History: I have had my 25C since I bought it new back in 1977, the calculator was unused for almost eight years sitting in my drawer. With my recent interest in old HP, I decided to revive it. After replacing the old cells with new NiMH cells, I charged it up, and it worked - a kind of.
Problem description: The storage registers would not reliably store and recall numbers. Program memory was erratic, you could key in a program, after a while, you got a lot of GTO 00 in it. I accidently discovered if the calculator was warm (if it is on charger), then the problem would disappear and everything worked. However, when it cools to room temperature, the problem would return.
Calculator affected: HP 25C
Fix: Thank you David Smith and Luiz for their responses, and especially Katie, for her repair tip on
http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/articles.cgi?read=267
This afternoon, I went by the electronics store to pick up the 20K resistor and the zener diode to conduct the EC on my 25C.
After an hour of work, I conclude that Katie's article is absolutely spot on. Although her article did not specifically include the 25C, the procedure she listed for the 25 work beautifully.
I solder a 1/8W 20K resistor across pin 11 and 12 on the ACT chip. I did it on the back side so I don't have to solder over the ACT chip. The 1/8W resistor is preferred, as it is small enough that it barely protrudes above the other pins, and the 25C does not generate nearly enough voltage to overload it.
With that done, the problem is now gone. Both my storage registers and program memory is back in working order, and the Continuous Memory works. Before closing the calculator up, I made another EC by adding the zener diode to protect the fragile electronics.
Observation: The strangest thing is I have never had this problem until now. Everything worked perfectly for 20 years, I only discovered the problem after it came out of an eight year storage. Katie referred to excessive noise on the low frequency clock signal, I began to wonder if the aging of the electronics might have something to do with the noise.
Lesson learned: If you own a 25C, before assuming that register and program memory problems are symptoms of fried memory chips. Try Katie's fix before you give up, you may get a happy ending with a $0.20 part.