Let's get a charge out of batteries.

42 responses total.

Why 13.8 volts? My HT batteries have charging ports that say "13.8 volts".
I have never seen a labeled 13.8 volt source for charging. I have measured
something above 13 volts from a 12 volt (car) battery, and I have a "CB
Power Supply" that is stated to provide "13.7 to 14.5 volts". I have also
charged one of those NICADS that say they want "13.8 volts" by plugging
them into a car cigarette lighter (overnight - engine off). So, what gives
with that "13.8 volts? Does it really mean what it says? I've never heard
of a wall-adapter rated for 13.8 v (though you can get them rated 14
volts). So, what's going on here, and why? 

Lead-acid batteries are only nominally 12 volts.  The exact voltage
depends on how charged the battery is, on recent charge/discharge
history, and the temperature of the battery.  This leads to a "cold
start" trick - if the battery is weak, and doesn't have quite enough
oomph to start the car, turn the headlights on for 30 seconds, then turn
them off & try again.  The drain on the battery will heat the battery
up, which may raise its voltage enough to enable the car to be started.
Automotive electrical systems usually "float" on the battery, so the
exact voltage depends mainly on that.

The generator (or alternator on most modern automobiles) is responsible
for charging the battery, but the battery must also not be overcharged -
attempts to do that actually produce steam, hydrogen, and oxygen.  The
latter two can actually produce an explosive mixture, while all of these
use up the water in the battery.  Fortunately, it's easy to determine
the state of charge in the battery, so a simple voltage regulator
suffices.  I believe a quiescent fully charged battery at room
temperature is about 13.8 V.  The voltage regulator will probably try
for about 14.1 V.  A nearly discharged battery may read 12 V.  Even a
completely discharged battery may read 9 V under no load circumstances.
This can result in odd behavior if the battery is nearly dead - for
instance, the instrument panel lights may light up as expected (only
slightly dimmer), but as soon as the starter is engaged only a click
from the starter solenoid may be heard.  If the battery doesn't have
enough oomph even to keep the solenoid closed, a rapid series of clicks
may be heard instead.  In many cars, the "battery" light on the
instrument car is an essential part of the charging circut - and must be
replaced if it does not light when the ignition is turned on.

When using that trick of turning the headlights on to warm the battery, it
is a good idea to do it only to batteries that otherwise work well.  I had
a battery for a while that would tend to completely die anytime it had a
lot of trouble starting the car, and turning the headligyhts on while it
was having problems would drain it very quickly.  OTOH, the battery that
replaced that one works much better.  I left the lights on with the car
turned off for three hours today (by accident) and the car didn't even
hesitate when I started it.

Marcus' comments are pretty reasonable. However, how did ICOM (and others)
settle upon specifically "13.8" to have embossed on their equipment, when
it is undoubtedly a range of voltages that can be accepted? The manual
even says "connect an external 13.8V DC power source" when, at least
literally, there is no such thing. Here's a related question: these
battery packs have two charging jacks - one for a wall-adapter for
charging, and the other for "13.8V". I know the former has the charge
limiting circuit in the adapter, and the latter in the battery pack, but
why provide both? A "13.8V" wall-adapter would suffice. 

At a shear guess - it may be historical.  The "wall adapter" may have
come first, followed later by the "13.8v" connector.  Or, the two may
have originally started with two different companies, and both may have
then become industry standards in tandem.  If the battery to be charged
is in fact lead acid, then "12 V" may in fact be very wrong, and it may
well need something that is actually about 13.8 V for correct operation.
Since there are many power sources that deliver a regulated 12 V output,
specifying 13.8 V may well be necessary to ensure people don't try using
things that won't work.

On the other hand, if the battery to be charged is a NiCd, then things
get even more interesting.  NiCd batteries have a very flat performance
curve, so it's difficult to tell from the voltage whether the battery is
nearly charged, or nearly discharged.  For maximum results, NiCd
batteries should be fully charged just before use, and then used until
they are completely discharged.  If NiCd batteries are repeatedly only
slightly used, then charged right back up again, they develop "memory",
and the voltage drops off and the battery acts "dead" much sooner than
it should.  This is very different from lead acid batteries, which like
to be kept fully charged, but don't like to be repeatedly deeply
discharged.

1991 ARRL Hndbook: "Most modern radio equipment is powered from 13.8V
...." The Handbook also shows several power supplies and lead-acid battery
chargers that provide 13.8 volts. 

