greg@amdahl.uts.amdahl.com (Greg Bullough) (12/02/88)
The other night, as I was walking by the room where my XT sat running, I heard a bang, like the sound a TV makes when its high voltage arcs. In fact I assumed the monitor had arced, due to dust or whatever. Next time I brought the system up, I noticed that the SEFCO clock wasn't working. I bought a new battery, and when I went to install it, found that the old battery had exploded and burned, with enough force to bend the battery clip back to a 45-degree angle. There were chunck of battery innards all over the inside of the inside of the PC. This was no quiet self-destruction. Investigation showed that the SEFCO clock card had +5 volts across the lithium battery terminals! Now it's hard to determine the circuit from the PC board, but there doesn't seem to be much that protects the battery from the supply voltage... Interesting, indeed. Greg
berger@clio.las.uiuc.edu (12/07/88)
There should have been a diode preventing the voltage from being
applied to the battery. Was the diode shorted, or missing alto-
gether?
Mike Berger
Department of Statistics
University of Illinois
berger@clio.las.uiuc.edu
{convex | pur-ee}!uiucuxc!clio!bergergreg@amdahl.uts.amdahl.com (Greg Bullough) (12/08/88)
In article <16800395@clio> berger@clio.las.uiuc.edu writes: > >There should have been a diode preventing the voltage from being >applied to the battery. Was the diode shorted, or missing alto- >gether? > Well, there's more to the story, now. The card was replaced by a very concerned dealer. I plugged in the new card, pulled the battery, put a VOM across the clip, and turned on the computer. VOILA! +5 VOLTS! Great design, huh? I cannot find a blocking diode, so I assume it was left out of the design. Can anybody correct me if I'm wrong. It IS a no-no to put a positive over-voltage across a lithium battery, isn't it?? Greg
pnelson@antares.UUCP (Phil Nelson) (12/08/88)
In article <dcKUudf=Im1010ncJtw@amdahl.uts.amdahl.com| greg@amdahl.uts.amdahl.com (Greg Bullough) writes: ..cut some stuff.. |Well, there's more to the story, now. The card was replaced by a |very concerned dealer. I plugged in the new card, pulled the battery, |put a VOM across the clip, and turned on the computer. VOILA! |+5 VOLTS! Great design, huh? I cannot find a blocking diode, so I |assume it was left out of the design. | |Can anybody correct me if I'm wrong. It IS a no-no to put a positive |over-voltage across a lithium battery, isn't it?? | |Greg In a word, yes. One thing though, if you are measuring with a good DVM, you might want to re-try with a load, say 10K ohms. A good DVM will read voltage supplied by minute (and inconsequential, in this case) leakage current. -- Phil Nelson at (but not speaking for) Tymnet, McDonnell Douglas Network Systems Company POTS:408-922-7508 UUCP:{ames|pyramid}oliveb!tymix!antares!pnelson LRV: Component Station
skl@van-bc.UUCP (Samuel Lam) (12/09/88)
In article <dcKUudf=Im1010ncJtw@amdahl.uts.amdahl.com>, greg@amdahl.uts.amdahl.com (Greg Bullough) wrote: > The card was replaced by a > very concerned dealer. I plugged in the new card, pulled the battery, > put a VOM across the clip, and turned on the computer. VOILA! > +5 VOLTS! Great design, huh? I cannot find a blocking diode, so I > assume it was left out of the design. > > Can anybody correct me if I'm wrong. It IS a no-no to put a positive > over-voltage across a lithium battery, isn't it?? Perhaps that particular card was designed to use NiCad battery (rechargable) instead of lithium battery (non-rechargable). The +5 volts that is used to recharge the NiCad while the computer is on will probably eventually blow up a lithium which found its way there. -- Samuel Lam {alberta,watmath,uw-beaver,cs.ubc.ca}!ubc-cs!van-bc!skl
mbe@dde.uucp (Martin Berg) (12/15/88)
In article <dcKUudf=Im1010ncJtw@amdahl.uts.amdahl.com>, greg@amdahl.uts.amdahl.com (Greg Bullough) writes: > .. > Well, there's more to the story, now. The card was replaced by a > very concerned dealer. I plugged in the new card, pulled the battery, > put a VOM across the clip, and turned on the computer. VOILA! > +5 VOLTS! Great design, huh? I cannot find a blocking diode, so I > assume it was left out of the design. > > Can anybody correct me if I'm wrong. It IS a no-no to put a positive > over-voltage across a lithium battery, isn't it?? What input impedance does your VOM have ? Could it be that it is so big that even a large resistor (say > 100 Kohm) from the +5 V to the battery could give this result ? If this is the case, the current flowing from to supply to the battery wouldn't hurt the battery. By big input impedance I mean something like 5-10 Mohm. -- mbe@dde.dk | "The answer is 42" or | A. Douglas ..mcvax!enea!dkuug!dde!mbe |
greg@uts.amdahl.com (Greg Bullough) (12/17/88)
In article <119@mother.dde.uucp> mbe@dde.uucp (Martin Berg) writes: >In article <dcKUudf=Im1010ncJtw@amdahl.uts.amdahl.com>, greg@amdahl.uts.amdahl.com (Greg Bullough) writes: >> .. >> Well, there's more to the story, now. The card was replaced by a >> very concerned dealer. I plugged in the new card, pulled the battery, >> put a VOM across the clip, and turned on the computer. VOILA! >> +5 VOLTS! Great design, huh? I cannot find a blocking diode, so I >> assume it was left out of the design. >> >> Can anybody correct me if I'm wrong. It IS a no-no to put a positive >> over-voltage across a lithium battery, isn't it?? > >What input impedance does your VOM have ? > I used an 100,000-ohm/volt FET VOM for this debacle. >Could it be that it is so big that even a large resistor (say > 100 Kohm) >from the +5 V to the battery could give this result ? There was a resistor from the battery to the supply. 100 puny ohms. That reminds me... I should check the total resistance from the B+ line on the pin-out of the card to the battery positive. I don't THINK there are other resistors in the circuit, but I may be mistaken. In any case, a little diode has provided the proper kind of resistance and settled the issue entirely. The potential on the battery never rises above it's rating, now. Nothing like a bit of dining-room-table engineering. >If this is the case, the current flowing from to supply to the battery >wouldn't hurt the battery. Ah, yes, but the key was that the current (or something else) hurt one battery VERY badly. And it takes a lot more heat than is ambient in that section of my PC to cause a spontaneous explosion of a lithium battery. Greg