At 12:31 AM 5/28/02 +0100, you wrote:
A little while ago I seem to remember someone in the UK saying they
needed a power adapter for one of the original HP calculators -
HP35/45/55 etc.
I have one if that person would contact me off list....
It wasn't me, even though I have a fair collection of HP calculators
(including, I think, all but one of the devices to use that PSU).
However, this seems like a good place to post some information on
repairing said PSUs.
This unit is one of the few that charges the battery properly. The PSU
has 3 connections to the calculator. Ground, a constant-voltage (4.2V)
supply to run the logic and a constant current (55mA) supply to charge
the NiCd. When the charger is not connected, then the battery is
connected to the logic supply line. When you plug the charger in, it
moves a contact in the calculator plug away from the pins and disconnects
the battery and the calculator.
For models _without card readers_ (35, 45, 55, 70, 80) it's safe to
connect the charger without a battery pack in the calculator. There is no
connection between the battery and anything else when the charger is
plugged in.
WARNING: this is true for all models EXCEPT the LED C models (25C, 29C, etc). DO NOT
try to run one of them without a GOOD battery or you'll burn up the memory chips. FWIW
In these the ROM, RAM and constant memory are all on the same ICs so you may destroy more
than just the constant memory.
For models with the card reader (65, 67), you must have a
good battery pack installed in the calculator before
connecting the
charger. Otherwise the card reader sense amplifier chip sees the full
voltage of the constant current supply (about 16V) and may fail. This
chip is, of course, HP custom.
Yes, However the same IC is used in the 67, 97 and the HP 41 card reader. So there are
at least a supply of replacement ICs available if you're willing to sacrifice
something.
Anyway, getting back to the charger. There are basically 3 stages to the
circuitry. An unregulated DC PSU giving about 16V (transformer, 4 diodes
as a bridge rectifier, 400uF capactitor). A constant voltage regulator
(resistor + zenner and an emitter follower). And a constant current
regulator (a couple of transistors and resistors). All discrete
components, nothing custom.
As to what goes wrong with them :
1) Open-circuits in the output cable, particularly close to the
calculator connector. Trivial in theory, harder in practice. You have to
slice open the connector as it's moulded to the cable.
This is by far the most common problem and they're usually intermitant so that makes
them even harder to spot.
2) The 400uF smoothing capacitor goes open-circuit.
Easy to spot since the output voltage drops under load. The cap is short and
replacements are difficult but not impossible to find.
3) Some versions have a fuse between the transformer secondary and the
rectifier. It's a soldered-in picofuse. Sometimes it fails for no good
reason.
4) The power transistors (especially the NPN one in the constant-voltage
circuit) can go open-circuit.
Here's an ASCII-art schematic of the classic charger....
HP82011 (etc) 'classic series' charger
--------------------------------------
(Unregulated PSU)
-----)||
)||(----+----->|--+-----+------ +16V
)||( | | |
)||( | +-->|--+ === 400uF
Mains)||( | | ---
)||( +--)--|<--+ |
)||( | | |
)||(-------+--|<--+-----+---+
-----)|| |
---
///
(Constant voltage regulator)
c e
+16V ----+------\ >---------------------- +4.2V Logic Supply
| \ / (NPN power transistor)
| -------
/ | b
\ |
/ 470R |
\ |
| |
+---------+
|
---,
^ 4.7V zener
|
---
///
(Constant current charger)
(PNP power transistor)
13R e c
+16V -----+----\/\/\---+--------> /------------- 55mA constant current
| | b \ / charger supply
| ----- -----
| e / \ c | b
+----------> \--------+
(PNP signal transistor) |
/
\ 4k7
/
\
|
---
///
-tony
Joe