I think I have a hint: apparently, the "float" voltage for the lead-acid
battery is very close to 13.8 volts. That is, at that voltage supplied
from a charger, the "charge" rate becomes very small. Actually, it doesn't
charge, but there is water electrolysis, but at an "acceptable" small
rate.  Therefore a 13.8V power supply is a "standard" charging circuit.
Presumably, then, other battery-pack designers worked to that standard. 
(The water-electrolysis voltage is really closer to 1.5 volts, but the
chemistry of the lead-acid battery supports 0.8 volts of "overvoltage", so
water electrolysis doesn't start until 2.3 volts per cell, or 13.8V for 6
cells). 

I *have* seen a labeled 13.8V power supply, though not specifically for
charging. The Astron power supplies for ham equipment provides 13.8V
+/- 0.05V. This doesn't get into the historical origin of the choice,
but illustrates the standard.

My Heathkit HW4M HT has two plugs too.  However, one is for the battery
charger and the the other, labeled "DC In", supplies power directly to
the HT, bypassing the battery.  Connecting power to this connector does
not charge the battery.  It is intended for mobile operation where it is
connected to the mobile systems power source, most commonly a lead-acid
battery/charging system.  Almost a decade ago, I did some work with UPS
systems for heart-lung machines.  The chargers for the 12 volt lead-acid
batteries would start out charging at a current = 1/10th the amp-hr. rating
of the battery until the voltage across the battery reached 14.2 VDC.  At
that point the voltage went back to 13.8 VDC and held there indefinatly.
This charging procedure resaulted in the quickest charging of the battery
without doing damage to it.  They would have charged just as well had we
simply a charger held at a constant 13.8 VDC with a cow-bar curent limiting
circuit to keep the current from getting too high at the start.  It would
take longer to charge though.  I suspect 13.8 was chosen since an ideal
charging circuit will maintain this voltage on your standard 12 VDC lead acid
battery.  BTW, 10.7 VDC (Under normal operating load.) is considered "deep
discharge" for a deep-discharge lead-acid battery.  Any less and you risk
permanant damage to the battery.  The "discharged" voltage of a regular
battery is somewhat higher.  I'd suspect somewhere between 11.5 and 12.5
VDC.

My ICOM 2GAT requires a separate adapter for running directly from the car
battery; the battery jacks are just for charging. Now that we realize that
13.8 volts is a (nominal?) standard for the float charging voltage for a
lead-acid battery - it doesn't seem to make much sense to specify
"exactly" that for the charging source for NiCd battery packs, which have
their own current limiting or more sophisticated charging circuits. Or,
does it?

The sophisication of most HT chargers consists or a resistor in series
with the battery.  So as not to get into trouble with UL and a host of
other regulation agencise, they prob. have to keep the wall-wart voltage
below 24 VDC.  Being able to charge the battery from a lighter plug is
also nice.  Hence this series charging resistor is not of a high value.
Small changes in charge voltage can mean big changes in charge current
for this reason.  They could put a more tolerant charging ckt. in but
that would take space and people already gripe about having to pay $50+
for a new battery.  (Well, I do!)

The nominal charging current of 10% of the ma-h capacity in ma, is
pretty rough too. Even with just a current limiting resistor, I'd
think there is some tolerance on the charging voltage. So, what would you
say? From the nominal pack voltage up to 20V? (ca. 15%, rather than 10%)

Using Ohms law, it should be pretty simple to calculate: 
Resister = ((Supply voltage) - (Battery voltage)) / (Desired charge current)
 
Unless it's a battery designed for fast-charge, don't go to far beyound the
10% charging rule.  Doing so WILL reduce the life of the NiCad pack
significantly.  Fast charge NiCad packs and chargers have termal sensors
to sense the temprature of the pack during charging.  They look for a sudden
change in temp. to determin full charge.

Figuring the charging current is simple too:
(Charging current) = ((Supplied voltage) - (Bat. voltage)) / Resistor

How long it takes to charge a battery pack is also important to most users.
I would guess that the charging efficiency is somewhere around 85%.
 
To find out how long it will take to charge the battery pack:
(Time to full charge) = (Amp Hr. capacity of pack) / (Charge current in Amps)
 / (.85)

With that, it sould be possible to design a reasonable charger to suit ones
needs ;-)

Just after I had hit Return on #11, I realized that I had not considered
the battery voltage"! This will be rather important for my 13.2V HT
battery pack which says charge from 13.8 volts. If it is just a resistor,
going to a 14.4V source would double the charging current. Since you can
get 14.4V while charging a charged lead-acid battery, I wonder if the
battery pack isn't a bit more sophisticated. Perhaps they choose the
resistor to give the maximum charging current with (say) 15V, and include
a 7815 15V reg ($0.67 retail) with the resistor to set the max charging
current at 15V (and it would take longer to charge with 13.8V source).  A
few tests of battery packs with a variable voltage source and ma meter
should tell the tale (but I don't have the former). 

