Any standard failure modes in old TI calculators?

[Joerg]

Gents,

My trusty old 1974-vintage Texas Instruments SR-50 is beginning to
sputter <sniffle>. It turns on and at first only reacts to very few keys
like "Clear" or "Equals". When pressing those it does erratic things
such as incrementing by some large number until overflow occurs.
Gradually its marbles come back, more and more keys respond seemingly
correctly and it does math correctly. But of course it can't be trusted
much in this sort of state.

Inside it looks clean, signals are healthy, solder joints look ok. Not
sure what's underneath the membrane keys, I haven't dared to tear that
apart because then it's probably all toast.

Does anyone remember if there's a standard failure mode on these I could
look for? Or maybe it doesn't like very long periods of non-use?

What amazes me is how much power they use. With all LEDs on it slurps
120mA out of a 3.8V supply. Cleared down to just one zero it's still 80mA.

 

[Spehro Pefhany]

Gents,

My trusty old 1974-vintage Texas Instruments SR-50 is beginning to
sputter <sniffle>. It turns on and at first only reacts to very few keys
like "Clear" or "Equals". When pressing those it does erratic things
such as incrementing by some large number until overflow occurs.
Gradually its marbles come back, more and more keys respond seemingly
correctly and it does math correctly. But of course it can't be trusted
much in this sort of state.

Inside it looks clean, signals are healthy, solder joints look ok. Not
sure what's underneath the membrane keys, I haven't dared to tear that
apart because then it's probably all toast.

Does anyone remember if there's a standard failure mode on these I could
look for? Or maybe it doesn't like very long periods of non-use?

What amazes me is how much power they use. With all LEDs on it slurps
120mA out of a 3.8V supply. Cleared down to just one zero it's still 80mA.

How are the batteries? On some of those old calculators they worked to
filter the power supply. Have you tried powering it with a lab supply?
That one has 3 NiCd cells, right?

I've still got an SR-52 (the one with the magnetic card reader) around
somewhere. But I never liked TI's keyboards.. felt cheap and were
never 100% sure one and only one keystroke got through. Went HP and
never looked back.


Best regards,
Spehro Pefhany

 

[Joerg]

How are the batteries? On some of those old calculators they worked to
filter the power supply. Have you tried powering it with a lab supply?
That one has 3 NiCd cells, right?

Yup, three cells and they had leaked so they are on the way to E-waste.
Luckily these leak into the bottom shell, not the electronics. Right now
it's on a bench supply and a 100uF cap where the battery was. If I get
it going again I'll give it three new NiMH cells and put them on a C/10
or a little less trickle. I know that's not too healthy for them but
they seem to last a long time in cordless phones that way.

I've still got an SR-52 (the one with the magnetic card reader) around
somewhere. But I never liked TI's keyboards.. felt cheap and were
never 100% sure one and only one keystroke got through. Went HP and
never looked back.

Same here, the HP-11C is my office workhorse. But I keep the SR-50 at
the lab. Mainly because my parents gave it to me in the mid 70's when it
was unbelievably expensive. I am forever thankful to them for that
because it opened the door to many intricated projects. And because of
the rather good quality LED it is much easier to read in bright light.
No RPN though which throws me a curve every time.

Right now it's totally on the fritz again, totally erratic :-(

 

[Joerg]

<http://www.datamath.org>
Note that there are about 6 mutations of the SR-50.
<http://www.datamath.org/Album_Sci.htm>
<http://www.datamath.org/SCI/WEDGE/sr-50.htm>

I've got the original, way on the left.

My guess(tm) would be that you have some manner of liquid or corrosion
under some of the dome switch keys. The changes in key behavior are
caused by electrolytic action. Disassembling the keyboard dome
switches is a mess.
<http://www.datamath.org/Story/JPEG_Klixon.htm#1KS>
However it's about the only way you're going to clean out the
accumulated corrosion under the dome.

Possible. What puzzles me is that the thing worked fine 15 minutes ago
and now is all dead again on the keys.

Nope. I can be done. Don't expect to re-use the clear plastic cover.
I just used clear packing tape. I made a 1:1 Xerox copy of the switch
contact board. I placed the packing tape backwards (sticky side up)
over the Xerox copy, and placed the domes in their proper location. I
then carefully stuck the tape onto the contact board. The first time
I did this, it took about 5 tries to get it right.

