Dropped HP41CX (carpeted floor), no longer working
#1

I dropped my HP41CX, and now it no longer works. Thankfully it was on a thick carpeted floor, and only from a height of 18" (I was sitting on the couch), but it will not turn on now. I took the back off, and all I could find wrong is 2 of the 4 "nibs" that hold the battery contact section in place were cracked, but still in place (but no longer straight-up). I carefully checked all contacts on the PWB's and the battery contact section and can find nothing else wrong. Any suggestions?

#2

If you firmly hold the two halves together (around the display), does the unit return to normal?

#3

As a long time TI guy I just have to offer the following quote from the Service Manual for the TI-58/59 which tells how to check for a faulty resonator:

Quote:
... perform "drop test" by: starting the diagnostic program, dropping the calculator approximately 6 inches onto a hard surface while the program is still running. If calculator goes into preload or gets wrong answer to diagnostic program, replace resonator.

Back in the early 80's I wouldn't do that with MY TI-59. But a friend who had easier access to a replacement since he was an instructor in the TI Productivity Program did it and his calculator passed. He went ahead and did the test from 12 inches and 18 inches and the calculator still passed. He didn't try a higher drop. He was dropping the calculator on his desk top. Maybe that wasn't a hard enough surface to really stress a TI-59.
#4

O.K.... I've collected about a dozen TI-59's, some from parts from several calculators. I've rebuilt keyboards, card readers, etc.

What's a "resonator"? (Or am I going to feel stupid for asking?...)

#5

Hi, Doc;

although general, this may help.

Cheers.

Luiz (Brazil)

#6

It's a little circular two-lead device that controls the clocking of the digital circuitry of the TI-58/59 series. It's a type of primitive quartz crystal, though I'm not sure quartz was actually used.

#7

Michael, how did you rebuild the TI keyboards? Is there a way to "debounce" them?

#8

Ceramic resonators are much more accurate than an RC-governed oscillator, but not nearly as accurate as a crystal. The cost is much lower than that of a crystal too. See for example http://dkc3.digikey.com/PDF/US2010/P1490.pdf . I can't think of any reason the TI-59's clock source would need to be more accurate than an RC would make it though. Does anyone know what the approximate clock frequency was?

#9

Resonator is often used as a synonym for oscillator. As I see it the TI58/59 has two: the clock generator and the power supply. I don't know if the 59's card reader circuit has a resonant circuit. I would guess the "drop test" as described by Palmer refers to the clock generator.

#10

From a document by Spike de Wal:


Quote:
Phi_1 & Phi_2 Clocksignals:

This is a 2-phase (non-overlapping) clock that switches between Vss and Vgg. It is derived from a 455KHZ MF filter through the TMC0301-chip but some models of the TI59 have deviating ratings like 462KHZ. The Clock operates in two modes:

• Calculating (fullspeed)

• Display or IDLE-mode (lowspeed)


#11

Gosh... it's been awhile. I was never able to strip one apart... if I recall correctly, they're held together with adhesive tape. I have transplanted working keyboards to non-working logic boards. I've removed the keys from the calculator to wash them and the cases and returned them one-by-one to their original locations. (Usually) Then I've replaced the deteriorated foam with a sheet of very thin foam of the type that people wrap items in for shipping. Generally, they have worked well.

Fortunately, these calcs are still pretty cheap. I bought a working one with manuals, a printer, and lots of modules just a few weeks ago for $5 from that auction site....

#12

But why would any of those components be more succeptible to impact? I would think that, other than literally breaking a component, the only thing that would shut the calculator down would be a mechanical interruption of the circuit such as bouncing the contacts off the battery, interrupting the switch, etc.

I guess I'm not understanding why the clock circuit would be any more likely to fail than knocking a pin from an IC or breaking a diode or something...

Or am I just taking all of this too literally?