Indeed!  The problem with a 7815 is that a 78XX type regulator requires
about 3V of "headroom" to regulate properly.  i.e. a 7815 would require
a 18V input to regulate to 15V.  There are "low dropout" regulators, but
even these require about a volt of headroom.  When you consider that a
silicone junction requires about 0.7V to overcome, you'll find that 
designing an active regulator is tricky.
One trick would be to have a switching circuit that "breaks" your 13.2V
pack in half, creating two 6.6V packs, and than placing the two halves
in parallel for charging.  If you figure out the switching for this, you'll
see it is not simple and might be prone to problems.  I've done this and,
if memory serves, it required a DPDT switch.
You'd be wecome to run some experiments in my basement some evening Rane.
Mad scinentists have all that kind of stuff ;-)

A battery pack for a 2GAT HT, which I thought I had charged not long
before, went dead after ca. 20 minutes, during the Walkathon, so I looked
into it further. It is a pack that holds 8 replaceable AA NiCads, and
hence has a nominal voltage of 9.6. I charged it fully and then discharged
it rather completely through 100 ohms; it went dead within an hour. I then
repeated the charge/discharge, as a "conditioning" exercise. After
recharging again, I decided to have a look at the discharge curve, so
recorded the pack voltage every half-hour while discharging through 100
ohms. The discharge curve looked like (t in hours, V in volts): 

t   0     0.5   1.0   1.5   2.0  3.0  3.5  4.0  4.5  5.0  5.5  6.0  6.5  6.75
V  11.22 10.46 10.14 10.04 10.02 9.96 9.92 9.85 9.75 9.64 9.48 9.21 7.27 5.34 

The area under V vs t, divided by R, is the A-H capacity, which for this
discharge is 0.646 A-H, very close to the nominal 0.6 A-H. Therefore, all
the cells took a full charge. This appears to be a way to test a NiCad
battery park, even though it is somewhat tedious to read the voltage at
intervals for nearly 7 hours (hmmm...a computer could do that, including
terminating the discharge at a desired voltage. Anyone know of a circuit
for building such?)

The charging current initially was 65 ma (ca. 0.1C), while at full charge
it was 45 ma., from a 14+ VDC source. The back resistance on the + side
was "infinite", and the forward resistance with an ohmeter was 150K, which
means there is at least a diode and apparently a current-limiting circuit,
and not just a resistor. 

I thought this was an interesting exercise. I'm going to try it next
on the regular BP-70 pack for the 2GAT.

I've never tried this exercise. It seems well done, but I'm perplexed by
the numbers you got. A nicad cell is rated at 1.2 volts. 8 cells should yield
9.6 volts. Yet your pack put out over 9.6 volts for 5(!) hours. This just
doesn't seem right.

You initially said that the pack went "dead" after a 1 hour discharge through
a 100 ohm resistor. What voltage level do you associate with "dead"?

After recharging, it was able to deliver charge for 7 hours? That's a
rather miraculous recovery. Is it possible you have a loose connection in
your charger that resulted in incomplete charges the first couple of times?
Since you were now in the process of actively "fiddling" with the whole
setup, you could have wiggled whatever was loose in the charger back into
a temporary working condition. Just an idea.

I just checked the 2 battery packs for my Makita drill. The fully charged
one, charged about 3 weeks ago, was at 10.30volts. The partially discharged
pack in the drill was at 10.1volts. I ran that pack down  to the point where
it could no longer turn the drill and then left it loaded for an additional
2 minutes. Immediately upon removal from the drill it was at 2.2 volts but
it was climbing back up. After 10 minutes it had "recovered" to 8.85volts
and was stable at that voltage. Interestingly, even though this cell was
"dead", it put out 8.85 volts at high impedence *and* into a 100ohm load.

After fully charging it, the pack was at 11.45volts. This dropped off after
a couple of hours to a level of 10.95volts.

Hmmm, so much for that 1.2volts per cell.

I think one thing to remember here is that a rechargable battery(lead-acid,
gel-cel, nicad, whatever) is always "high" when it first comes off the
charger. Close to it's charging voltage. But pretty quickly stabilizes
at it's normal voltage. How long after coming off the charger did you run
your discharge test Rane? Another idea is to try the same sequence with
no load and see what the battery's output curve is like.

  (And this from the guy who said counting people at the Ypsi post
office was a dubious expenditure of time?  ;-)
 
  Btw, what do folks think of the new "rechargable alkaline" batteries?
I've heard they have shorter lives overall, but longer per-charge use.
I'm thinking of phasing out most of my NiCads in favor of alkalines,
to avoid the inconvenience of 10 minute battery charges.