Yikes, that looks like a bear of a job.

Not that I know of. I've seen keyboard problems, dead leaky
batteries, and flimsy plugs on the chargers, on the older TI
calculators. However, I've worked on many more HP calculators than
TI.

The charger plug was the first thing that went but that was easily fixed.

Good luck.

Thanks. Not looking forward to this one though.

 

[Spehro Pefhany]

Possible. What puzzles me is that the thing worked fine 15 minutes ago
and now is all dead again on the keys.

If one of the keys is 'stuck' on due to (say) spilt wine or perhaps a
fine single-malt scotch (not that I'd have any direct experience with
that sort of disaster), then the scanning stops and the calculator may
appear dead, or may do something else (such as count up) if it
triggers auto-repeat.

If something like that happened, you might be able clean it out or
not. Once it 'tracks' the conductivity seems to stay. If you can get
it open without wrecking it, it should be possible to clean. IIRC it's
metal clicky domes in a polyester overlay on the TI calc's but it's
been a LONG time, my memory may have faded.

I wonder if putting the keyboard through the dishwasher and then
subjecting it to a vacuum to dry it inside would be less risky than
trying to open it up.

I've also seen the keyboard PCB crack on remotes, but that's when they
use the crappiest paper-based phenolic you can buy. Your TI probably
has nice FR4.

Some remote controls just seem to spontaneously fail:
http://www.speff.com/panasonic.jpg




Best regards,
Spehro Pefhany

 

[Spehro Pefhany]

If the keyboard domes were vented, that might work. However, without
a way for the suds to get under the domes and back out for drying, I
don't see it doing much good.

Right, the polyester is glued right to the PCB, right? Should seal it
against ingress of fluids, but also prevents cleaning.

Yep, it's FR4. Later cost reduced models were phonolic.


What size nail were you hammering with the remotes?

Bitch chewed it. Female Golden Retriever, to be specific. I'm sure
that there are people who would try to return something like that
under warranty. After all, Panasonic didn't have to make it so tasty
and bone-shaped.


Best regards,
Spehro Pefhany

 

[Uwe Hercksen]

No RPN though which throws me a curve every time.

Hello,

if you are really used to RPN, you have problems using another
calculator without RPN.

Bye

 

[Joerg]

Hello,

if you are really used to RPN, you have problems using another
calculator without RPN.

True, and I do. But it does train the brain so I like the challenge,
just like driving a car with right-hand steering. I also am sometimes
forced to use non-RPN because I often don't take my HP on the road. The
usual, some innocent EMC job and then they need me to also fix internal
noise and calculate a filter, on whatever calculators they have. Lately
on guy handed me his smart phone turned calculator. No RPN either and
then you have no time to load a new app.

No joke: This can help in at least pushing out diseaese like Alzheimer's.

 

[Joerg]

If one of the keys is 'stuck' on due to (say) spilt wine or perhaps a
fine single-malt scotch (not that I'd have any direct experience with
that sort of disaster), then the scanning stops and the calculator may
appear dead, or may do something else (such as count up) if it
triggers auto-repeat.

If something like that happened, you might be able clean it out or
not. Once it 'tracks' the conductivity seems to stay. If you can get
it open without wrecking it, it should be possible to clean. IIRC it's
metal clicky domes in a polyester overlay on the TI calc's but it's
been a LONG time, my memory may have faded.

Unfortunately that's the case. Jeff has described how to fix it but it
seems to be a rather onerous procedure.

I wonder if putting the keyboard through the dishwasher and then
subjecting it to a vacuum to dry it inside would be less risky than
trying to open it up.

I've also seen the keyboard PCB crack on remotes, but that's when they
use the crappiest paper-based phenolic you can buy. Your TI probably
has nice FR4.

Some remote controls just seem to spontaneously fail:
http://www.speff.com/panasonic.jpg

That looks like the "Golden Retriever effect"

 

[Joerg]

Take another calculator and hold down one key, any key. Now, try
punching the other keys and see what happens. On all 3 calculators in
my palatial office, they all refused to respond.


