HP Forums

Full Version: My HP 55 died
You're currently viewing a stripped down version of our content. View the full version with proper formatting.

My old HP 55 finally died. Sometimes I get a 3 in the display when I turn it on, sometimes I don't. I've taken it apart, cleaned the contacts and looked it over. But I can't find anything obviously wrong. Anybody got any ideas on fixing it?

Hi, Do you get 3.00 or 0.30 or something like that or just 3?
Please give us some more info. concerning the symptoms.

It's just a 3 when I turn it on.
Sometimes it appears in the 4th LED position , sometimes at the far right. Most of the time the display is just dark. This is true for RUN, PRGM and TIMER modes with the battery in and the external power source plugged in.
I did take it apart and clean the contacts with a Q Tip soaked in Archer cleaner/degreaser.

When that 3 appears, is it very bright?


Viktor

Yes. If it appears at all it's at full strength.

My guess is ..kiss it goodbye?? I waiting with baited breath for your answer, since I have a non-functioning 21 that shows a single "very bright" digit (plus some randomly/partially lit segments on the other digits).

I'm guessing my 21 was, at some point before I got it, powered from the AC adapter without a good battery-pack in place, which most likely fried the LED driver??

OTOH, unlike my 21, Walcott's 55 shouldn't necessarily have suffered the same fate as mine, as the "Classic" series are "able" to operate w/o a battery in place because the adapter is well-regulated. Right? Wrong?

Thanks,
Matt

My guess is that it is one of the LED driver chips.

Where exactly is the LED driver? On the HP-21 and classic calculators?
Thanks.

So ... anyone know how to replace a blown LED driver for this thing, or who can do it?

Matt,

On your HP21 (or the HP55 for that matter), have you tried replacing the batteries? My experience with my own HP21 suggests that I get a frazzled display when I run AC and a battery pack with very old NiCds. On the other hand, the same calculator display looks perfectly normal when I use a battery pack with fresher NiCds and AC power.

Someone posted an informative message earlier that in the Woodstocks at least, the battery acts as a "buffer" or "filter" for the AC line voltage. Therefore, it makes sense that if the batteries are not in good shape, the display output may suffer.

Just my $0.02

Todd Garabedian
Glastonbury, CT

Most likely the problem is NOT an LED driver chip. Bad drivers typically would show up as a single missing segment across each digit in the display, a single missing digit, or a group of missing digits.

If the single digit you see if VERY bright, the circuitry that scans the display is probably bad. Normally each digit is on about 8 percent of the time. If the scanning is locked up one one digit, it will appear about 8 times as bright as normal (and will probably burn out in short order).

If the single digit appears normal intensity then the circuitry that feeds the display scanner is broken (probably the CPU). The fact that you didn't mention that the digit is glowing like a small sun and that the bad digit moves around on power-up makes me think this is it).

I'd check and clean the contacts again (particularly the gold forked posts between the keyboard and the CPU board). Also look for a stuck key. Try another battery. Swap the CPU card with another machine (you do have a spare, don't you?)

You can't miss them on the classic calculators. They are right below the display. There is an anode side driver and a cathode side driver.

Todd,

Thanks for the advice, I'll take a closer look. I was using a battery pack from an HP 25. With both calculators side by side, and swapping the battery back and forth between calculators, the 21 had the symptoms I described, whereas the 25 operated normally. But I do know the battery wasn't fully charged. I think I'll try repeating the test after recharging the battery overnight. It's certainly worth a shot!

Regards,
Matt

It is of normal brightness.I cleaned the forked posts between the keyboard and the CPU board and I no I don't have a spare :)

I'm beginning to think it's time to chuck it out. All things come to an end.

Walcott,

A single, unusually bright digit and a dead keyboard together are clear indicators that the display and keyboard are not scanned. Possible reasons include a dead main oscillator, power supply problems, dead CPU, even possibly (but not probably) a dead cathode driver. This is usually when I push the ON button on my oscilloscope...

If the digit appears with normal brightness, the problem is probably something more subtle, since obviously, the display is then still scanned properly, so the clock frequency is present. In other words, the problem is likely a faulty CPU or ROM chip, neither of which is good news.

You were already advised to check for bad/dirty contacts; another routine thing I do is to check for faulty semiconductor junctions using the diode setting on a DMM. (This measures the 'forward bias': basically, it lets a small current flow through the junction in the 'open' direction, and measures the voltage drop, typically around 0.7V for a silicone junction.) Most diodes and transistor junctions can be checked in-circuit this way, and if there's, say, an open diode (measures as infinity) or a shorted pn junction (measures as 0 or something very close to 0) in a transistor, they can be detected. If you get lucky, you might find that the problem is only an easy-to-replace diode that has gone bad.


Viktor

Hi Walcott!

I only now read your post and unfortunately don't have any good suggestions on how to get your calculator working, but I DO suggest you don't "chuck it!"

Offer it for bids on ebay! Even if you don't care to get any money for it, you might!, and even more importantly, you'll help others who might be waiting for a broken machine to salvage parts from to repair one that they have that doesn't work! I even have some interest, as I have a 55 that works, but has a somewhat stuck key. By putting it on ebay, you'll reach MANY HP users!

Best wishes!
-Mike

Walcott sent me the calc for evaluation. It has a bad chip (HP# 1820-1128). This chip was also found on HP-45s. Does anybody out there have a spare?
Thanks

The 1820-1128 chip is a combined clock buffer and reset circuit.
It contains the clock buffer circuitry that's in the 8 pin chip on an HP35 and the reset circuitry that's built from
2 discrete transistors, etc in that machine. The pinout of the 1820-1128 seems to be :
1-Reset Out ; 3-Clock Phi2 In ; 4-Gnd ; 6-Vgg; 8-Clock Phi1 In;
9-Clock Phi1 Out ; 11-Vss ; 13-Gnd ; 14-Clock Phi2 Out ; 16-Reset Capacitor
What is the problem with your chip? If it's one of the clock outputs, you might
be able to kludge in the 8 pin buffer from an HP35. If it's the reset circuit you
might be able to make it from discrete components.

The display when firt turned on is usually blank. Sometimes a single number (most often a 3) will appear in two different positions, and sometimes when you switch it on (very rarely) just right it will turn on properly. My guess is it is the clock circuitry.I just found a HP-45 CPU board in my parts bin and will try switching out the chip tonight.
By the way Tony, I have a faulty HP-15C. It appears that one of the three chips are faulty on the board since I checked the two caps and resistor. Are these chips specific to the 15C or does the HP-12C have similar chips?

How do you know it's the clock buffer/reset chip? It could be a lot of other things.
The fault actually sounds more like a reset problem than a clock problem (if one of the clock signals is missing, then you
get no display at all). But if you have a 'scope you might look at the waveforms on the Phin output pins. If either is missing, check the corresponding input pin -- if that's
also wrong, you've got an anode driver or connection problem.
But as I said, this sounds like a reset problem. Before changing the chip, check the 2.2uF capacitor connected between
pin 16 of this chip and ground. It might be low in value or something.
Wit regards to the 15C problem, start by resoldering all the connections to these chips. I've had a few voyagers that
suffer from dry joints. You might be luckky. I beleive the 3 chips are :
R2D2 (ROM, RAM, Display Driver) -- the largest chip, then a ROM, then the Nut CPU (on the far side to the
battery holder). The Nut is the same as the one used in other Voyagers (but not all versions -- later 12Cs have everything on one chip), the other 2 chips are unique to the 15C.

Replacing the HP part # 1820-1128 chip did the job. The calculator is working normally again. Thanks for the input on the chip's function.
Without seeing a schematic, I had theorized that perhaps the chip was providing some sort of sync pulse.

Sounds like it was the reset section that had failed. IIRC, these
'Classic' series machines use the reset pulse to synchronise various
sections (it's connected to the ROMs, for example). A missing reset signal may cause the 'CPU' to
start at the wrong location, I think. Keep the old chip, though. If the clock
buffer section is good, it could be kludged into an HP35 if necessary. Or you could
rebuild the reset circuit in discrete components like the HP35 does.

My bet is the chip is probably good and the reset cap was the source of the problem. The reset section inside this chip so simple it is VERY unlikely to fail. I think it consists of two transistors.

Also, a friend of mine fixed a similar problem on one of my machines by soldering a 6.8 uF cap across the power pins of this chip. So definitely keep the old one handy.

I think you're right about the reset circuit in this chip. The HP35 uses a few
discrete components (2 transistors, a handful of resistors, and a 2,2uF capacitor, one end of which is grounded). it seems to be a
simple schmitt (?spell) trigger circuit. On the HP45, etc, this
circuit is hidden inside this chip, with the capacitor (the only component that couldn't
be put on the chip) connected externally. If it's not quite the same circuit, it's going to be fairly similar. Incidentally, the CPU hybrid in the HP65 also has a 2.2uF
capacitor connected from one of its pins to ground -- yes, it's the reset capacitor. Probably much the same circuit on one of the chips
inside the hybrid...

The chip was definitely bad. I switched it out with another from a HP-45 and it worked.