Erroneous HP-16C

[damaltor]

Hi everybody,

I have in front of me an HP-16c, which behaves oddly.

The good:
First of all, both self tests run without errors. But while using the device, the following happens:
- I calculated 15 minutes of random integers. No problems. HEX/DEC/BIN/OCT work like a charm.
- I noticed that converting to DEC takes a while (>3 seconds), while BIN/DEC/OCT are shown almost instantly. Is that normal?
- Floating point calculation works. I calculated 10 minutes of random numbers and tried most functions.

The bad:
- I cannot change the word size in integer mode. If I type "8 WSIZE" the calculator display blanks and stays blank. If I pull out a battery, and wait a while, and put it back in, random things happen - blinking display, or wrong calculations (3+3=2.5) or such. I can bring it back to function with a memory clear (ON+MINUS).
- If I go into PRGM mode and clear the programs, the calculator correctly shows "000-". If I start to type in programs, it always starts at line 85, and if I continue to program, the line number wraps around after line 91:

Code:

Key     Display
P/R     000-
1       085-    1
2       086-    2
3       087-    3
4       088-    4
5       089-    5
6       090-    6
7       091-    7
8       085-    8

The interesting thing is, the program works anyway: if I now press R/S, the display shows "12345 d" which is correct as the word size is 16.

Also, some key codes are shown wrong:

Code:

Key     Display
x<>I    085-000 72 (Should be: 001- 42 22)
x<>y    086-000 34 (Should be: 001-    34)
CLX     087- 43 35 (display is correct)

Most keys including the gold f-shift are wrong as far as I can tell (though the program seems to work fine as far as I tested, but I guess that would not work for longer programs).

Any ideas? My guess: The RAM is broken or desoldered, as the self test runs fine I guess that the ROM should be ok (though I dont know what exactly is tested in the self test).

[Katie Wasserman]

Quote:If I go into PRGM mode and clear the programs, the calculator correctly shows "000-". If I start to type in programs, it always starts at line 85, and if I continue to program, the line number wraps around after line 91:

If this happens after a full reset and with good batteries installed, I'm pretty sure you do have some bad memory in there.

[damaltor]

Well well. I have now opened the device. It is the version without the plastic film. I noticed that the "south" end of the pcb is covered with clear tape. All solder contacts look fine, I will resolder them though. maybe it helps...

[damaltor]

It did not help at all. I found stuff out though:

I did now notice that if I cannot switch it back on, I can reset it with the magic reset D+ON, which I found in a paper for synthetic programming on the hp 15c (on which the combo is y^x + ON which are the same keys).

https://www.brouhaha.com/~eric/hpcalc/hp...ant83.html

even more interesting, the calculator then returns and works fine, it even changed the wordsize correctly. So to change the wordsize I have to type 8 WSIZE (calculator switches off) D+ON (calculator switches back on) CLX (because there are weird values in x, see above link).

So, the device obviously is broken, but kind of usable. Does anyone have source for another broken device or the ICs?

edit: i have here the combination of 1LE2-0322 and 1LF5-0301.

[Dave Frederickson]

(01-13-2016 07:54 AM)damaltor Wrote:  Does anyone have source for another broken device or the ICs?

Check The Sphere Research Canadian Test Equipment Site for parts, eBay for a donor calc.

[damaltor]

Thanks, that sounds good!

Update: I have now transplanted the CPU from a (working) 11C into the 16C and vice versa. And voila:

The 16C now works perfectly!

The 11C instead does now the exact same thing which the 16C did before: his program memory starts at 85 and the keycodes are wrong etc. Still, the 11C passes both self tests.

I guess the CPU is faulty. As the 11C is easier/cheaper to replace than the 16C i will happily leave it like that. Still i would like to fix the 11C, but the CPU is almost definitely faulty. I need a 1LF5-0301!

[Katie Wasserman]

(01-13-2016 04:01 PM)damaltor Wrote:  I guess the CPU is faulty. As the 11C is easier/cheaper to replace than the 16C i will happily leave it like that. Still i would like to fix the 11C, but the CPU is almost definitely faulty. I need a 1LF5-0301!

Good to hear that you got the 16c working. 12C's often go for very little on ebay, look for one of those for a donor 1LF5-0301.

[damaltor]

Thats a good idea, thank you!

[jebem]

Most probably there is no need to point it out once again in these forum pages, but it doesn't hurt either.
Excellent Eric's site contains a lot of information on these machines as well.

The 12C series saw a number of CPU variants installed, so I guess that the 12C to look for as a donor should be as old as possible, to increase the chances of getting that 1LF5-0301 chip.

[damaltor]

Yeah, i have seen that. As the 12C machines are "the cheapest" though, i might trial and error until i have one that fits - though i dislike sacrificing another old machine, so i guess i will wait until someone sells a broken or beaten up one. Those have their history, too, but still for me personally the 11C is of more value than the 12C. also, the 12C is as far as i know the most popular and most produced, least rare device, so it might be okay to sacrifice one.

[Paul Berger (Canada)]

(01-14-2016 12:44 PM)damaltor Wrote:  Yeah, i have seen that. As the 12C machines are "the cheapest" though, i might trial and error until i have one that fits - though i dislike sacrificing another old machine, so i guess i will wait until someone sells a broken or beaten up one. Those have their history, too, but still for me personally the 11C is of more value than the 12C. also, the 12C is as far as i know the most popular and most produced, least rare device, so it might be okay to sacrifice one.

I have a 12C here that has the CPU chip you desire in it, and a bad display. It was the display donor for a 16C I repaired for Robert Prosperi, I went back and looked at some earlier email and he said the 12C was fully functional. I put batteries into in just now and it does appear to work although it is impossible to confirm completely due to the condition of the display. When I pressed f-P/R it went into program mode and display step 000, I entered a short program and it appeared to run ok. It is your for postage, PM me with shipping information if you are interested.

Paul.

[damaltor]

PM sent - thanks in advance!

[rprosperi]

(01-14-2016 02:41 PM)Paul Berger (Canada) Wrote:  I have a 12C here that has the CPU chip you desire in it, and a bad display. It was the display donor for a 16C I repaired for Robert Prosperi, I went back and looked at some earlier email and he said the 12C was fully functional....

Indeed, that 12C worked perfectly, it was just a sacrificial donor for the 16C LCD repair. Also, I can confirm it was an early machine so CPU is almost certainly correct. Good luck with it, and thanks Paul both for remembering and sharing.