HP Forums

Hello everybody,

I just want to share my joy over the successful reviving of a HP 9810A.

This device was in my cellar for several years and didn't work since I got it from its former owner. Some time ago I did a first try to repair. At this time I got some useful information from this forum. Unfortunately I had no success then.

At the begin of this year I found some time again. I set the goal to present this calculator on the European Vintage Computer Festival (VCFE), which will start at the first of May.

It took about two months of hard work. I found three defects in the machine: one dead ttl logic gate (74H30) and one dead transistor in the address drivers of the DRAM circuits. The third defect was hard to find: The calculator worked, but the display was wrong. After comparing the memory accesses with these listed by the HP9800 series emulator, I located the fault in one of the microcode ROMs. At least two addresses give the wrong data. One of them caused a wrong sequence in the microcode. It should be mentioned, that this circuit is an Intel 3301A, which is a mask programmable ROM. I never assumed that these could get defective, they should live almost forever.

I "repaired" this by decoding these addresses and correcting the output of the ROM by some external circuit. The solution does not really look pretty:

but it works! After the VCFE I will do some cleanup.

I would like to say Big Thanks to Tony Duell who gave me the first hints and especially to Dr. Andreas Bürger, who wrote the great emulator for the HP9800 series.

I found the most information in the patent US-3,859635, which can be found here. It contains the schematics of the circuits, the dumps of the microcode roms, some listing of the application code and lots of explanations to data stuctures and sequences. The HP 9810 is obviously one of the best documented calculators.

There is an repair manual from HP, which can be found on the museums DVDs. This manual is next to useless now, because it requires a complete set of circuit boards to find the defective one by swapping one by one.

The next step will be the connection of the plotter HP 9862A which I have too. For this I will have to rebuild the missing connection module.

Best regards,

Thomas Falk

Thomas, congratulations for your success!
After having done a similar "revival" project on a very sick 9815A, I can understand the pride and joy which comes when bringing an HP calculator back to life. And when it finally works, you have to tell the world about it. It's part of the experience!

Joel Setton

Impressive work Thomas. Congratulations. Hope to see some pictures of the beast in working order.

Hi,

thanks for the gratulations.

Here are two pictures. The first shows the display after repairing the memory system. The numbers shown where displayed after pressing the Clear key and didn't change much after pressing other keys. Only the program display was right. On top you see the connections for the logic analyser at the microprocessor board and the memory block. The red/black wires are connections into the memory block to be able to measure on the working calculator.

The second picture shows the calculator after fixing the error in the ROM during the last checks before closing the cover. At the left is a logic analyser and the oscilloscope. Both are useful for different kinds of faults.

If you want to see the calculator live, you should come to Munich at the next weekend to the VCFE

Greetings,

Thomas