#13

Hi, Doc;

most (if not all current) components are actually sealed, meaning no 'empty spaces', like the ones found in vacuum tubes, are left inside of them. In fact, solid state 'everything' is a tendency, as we saw mechanics being continuously replaced by electronics (see CD´s, MPwhatever, etc.). Even solid state relays are replacing the mechanical ones.

The picture below shows the HP21 oscillating circuit (I am not sure if this is a resonator; does not look like a crystal, though)

As you can see, the little thing is composed by a conductive piece inside of a glass tube and 'touched' by the opposed contact terminals. The inside of the glass tube has vacuum (or is filled with some inert gas), and it is somehow easy to break or get loose after some time. Soldering by itself may damage the component if not controlled accordingly.

Hope this helps (and is accurate...)

Lui z(Brazil)


Edited: 5 Jan 2010, 12:42 p.m.

#14

Although the ceramic resonator relies on mechanical piezo-electric properties [ref], I would assume that the resonator would be relatively immune to the shock from the described drop test. Perhaps the TI part had a known failure mode that made it susceptible to such shocks?

I agree that contact bounce from the battery terminals could also be a likely cause of errors in such drop tests.

#15

I get it now... thanks.

Palmer's post didn't actually say that dropping the calculator would break the resonator. It said that the drop test could detect a faulty one:

"Quote:
... perform "drop test" by: starting the diagnostic program, dropping the calculator approximately 6 inches onto a hard surface while the program is still running. If calculator goes into preload or gets wrong answer to diagnostic program, replace resonator."

THAT makes sense to me, now that I understand the principal of the ceramic resonator. If it were mechanically faulty, which I can see happening from the "right" kind of impact, this test would DETECT it.

O.K... and I was just about to sacrifce a TI-59 to see how far I'd have to drop it and what would break first.....

#16

Please, read 'solid state' instead of 'sealed' in the first sentence:

Quote:
most (if not all current) components are actually sealed(...)
Thanks!

Luiz (Brazil)

#17

Hi, doc;

if the resonator has a bad contact inside of it, chances are that operating the calculator after a while might heat it up enough to disconnect the resonating part of it and cause a malfunction. By dropping the calculator, it would misplace it for good, I guess.

Cheers.

Luiz (Brazil)

#18

Looked at my TI-58 last night (currently open for repair), it has "384" marked on it, implying 384kHz.

Edited: 6 Jan 2010, 6:06 a.m.

#19

Hi,

I looked at my TI-58 last night (it's open for repair), and the resonator is a platic cube similar to this:




Except it's blue and marked "384" which I'm guessing implies 384kHz.

Bart.

Edited: 6 Jan 2010, 6:06 a.m.

#20

Hi, Bart;

the one I found in my TI59 is different, but I have a TI59 carcass that has a resonator that looks like this one pretty much. Please, see the pictures below.

This is the resonator present in my working TI59

This is the resonator present in a TI59 carcass

This is just to compare the two calcs (click to enlarge)

As I could not find a similar component, either the TI58C has another oscillating circuit or the resonator has a different body. I did not even find a component with a color scheme of the kind orange-gray-yellow (384)

Cheers.

Luiz (Brazil)

Edited: 6 Jan 2010, 10:11 a.m.

#21

Hi Luiz,

Interesting to see different versions of a machine. Here's a picture of my 58:


The blue "box" next to the 8-pin DIP is the resonator.

And yes, the blue stuff you see on the tracks is from corrosion due to battery leakage - the reason I have it apart. Whether I'll be successful at repairing it only time will tell - it's one of the "as and when I have the time for it" projects. So I follow and learn much from these threads.

Thanks,
Bart

#22

Hi, Bart;

The one you see in the picture is a TI58C (is yours a 'C', too?), and it also had some blue powder (seems to be copper oxide, is that so?). Now, please, have a look at the picture pointed by this link.

