The Museum of HP Calculators

HP Forum Archive 17

[ Return to Index | Top of Index ]

HP 9815A repair
Message #1 Posted by frankabc on 22 Aug 2007, 11:38 p.m.

We need assistance in diagnosing our HP 9815A calculator computer. If anyone can help us fix ours or locate a replacement HP 9815A, we will be VERY appreciative. Thank you. Frank Simpson, fbsimpson@pharmacy.wisc.edu , tel. 1-608-849-3382.

      
Re: HP 9815A repair
Message #2 Posted by Tony Duell on 23 Aug 2007, 4:33 a.m.,
in response to message #1 by frankabc

You can get 'my' schematics of the HP9815 from the Australian site. Somewhat strangly for an HP calculator, the 9815 is based on a standard microprocessor, a 6800. There are a few 'suprises' in the electronics, but nothing too strange.

You don't give us much to go on. What does the machine do? Do you get anything on the display? What experience of electronics troubleshooting do you have? What test gear do you have access to?

            
Re: HP 9815A repair
Message #3 Posted by frankabc on 23 Aug 2007, 3:55 p.m.,
in response to message #2 by Tony Duell

Thank you, Tom, Our misbehavin' HP 9815A does a few interesting things. Run Mode: First, when you key in 2, 2.00 enters automatically. When you try to key in 25, you cannot, because it will display 2.00, and then 5.00. Though the enter key DOES work, and can be used, it is not necessary to hit enter to do arithmetic. So you can now key 2, 5, +, and the result, 7, appears on the display. Normal RPN would be 2, ENTER, 5, +, display 7. Auto Start mode: When you flip the power-on switch, the 9815A will attempt to load a program from a DC100A tape cartridge. This seems to be working. Unfortunately our tapes are 20 years old, and we fear that they might not have readable programs any longer. The priter prints "file not found" or, on more encouraging attempts, "Checksum error". But we cannot tell if the tapes are good, or not, because we do not know it the computer is good, or not. The cartridge drive itself seems to be working. Program mode: One cannot simply enter the program mode with the switch. If I go through the following mindless steps, I can enter the program mode, every time.

Switch in run mode turn power on 1 Store 1 . rcl 1 . then switch to program mode.

Once in the program mode, the display reads 0000 on the left, and 0136 on the right. There should be 432 programmable steps, minimum, not 136. Furthermore, our machine has the 001 memory and 002 I/O options. The 001 option should increase memory to 2008 steps. Again, not 136. On good days, I can program a few steps. For example, I can key in steps to recall numbers stored in A -> J registers, multiply, end, etc. But when I switch back to run, she does not run. The printer prints "memory overflow", on almost every interraction. I have attempted to clear the memory by the protocol in the manual, and sometimes I can clear it. I have pressed the list button, and the printer prints the memory steps and contents, usually to 136 steps, but on one occasion, out to 216 steps. On that 216 day I was able to reallocate the 80 store registers to programmable registers by the protocol in the manual. Note that 80 + 136 = 216. I only have gotten 216 programmable steps once, and have never again been able to repeat it. I also have wondered why the priter prints "memory overflow" even after only turning power on in run mode. Isn't the memory supposed to clear when power is turned off, and then back on? Well, I decided to press list, immediately after turning the power on. The print out informed me that memory really was already full of all sorts of stuff. That is about as far as we have gotten. In answer to your questions. I have an electrical engineering friend who is going through the machine. Don used to repair electronics for our Biochem Department. We have a digital volt meter, and that is about it. Don has a logic probe at home. We have all of the manuals, including the service manual. We have found the Utilities and Test Cartridge Tape (09815-10004), but we are not sure if it is still readable. It yielded "no file found" when inserted into the tape drive. It would be nice to obtain a known-to-be-good copy of the Utilities and Test Tape from someone. We also cannot locate the 9815A Test ROM Assembly plug-in (98140-66501). It would be very beneficial for us to find these two troubleshooting aids. We are wondering if Digital Signature Analysis is possible with the 9815A. It looks like HP was developing Signature analysis around 1980. We also are wondering where your web site is with the schematics of the HP 9815A. We greatly appreciate any assistance and suggestions that you may offer. Thank you, Frank Simpson fbsimpson@pharmacy.wisc.edu 608-849-3382 (messages)

                  
Re: HP 9815A repair
Message #4 Posted by marais on 23 Aug 2007, 6:15 p.m.,
in response to message #3 by frankabc

Does the machine exhibit the same behaviour if you pull the 001 RAM option?

Andreas

                        
Re: HP 9815A repair
Message #5 Posted by frankabc on 23 Aug 2007, 6:52 p.m.,
in response to message #4 by marais

Many Thanks Marais for responding. This forum is great! And very Sorry Tony, for my mistakingly reading your name as Tom, Yes, Marais, we wondered the same thing a few days ago. The 9815A is showing the same, identical behavior since we pulled the 001 memory option. We still have the same 136 steps of memory displayed when in program mode, whether the 001 is present, or removed. Frank

                              
Re: HP 9815A repair
Message #6 Posted by frankabc on 23 Aug 2007, 7:31 p.m.,
in response to message #5 by frankabc

Dear Mr. Duell, We have located your schematic drawings for the HP 9815 via google by typing, hp Australia museum, followed by searching the museum site for, Duell. Very nice! Thank you! Frank

                                    
Re: HP 9815A repair
Message #7 Posted by Tony Duell on 24 Aug 2007, 1:22 p.m.,
in response to message #6 by frankabc

Quote:
Dear Mr. Duell,


If you want to be formal, it's 'Dr Duell', but 'Tony' is fine :-)

Quote:
We have located your schematic drawings for the HP 9815 via google by typing, hp Australia museum, followed by searching the museum site for, Duell. Very nice! Thank you! Frank

That '136' steps left is clearly wrong. I susect a RAM problem. Now, had it been a 9815S, I'd have immediately suspected the RAMs on the CPU board, as they're 2114s in that machine, and a more unreliable chip I've yet to see. But in the 9815A, they're 2111s I think (256 locations of 4 bits each).

Given that it's a 6800 system, it's likely that system variables are stored in page 0, since there's a shorter,faster way to access locations there. So I'd replace those RAMs first.

                                          
Re: HP 9815A repair
Message #8 Posted by frankabc on 25 Aug 2007, 2:26 a.m.,
in response to message #7 by Tony Duell

Dear Dr. Tony, Don understands what you have said. I don't. My degree is in biochemistry, not EE. Thank you VERY much for your considerate thoughts, and for your amazing schematic diagrams. We will let you know how it turns out. Dr. Frank


[ Return to Index | Top of Index ]

Go back to the main exhibit hall