suggestions ?

And anyway, how is it possible that the leakage did something to the pcb if the pcb did not reach that far, under the batteries ( see lower part on following internet pic ) ?



Thank you
A.

P.S: sorry for my writing, I'm not english native speaker.

[mfleming]

(11-17-2020 08:51 PM)ytterligare Wrote:  Anyone can explain how to "unhook the positive battery terminal" ? Any other suggestions ?

The positive battery terminal has a hole in the back, and a plastic hook in the wall behind it goes into the hole. You need to lift up the terminal to free it from the hook and then pull the terminal away from the wall so that the terminal clears the hook when you lower the terminal back down.

Try pulling the terminal away from the wall using a small screwdriver blade or similar tool inserted in the top loop of the terminal. You will feel the hook stopping the movement, and you can probably see how the hook holds the terminal in place. A thin blade inserted between wall and terminal will stop at the hook instead of going all the way down to the bottom of the compartment. Unhooking is basically an upward then forward movement. Rreverse the movement when reassembling.

HTH!
~Mark

[franz.b]

the misfire could depend on the common degradation of the rubber pressing the keyboard contacts. You have surely heard of it. Remove the pcb, clean the contacts and replace the rubber. On an hp27s I have recently had the same problem, if you are not sure if the contact is good follow the contact tracks of the on key and, once the pcb is powered, short circuit them to see if the calculator turns on.

[OtakuLab]

(11-17-2020 08:51 PM)ytterligare Wrote:  And anyway, how is it possible that the leakage did something to the PCB if it didn't reach that far under the batteries?

That depends on the position in which your calculator was kept all the time. I hope for you that this has not resulted in any further damage and that a thorough cleaning is sufficient.
__________________________________
"Move forward and do what you think is best. If you make a mistake, you’ll learn something. But don’t make the same mistake twice." - Akio Morita

[CMarangon]

Hello!

Previously in HP Museum Forum.

>>>And anyway, how is it possible that the leakage did something to the pcb if the pcb did not reach that far, under the batteries.

And now the conclusion:

Well, the first thing I would do would be to reset the calc. Thing that I suppose you already have made a lot of times.

I don't know how can be possible dont work after leakage, but I have lost radios, calculators and some electronics due of leakage of batteries, they was 5 or 6 electronic equipment. For it it is recommended to remove batts, when calc reamins a long time with no use.

I suppose that when batts leak, appears a kind of "paste" and it also generate some kind of acid gas that corrodes some terminal or even some track of PC Board.
This paste also can make a short circuit between tracks. Even moisture can do it, so much more the "paste"

First thing you can do is to use a multitester and test if terminals are short circuited or have
infinit resistence (circuit is open).

Well, this demmands advanced knowledge of Electronics. I have a basic knowlegde of Electronics, which allows me to assemble kits and make some power source and home low power speakers systems.

So, the best I can do is wish to you good luck!


By the way, where is that engeneer that used to write...
"I don't speak for HP when I post here" ?

[OtakuLab]

(11-18-2020 03:54 PM)CMarangon Wrote:  I don't know how it is possible that they will not work after a leak, but I have lost radios, calculators and some electronics to leaking batteries, so 5 or 6 electronic devices. For this reason, it is recommended that you remove all batteries if your device will not be used for a longer period of time.

I suppose when batteries leak they lose a kind of "paste" and there is also a kind of acid gas that corrodes a terminal or even a conductive track on the circuit board. This paste can also cause a short circuit between conductive tracks.

That would already describe the problem sufficiently. Even if a device remains in operation for a long time, the batteries should always be replaced after three years at the latest. Fresh batteries also ensure increased operational reliability and you save yourself such trouble. The TO has surely recognized this by now.
__________________________________
"Move forward and do what you think is best. If you make a mistake, you’ll learn something. But don’t make the same mistake twice." - Akio Morita

[ytterligare]

Ok, first thing I wanted to thank everybody that took time to reply to my post.
Well, finally I was able to "crack it open", and gain access to the innards, and what I saw was not good at all.






And yes, the PCB itself was not harmed by the leakage, but the keyboard plate surely was !!
Furthermore, the "ON" key happen to be right above the leaked surface...my suspect is that some acid did enter inbetween the keyboard layers.

To exclude any battery terminal involvement, I powerd the calc with 4.5 V from an external source and it did not power up...



Now, hoping to help gaining access to the keyboard, I removed the main PCB by unlatching it from the screen/keyboard/frame bending the six metal latches, four of which probably are needed to power the PCB.



...but the keyboard seems not accessible at all, unless I pop all or partially the brown plastic rivets that set the internal metal keyboard plate in place.
Once popped, if I'll ever be able to do it, I suspect the keyboard will be too lousy to guarantee the right tightening .

So, what to do now ?
I was think to first try to dunk the leaked corner in pure vinegar or solution and let it infiltrate in the keyboard layers, the same way the alkaline substance did...
Leave it there for few hours/days ( any suggestion ?) then let it dry up and submerge it into isopropyl alcohol for a while...

Is it a "no-no"...?

Andrea

[franz.b]

in my opinion, before trying to clean the keyboard, make sure it turns on. To do this, follow the contact tracks of the ON key and, after having powered the pcb mounted on the plate in order to make the screen work, short-circuits the two tracks to see if the calculator is alive.
To do this, find a diagram of the keyboard contacts, but I suppose the first two contacts are those of the ON key.
to clean the keyboard I do not think it will be enough dips in various substances but it is worth a try. I fear the conductive tracks are corroded, but you can, I believe, reconstruct them with a conductive marker. opening it will be a lockdown job... and even worse closing it.

[ytterligare]

(11-18-2020 07:56 PM)franz.b Wrote:  in my opinion, before trying to clean the keyboard, make sure it turns on. To do this, follow the contact tracks of the ON key and, after having powered the pcb mounted on the plate in order to make the screen work, short-circuits the two tracks to see if the calculator is alive.
To do this, find a diagram of the keyboard contacts, but I suppose the first two contacts are those of the ON key.

Hello Franz, thank you for your insight.
I'm venturing in unchartered territory ( for me ) so, here's what I found : it is for a 48G but I believe that should apply to 48S too.

This should be the schema for the keyboard contact :



I already have checked and the far left contact ( far right on the above picture ), once the calc is closed, is electrically coupled to a track that merges with GND, so the far left -1 should be the ON key.

Now, given the following drawings :



and



Am I to assume that in order to power on the calc I should short the last two contacts ( 1+2) or what ?

Anyhow, once the PCB is in place to ensure the screen is connected, it is not an easy task to find where, on the opposite side ( the one facing me ) to pinpoint the two contacts, if replicated....

[franz.b]

(11-18-2020 11:42 PM)ytterligare Wrote:  Am I to assume that in order to power on the calc I should short the last two contacts ( 1+2) or what ?

Anyhow, once the PCB is in place to ensure the screen is connected, it is not an easy task to find where, on the opposite side ( the one facing me ) to pinpoint the two contacts, if replicated....

yes, I assume contacts 1-2 are correct.
to find the correspondence on the dark side of the moon... no! of the pcb follow with the tips of a multimeter the track starting from the keyboard contact until the track spills over to the other side, and mark that point with a marker on the side that you see when the pcb is mounted. you do it for both tracks so you have two points to touch. I did this for my 27s and it was a useful procedure.

[ytterligare]

(11-19-2020 07:14 AM)franz.b Wrote:  I did this for my 27s and it was a useful procedure.

Ok Franz, I did not want to bend the metal latches agains, so I used somthing less "professional" which was not as effective ( and got my fiancee angry ) :



as a matter of fact, the contact was less than perfect, but apparently, shorting the spilling tracks on the other side of the moon/pcb did the trick the message is clear : she's alive !!



Hopefully by latching properly the PCB, then the screen will be ok again...but the main problem remains : how to clean the contacts....there's a big chance that the keyboard tracks are irremediably corroded, and they are almost impossible to reach if not popping a lot of plasic rivets which are then not replaceable....

[franz.b]

it seems like a great idea! I had the same problem when I did it and I invented to put pins under the hooks, but your method is simpler and more practical! now at least you know it's alive in fact and worth the keyboard repair.
I don't know exactly how to proceed, but I have seen disassembled keyboards online and maybe you can find some guides. I think you need to cut all the plastic rivets. Then to reassemble I think I saw a hot glue method, but I am thinking that I would try to put plastic or metal strips with holes in correspondence of the rivets, I would drill the rivets partially and I would screw in some small screws.

[ytterligare]

Those are all good ideas, but first thing first, I'll try to get rid of the crystals by submerging the "offended" part in a solution of Vinegar (60%) and water (40%).
Let's see if with all the cleaning I reach some result.. ;-)

Thank you !

A.

[franz.b]

certainly, the chemical method works you solve with a few flicks. I would, however, remove the external keyboard cover which is simply glued. you could ruin it with vinegar and then the liquid will penetrate better in the depths.

[cruff]

(11-19-2020 06:36 PM)ytterligare Wrote:  Those are all good ideas, but first thing first, I'll try to get rid of the crystals by submerging the "offended" part in a solution of Vinegar (60%) and water (40%).
Let's see if with all the cleaning I reach some result.. ;-)

Not sure why you would need to dilute the vinegar, you could use it straight. Unless of course if you really have a concentrated acetic acid instead, which should be diluted.

[lyuka]

As the keyswitch of the pioneer series are made of carbon contact,
any kind of acid treatment won't work, or do harm to the membrane circuit.


HP Pioneer series calculator internals

It was possible to transplant the membrane switch, but it was very difficult.
Deterioration of the membrane switch itself is wear and cannot be repaired.

[ytterligare]

(11-20-2020 05:00 PM)lyuka Wrote:  As the keyswitch of the pioneer series are made of carbon contact,
any kind of acid treatment won't work, or do harm to the membrane circuit.

It was possible to transplant the membrane switch, but it was very difficult.
Deterioration of the membrane switch itself is wear and cannot be repaired.

I give up.
After keeping few hours the lower part of the keyplate submerged in vinegar 6% ( lot of bubbling and chemical reaction), rinsing with water and isopropyl alcohol, the first row of the keyboard is still dead, no sign of life whatsoever.

Alas, one 48s less, survived in the world
Will probably sell it as scrap for repair on ebay.

Thank you everybody !
A.

[mfleming]

Console yourself with the knowledge that you've gained valuable experience in doing this sort of repair. Most people, including myself, generally mess up a case opening on the first try and end up with a calculator that just doesn't fit back together right. At least you won't experience that disappointment

Now and then 48s(x) calculators sold as "for parts or not working" show up at very low prices, often less than $20 USD. They just need to be opened and have their keyboard pad replaced to be useful again.

Good luck with the next repair!
~Mark

[acoto]

Before closing the calculators, may like to buy a HP 39G, 40G or 49G on Ebay, and transplant the black display to the S or SX. ( Note, G+ or GS displays will not work, need a Saturn based donor, ARM changed the hardware architecture a lot and display is not compatible)

[franz.b]

I would try to completely disassemble the keyboard, you wouldn't lose anything if you still consider it a scrap.