TI58C and TI59 corrosion compare(Big pic: about 1.9 Mbytes)

The TI59 board (right side) has a lot of oxidation, and the TI58C board (left) seems cleaner, right? They had both about the same oxidation level before I left the TI58C resting for about 20 hours in a solution of 30 to 40% alcohol vinegar with hot water (could touch it without feeling pain...). You can see some of the copper oxide in the TI58C´s PC100A contacts, and also notice the missing copper (parts totally destroyed).

Battery leakage is mainly basic (PH), and the alcohol vinegar is acid, so they react with each other and the remains of the reaction (a salt, right?) is easily removed with a paint brush, current water and neutral detergent. As you can see, I removed the LED display, power supply and some components (transistors, elect. caps, "SSS" module contacts, etc.) prior to clean it up.

I have been doing this for some time (woodstocks and these TI's, mostly) and have successfully removed copper oxide from them. In one particular case, a TI59 board with extensive oxidation damage, had the remains of many copper tracks being dissolved or popped out of the mainboard.

The hot water accelerates the reaction, but after some time its temperature raises down and the reaction decelerates as well. Anyway, I always keep my eyes over the board during the process and never let it stays for more than 20-22 hours.

Hope this helps.

Cheers.

Luiz (Brazil)

#23

Thanks Luiz for the good advice! Mine is not a 'C'. The casing and keyboard assembly are in very good condition, so I hope I can get it to work again.

Bart

#24

... but I never used it (actually, I´ve never even saw a bottle of it...). Chances are that it is locally applicable.

About the TI58 and TI58C: as you surely know, the constant memory implies another power supply design and low-power RAM chips, hence a different PC board layout at least to accommodate the 'new guys'. I read (at Joerg or Viktor site, cannot be sure about it) that current through these RAM in low-power state ('resting') is almost in the same range as the one observed in internal, natural battery discharge. I also see that the TI58 has the ON/OFF switch connected to the same position as the ON/OFF switch of the TI59; the TI58C has it connected to the opposed side of the PCB.

Cheers.

Luiz (Brazil)


Edited: 7 Jan 2010, 6:37 a.m.

#25

Quote:
Perhaps the TI part had a known failure mode that made it susceptible to such shocks?

I agree that contact bounce from the battery terminals could also be a likely cause of errors in such drop tests.

I think that you are correct about a "known failure mode". The quoe that startd all of this was part of a larger section devoted to repair and modification of existing boards. A more complete quote is:
If the calculator has a -3 thru -6 board, the following modifications must be performed:
* Add a 10 Meg resistor across pins 3 and 5 of the clock chip.
* Check CR7- must be two PG 1992 diodes in series
* Check resonator: Replace if yellow, uncoatd resonator is found. If rectangular resonator is used, perform "drop test" by: starting diagnostic program, dropping calculator approximately 6 inches onto a hard surface while program is still running. If calculator goes into pre-load or gets wrong answer to diagnostic program, replace resonator.
* Check Pot - must be 1.0K, 1.5K or 2.5K with a 1.21K resistor in parallel.
* Check C7 - add 22 uf 15v capacitor if missing.
* Check for proper capacitor(s) for each clock chip. for
240 - 75pf between pins 4&5 of clock chip.
- 75pf between pins 3&5 of clock chip.
300 - 75pf between pins 3&5 of clock chip.
301 - 33pf between pins 3&5 of clock chip.
* And on and on with change affecting card reading and display.
When I first saw all of this back in the early 1980's my first thought was that those were the kind of instructions which can be generated when working with a large production run

On resonator frequencies from page 4 of the Service Manual

Quote:
The ceramic resonator Z1 resonates at a frequency of 455 Khz +/- 1%, which establishes a stable frequency source for the clock circuit. U11 divides the 455 kHz by two to produce a 227.5 kHz +/- 1% two-phase clock with a 20% "active" time. (The TI-58 can use a 384 kHz reonator which is the preferred part.)

Back in the 1980's when I was the editor of TI PPC Notes I received a letter from a Canadian who had what he called a "5/6 speed TI-59". It turned out that he had a TI-59 in which a TI-58 resonator had somehow been installed. As a result of looking into that we were told that the TI-58C uses a resonator with a frequency of about 404 kHz.