It's not too horrible. However, there is an alternate procedure that
might have some merit. Find a syringe with a rather fine needle. The
type used for insulin injections is about right. Clean out the
contents and fill it with alcohol or other evaporative solvent. Punch
the needle through the plastic over and UNDER the dome switch. Inject
the solvent, shake well, do your best to drain, and with luck, it
should work. You can dry drawing out the alcohol with the syringe,
but that will contaminate the contacts, requiring a purge between
every cleaning. My guess(tm) is that there's only one key that's
actually full of conductive corrosion.

At least there's a pinout for the scan chip:

http://www.datamath.org/Chips/TMC0501.htm

If you don't care about looks and resale value, you can drill a small
hole in the side (not the top) of the dome. It will take some care
not to hit the switch contact on the contact board or go through the
board. You can then use the syringe trick quite easily. Just wrap
some electrical tape around the drill to keep the drill from going in
too far. A Dremel tool should work.

That would be another option. I don't think these SR-50 have much of a
resale value and I am not concerned about that anyhow. I could just buy
a new calculator for $30-$40 but it has some sentimental value to me.
That sentimenatl value has its limits though

 

[Reinhard Zwirner]

John Larkin schrieb:

[...]

I have an HP35 that still works! And I still use it.

+1

Ciao

Reinhard

 

[mike]

Yup, three cells and they had leaked so they are on the way to E-waste.
Luckily these leak into the bottom shell, not the electronics.

I'd expect some evaporated electrolyte condensed somewhere it doesn't
belong.
I expect you'll have to clean all the boards very thoroughly.

 

[Jeroen]

John Larkin schrieb:

[...]

I have an HP35 that still works! And I still use it.

+1

In the case of an HP35, that would be '1+', I suppose.

Jeroen Belleman

 

[[email protected]]

True, and I do. But it does train the brain so I like the challenge,
just like driving a car with right-hand steering. I also am sometimes
forced to use non-RPN because I often don't take my HP on the road. The
usual, some innocent EMC job and then they need me to also fix internal
noise and calculate a filter, on whatever calculators they have. Lately
on guy handed me his smart phone turned calculator. No RPN either and
then you have no time to load a new app.

I carry my HP11C (emulator) everywhere I go. Well, I don't carry my
cell phone around the house. ;-) Loading a new app on my cell phone
takes all of 30 seconds.

I can't stand "algebraic" calculators. No two are the same. There are
always minor differences in the way operations are chained.

No joke: This can help in at least pushing out diseaese like Alzheimer's.

Yeah, suicide is faster.

 

[Joerg]

I carry my HP11C (emulator) everywhere I go. Well, I don't carry my
cell phone around the house. ;-) Loading a new app on my cell phone
takes all of 30 seconds.

Can't do that. My cell is a no frills phone-phone. No apps possible,
whatsoever. Kids think it's a steampunk model.

Slowly I get the feel this old SR-50 is toast. Even the pins of the SCOM
chip appear to have started some sort of dry-rot, all the way up to and
possibly into the chip.

[...]

 

[[email protected]]

Can't do that. My cell is a no frills phone-phone. No apps possible,
whatsoever. Kids think it's a steampunk model.

We had those until we took the plunge and bought Android phones in
January. We make very good use of them and haven't been sorry at all.

Slowly I get the feel this old SR-50 is toast. Even the pins of the SCOM
chip appear to have started some sort of dry-rot, all the way up to and
possibly into the chip.

Working as designed. ;-) I hated SR-50s. I would have stuck with my
slipstick if they were the only alternative.

 

[Martin Brown]

Gents,

My trusty old 1974-vintage Texas Instruments SR-50 is beginning to
sputter <sniffle>. It turns on and at first only reacts to very few keys
like "Clear" or "Equals". When pressing those it does erratic things
such as incrementing by some large number until overflow occurs.
Gradually its marbles come back, more and more keys respond seemingly
correctly and it does math correctly. But of course it can't be trusted
much in this sort of state.

They wear out after a certain amount of use. My SR-51 eventually expired
but largely because it fell off my belt clip and got run over.

I am surprised one lasted as long as this.

We once tormented one by turning the clock speed right down so you could
see the refresh. ISTR the TI patents are available (not sure if online)
so you should be able to get full circuit diagrams for the
implementation. Actually I think it was a 56 we did that too.

Inside it looks clean, signals are healthy, solder joints look ok. Not
sure what's underneath the membrane keys, I haven't dared to tear that
apart because then it's probably all toast.

I suspect wear and tear - nearly 50 years operation isn't bad. Most of
my calculators have died within a decade. Although I do still have a
working SR-59 and print cradle in the loft somewhere as I could not
bring myself to part with it.

Does anyone remember if there's a standard failure mode on these I could
look for? Or maybe it doesn't like very long periods of non-use?

I think it is the even longer periods of use inbetween that does for them.

What amazes me is how much power they use. With all LEDs on it slurps
120mA out of a 3.8V supply. Cleared down to just one zero it's still 80mA.

That is 6 segments on so 12mA a segment average. A "1" might well be
lower by a factor of 3 depending on how much juice the core takes.

Regards,
Martin Brown

(also a fan of TI calculators)

 

[Joerg]

They wear out after a certain amount of use. My SR-51 eventually expired
but largely because it fell off my belt clip and got run over.

I am surprised one lasted as long as this.

We once tormented one by turning the clock speed right down so you could
see the refresh. ISTR the TI patents are available (not sure if online)
so you should be able to get full circuit diagrams for the
implementation. Actually I think it was a 56 we did that too.

I suspect wear and tear - nearly 50 years operation isn't bad. Most of
my calculators have died within a decade. Although I do still have a
working SR-59 and print cradle in the loft somewhere as I could not
bring myself to part with it.

Yeah, one of the reasons I decided to hand it over to E-waste now is
that some of the domes appear to have a hole rubbed in, and not just
through the plastic. This calculator got some extreme workouts because I
did not have a computer for my early filter designs. My dad had an
IBM-5100 but it ran APL which I never really warmed up to.

I could have swapped the CLEAR and EQUALS domes with ARCTAN or other
lesser used ones. But there comes a limit. Walmart sells a nice HP
scientific calculator for around $50 and it has RPN.

I think it is the even longer periods of use inbetween that does for them.

That is 6 segments on so 12mA a segment average. A "1" might well be
lower by a factor of 3 depending on how much juice the core takes.

They were never very bright but nice small segments and very contrasty.

Regards,
Martin Brown

(also a fan of TI calculators)

Me, not so much, I have to say that afetr I married I began to prefer
HP. Feels like better quality and I started to really like RPN. My wife
brought an HP-11C into our marriage and that's the one I use right now.
We could never divorce because then I'd have to give up the 11C, and
they don't make those anymore

 

[Joerg]

We had those until we took the plunge and bought Android phones in
January. We make very good use of them and haven't been sorry at all.

We are on $5/mo pay-as you-go plans. Plenty for us. Virgin does have
smart phones but then AFAIK it pops up to $40/mo or so and I can't see
us using that. I hate it when people fidget around with their handheld
gizmos all the time. Some are literally obsessed. Middle of a nice
dinner ... *BING* ... "Oh, that could be so-and-so sending a text" ...
and out comes the schmart fone. Yuk.

Working as designed. ;-) I hated SR-50s. I would have stuck with my
slipstick if they were the only alternative.

It's been good to me. I could not possibly have calculated all those
filters with a slide rule. Plus back then HP's were truly unaffordable
for the masses, you had to be on a professor's pay grade for those.

 

[Martin Brown]

They were never very bright but nice small segments and very contrasty.

Me, not so much, I have to say that afetr I married I began to prefer
HP. Feels like better quality and I started to really like RPN. My wife
brought an HP-11C into our marriage and that's the one I use right now.
We could never divorce because then I'd have to give up the 11C, and
they don't make those anymore

The first calculator I ever encountered belonged to a school friend. It
came back from the US as his dad was a drugs rep. HP20 I think.

I RPN Obscure Very Find - Yoda.

I was just lazy and liked entering formulae in normal order without
having to think. I got away with mental arithmetic and writing SR for
"slide rule" in the margin until first year uni. Answers to 3 sig fig
and SR did not go down at all well with our crystallography supervisor
so I got a calculator. SR51 - I have fond memories of it too.

Regards,
Martin Brown