HP Forums

Full Version: Re: HP 25-C Repairs
You're currently viewing a stripped down version of our content. View the full version with proper formatting.

IMHO, in the case of the Continuous Memory versions of the Spice (HP3xC) and Woodstock (HP2xC), connected them to the charger without a battery pack in place is likely to
damage the (HP custom) RAM chip, particularly if the machine is turned off at the time. Connecting the charger to a machine without a battery pack that's turned on is somewhat safer.
The reason is that there's a backup supplg from the battery pack to the RAM chip, which is what maintains the memory. This voltage is limited by the battery itself. Without a battery in place, the voltage will rise to about 10V when the charger is connected (it's supposed to be about 2.5V -- the battery voltage), and this
will kill the RAM chip. The RAM chip on its own doesn't draw enough current to pull the charger voltage down (these chargers have an internal resistance of about 15 Ohms, BTW).
If you have the machine turned on, it is likely to draw enough current to reduce the voltage to something a little safer (although I would still be careful!).
It's best to make sure you have a good (not open-circuit) battery pack in place before you connect the charger.

I'd like to add to this that in my experience, the most likely chip that gets killed in the Continuous Memory Woodstocks is not the RAM chip; it is one of the ROM chips. Which makes things even worse... whereas the RAM chips are somewhat interchangeable (the same RAM chip is used in the 19C and the 29C, and also in some versions of the 25C; other versions can be easily modified to accept the double-capacity RAM chip) the ROM chips are specific to each model (there's some overlap between the 19C and the 29C).

The charging circuit in the Woodstocks is really despicable. Current limiting is accomplished by a simple resistor, that asymetrically heats the battery pack; there's no overvoltage protection; and the battery contacts are tin plated, subject to oxidization. This, IMO, makes premature death of the battery pack inevitable, and eventualy failure of the calculator due to overvoltage likely. Compared to the otherwise amazing design of HP calculators, I really fail to understand how this came to be!


PS: I always charge my Woodstock battery packs these days in a "for parts", non continuous memory machine :-)

I am suprised it's the ROM that fails -- it should get no power at all when the machine is turned off. The RAM on the other hand
gets a backup supply from the battery pack, and this rises when the macbine is off, with no battery, but with a charger connected. That could
certainly damage the RAM chip. My experience is more with the HP3xC series which use an electrically
identical battery pack and charger circuit. Certainly in the 33C, it's the combined ROM/RAM chip
that generally fails and I assumed it was due to overvoltage on the supply pin.
The HP2x and HP3xE/C charger circuits are very simple. Just the transformer secondary winding,
a diode, an 8.2 Ohm resistor and the battery pack in series. The transformer has
a fairly high internal resistance (10-15 ohms) which will limit the charging current somewhat.
But the charger is hardly kind to the cells.
It would be very easy to make up a separate charger for the HP2x and 3xC/E battery packs. Use the HP adapter, the diode and the resistor as above. The hardest part would be making something to connect to the terminals of the battery pack.
I can't do ascii-art schematics on this forum, but I can e-mail one to anyone who's interested. The circuit really is simple.

Well, if you're making an external charger, you might as well do it right and build a "proper" charger with regulated charging current: it's very easy to do, you can build a simple circuit using a common 7805 voltage regulator chip, for instance (all you need is that 3-pin chip, a suitable resistor, and a capacitor.) Or better yet, if you've already opened your Woodstock battery pack to replace dead cells, you might as well make it so that you can remove the cells anytime, and just charge them in an "AA" NiCd battery charger. That would also solve the problem of connecting to the battery pack.

I don't know why it's the ROM that fails, but I've seen something like half a dozen or more dead "C" Woodstocks, and it was always one or more ROM chip that has failed. One possibility (I am just guessing here of course) is that, since the ROM and RAM chips share part of the bus, when the RAM receives excess voltage, while it itself might survive, it puts out something excessive on the bus that kills the ROMs?

That said, I have also encountered fried RAM chips occasionally, but in those cases, the calculator generally remained functional (or it was possible to bring it back to life by removing the RAM), it's just that its registers or program memory wasn't working properly.