#26

My first HP was an HP-67 that I bought in 1977. Later in 1977 I also bought a TI-59 and PC-100A. The TI-59 was in fact a much more advanced, sophisticated, and capable machine than the HP-67, especially with the solid-state software modules and the PC-100A. However, I went through five TI-59s and three PC-100s (ending up with a PC-100C) between 1977 and 1980. The TI hardware was just too "iffy" to trust for any critical application. The programs I wrote for use on my nuclear submarine had to be on the much more reliable HP hardware of my personal HP-67, and a Navy-purchased HP-97.

There were some significant differences between the PCBs of all of the five TI-59 units that I used. That seemed to indicate to me that there were many continuing problems with the TI hardware. Still, I'm fond of the TI-59.

#27

Quote:
I went through five TI-59s and three PC-100s (ending up with a PC-100C) between 1977 and 1980.
Interesting. They must have gotten the problems worked out. I got my 58c at the end of '81 and 59 a little later and never had any trouble with either one. I used them a ton.

Edited: 8 Jan 2010, 12:00 a.m.

#28

I bought a TI-58 before the TI-59. Shortly after, I traded the TI-58 to a shipyard engineer for his old obsolete HP-35. I still have it, an early red-dot 1143A version. It was in used condition but complete with the manual and all accessories.

I didn't know then that there was anything special about the red-dot version. I'm happy with that trade that I made 33 years (one-third of a century!) ago. I suspect that the TI-58 has been in a landfill somewhere for most of that time.

#29

After the fourth TI-59 failure (and my unit just past the warrantee for the last replacement), I wrote directly to the TI CEO (Mark Shepard, IIRC) describing the history of failures. I'm sure he never saw my letter, but I received a call from a TI engineer. TI covered the cost of shipping the TI-59 and PC-100A back to them while they sent new units that the engineer claimed had been thoroughly run thru testing and burn-in. The replacements were still working 18 months later when I "loaned" it to my brother in college (never to be seen again). :-(

#30

I, too, must have been lucky. I bought mine in high school in 1978 and it lasted flawlessly through college. It was "babied", however, and always kept in a padded case. I had ended up with one of those wonderfully padded SR-52 cases with two inside pockets.

I'm still considering a "drop testing" one of my now many 59's just for fun... I mean, study, to see what kind of physical tolerance they really have. I doubt they could go through snow blowers like the HP-25's....



Possibly Related Threads…
Thread Author Replies Views Last Post
  HP Prime: FLOOR, iPart , and their use in programs Alberto Candel 6 2,555 12-01-2013, 10:17 PM
Last Post: Alberto Candel
  HP Prime Save Power Turn Off Not Working Timothy Roche 12 4,078 10-27-2013, 01:41 PM
Last Post: Michael de Estrada
  Temporary User Mode Key Programs not working in RPN BruceTTT 7 2,624 10-14-2013, 01:46 PM
Last Post: BruceTTT
  HP41CX: Strange serial #... Stephan Matthys 1 1,132 09-13-2013, 01:38 PM
Last Post: Stephan Matthys
  HP to be dropped from Dow Jones industrial average BruceH 21 5,012 09-12-2013, 02:45 PM
Last Post: Chris Tvergard
  HP-97 suddenly not working Brad Barton 10 2,724 02-25-2013, 08:15 PM
Last Post: Brad Barton
  [41CL] Problem with PowerCL ... not working Doug (NYC) 5 1,932 01-22-2013, 07:18 PM
Last Post: Doug (NYC)
  42S exit key not working dwsmith 4 1,626 11-05-2012, 10:59 AM
Last Post: dwsmith
  HP-46 not working Ignacio Sánchez 1 995 10-24-2012, 11:17 AM
Last Post: Donald Williams
  [WP34s] last QtGui version not working fhub 8 2,448 08-31-2012, 05:08 PM
Last Post: fhub

Forum Jump: