RIP 46

[Keith Midson]

A sad day - my "built like a tank" HP46 suddenly decided to pour out grey smoke out the back and make a horrible bang noise before switching itself off forevermore ...

So I guess I will need to get a new calculator now. It will be hard to find something that has the same presence in its case!

Cheers, Keith

[bill platt]

Oh NOOOO!

And that is the first time I even remember anyone mentioning the 46! Had to look it up!

[Katie Wasserman]

A common problem on the HP-46 is the large electrolytic capacitors in the power supply failing. A loud bang with smoke could very well be an indication of this. Such a failure would likely have also blown the fuse. Open it up for inspection, a blown up capacitor will be obvious. They are easy to replace if you're at all handy with a soldering iron.

[Keith Midson]

Yes the fuse self destructed quite spectacularly as well. My 7 year old son was devastated! Good to know about the capacitor - I assumed that it had died for good. I will have a look and report what I find.
Cheers, Keith

[Tony Duell]

You can get schematics from the Australian Museum site.

I would start by taking off the top and bottom covers and then look for obviously burnt or damaged componets. Yes, some faults do leave a visible trail!. But remember that the burnt part may have been damaged by a fault elsewhere.

Continue by unplugging the transformer secondary edge connector from the back of the logic PCB. Replace the fuse with one of the correct rating and apply mains. If the fuse blows, then you have shorted turns in the mains transformer :-(. If the fuse holds, check the secondary winding voltages.

Then check the PSU components on the logic PCB. Others have mentioned the smoothing capacitors (the value is not too critical, if you can't find a 5000uF, don't worry, a 4700uF is fine). Also check the rectifier diodes (I've had problems with these in other machines) and the pass transistors.

[bill platt]

Hi Katie,

Remember how Valentin felt strongly that in some period of time, all these wonderful old devices would die? Was he speaking of the simple dead bug pieces like electrolytic caps or was he thinking about something else?

As our machines age, I wonder if it is possible to come up with a "safe pre-test" protocol for checking the machine's capabilities before actually powering up?

[Mark Storkamp]

This is all making me wonder about the possible modes of failure. I know switches are prone to metal fatigue, rubber can deteriorate, capacitors can dry up, and tin whiskers can sprout in chips.

What are the other possibilities? Can P and N regions migrate? Does the O in MOS break down? How long do LCDs last? How many years until 37% of the units have died of natural causes? I suspect the newer models have a shorter life span for a number of reasons.

[John B. Smitherman]

The newer models have fewer parts which should help with their overall longevity.

John

[Mark Storkamp]

Quote:

---

The newer models have fewer parts which should help with their overall longevity.

---

But the dies are smaller and more tightly packed, so slight manufacturing tolerances are going to have a greater impact. They typically run at a lower voltage so less variation over time will put them out of spec. Of course, this is all conjecture.

On the other hand, they also (in my case) tend to just sit collecting dust while the older ones get used far more often.

[Katie Wasserman]

HP's experiment with NMOS in the HP-27 proved that early NMOS devices fail sooner than PMOS devices of the same era. That's why there are so few working HP-27's around and yet most of the other Woodstocks still work.

Quote:

---

and tin whiskers can sprout in chips

---

... and now they can sprout on every soldered connection since RoHS went into effect. Actually I think the jury is still out on this, but there is great concern from what I've read about this possibility.

Quote:

---

"safe pre-test" protocol

---

As Tony suggests, you could discounted the power supply from the logic and check that first. This would at least limit the damage to just the power supply, which is usually easy to repair. This works on large desktop machines where the power supply is almost always easily separable from the logic. On pocket calculators of
unknown condition I power them from a external power supply with current limiting. This allows me to ramp up the current slowly to limit the chances of burning something out. This is a bit tricky to do since the calculator won't usually turn on until you give it enough current, but with practice you learn what's too much and want's about right.

-Katie