Depends on what you are using them in. The alkalines would probably work
fine in a low current device, but I suspect they wouldn't work as well
in a battery-powered power tool. Nicads are kind of halfway between being
a battery and a capacitor. You can pull a great deal more current off
of a nicad per unit time, than you can from a regular battery. In effect,
you're sucking electrons directly off of a charged plate, rather than
waiting for the comparitively slower effects of a chemical reaction
to proceed. This comes in handy in power tools that need to deliver high
torque into high loads.

Re #17: I have practical use for batteries and their behavior. The safety
of people at the Walkathon could have depended upon my battery status
(though it was the supply of oranges that really did). I have absolutely
no practical use for counts of the traffic at Ypsi post office. You must,
since you comment on this. What is it? 

I have repeated my experiment with the BP-70 battery pack for a 2GAT,
which has a nominal voltage of 13.2, and capacity of 0.27 A-H. The
discharge curve with a 149 ohm load was: 

t(min)     0   20   40   60   80  100  120  140  160  180  200  220 227     
V(volts) 15.0 14.0 13.6 13.4 13.3 13.2 13.1 13.0 12.8 12.3 11.3 7.7 6.0

Integrating the curve, as before, gives a capacity of 0.32 A-H. Another
"OK" battery pack. 

These discharge curves resemble those for AA NiCads shown in the Radio
Shack Battery Guidebook - an initial sharp drop (after charging), followed
by a slow drop until "capacity" is reached (though as shown in the
Guidebook, "capacity" depends on the discharge rate). 

I started these experiments immediately after charging. Higher than
nominal voltage arises from the electrolyte composition between the plates
not being uniform due to the internal flux of ions during charging. In
principle, if I had waited after charging (long enough) nominal voltage
would be attained. The same occurs during discharge: ionic species are
consumed or produced at the electrodes, but do not diffuse rapidly enough
to be uniform in composition everywhere. Battery voltage depends upon the
electrolyte composition at the electrodes (as well as gradients in this
within the bulk electrolyte), so the voltage has some dependence on
charge/discharge history. (The phenomenon of a non-uniform distribution of
electrolyte composition is called "polarization".)

The observation that the 9.6 V (nominal) pack "went dead within an hour"
may have been due to my not charging it long enough - my experiment was
initially not as controlled as would be desirable! I was going to consider
"dead" as 1 volt/cell (the books say to discharge only to this level), but
I ran the pack down to ca. 3 volts. 

After discharge of the pack with separate AA cells, I measured each cell
with a RS "battery tester". One tested OK, 2 tested "so-so", four tested
"low", and 1 tested "dead" (these are as shown by the red and green LEDs).
This is not surprising as the cells cannot have *identical* capacities,
and the final discharge cut-off is very steep. 

Re NiCads vs rechargable alkalines: the suggested load current range for a
AA NiCad is 0 to 2.0 amperes, and for an (ordinary) alkaline is 0 to 0.25
amperes (and for a carbon-zinc AA, 0 to 25 milliamps!). I don't have the
value for a rechargable alkaline, but I think it is the same as for an
ordinary alkaline, the difference being only that it can be "recharged".
(They cannot, in actually, be "recharged", but only depolarized, which
appears to extend their lives, since even ordinary alkalines have *much*
more zinc in them than is used.)

The discharge from NiCads comes only from the chemical reaction at the
electrodes - not any capacitive effect. However the NiCad has a great deal
more electrode area: they are built with strips of Cd and Ni (converted to
Nickelic hydroxide) separated by a thin layer of electrolyte, and rolled
into a cylinder. This construction is like an electrolytic apacitor, which
is what I think Grex was referring to. 

Incidentally, my experience has been that it is best to use NiCads until
they are nearly fully discharged, and then recharge, rather than leaving
them to float on charge. This, unfortunately, is not the best mode of
operation of insuring an always-ready, fully-charged battery. In some
applications, one can do a "deep discharge" at intervals, and recover
NiCad capacity. I have a utility for controlled deep discharge of a Mac
Powerbook battery, but have not seen the equivalent for general use of
NiCads. 

re 17:
        I tried using a Rayoovac Renewal battery (rechargable alkaline) AA
battery in my pager, and it died after two days and two pages.  I decided
that, even though it might be cheaper just to have a few Renewals and
rotate them, rather than using non rechargable alkaline batteries, it
really wasn't worth the hassle of having to change the battery that often.
 I went back to nonrechargable alkaline batteries, which tend to last two
or three weeks.

In #19 I called Greg "Grex" - a Freudian slip.

Heh, yes, I was going to mention that.

Yes, I was thinking of nicads being similiar to electrolytics in their
construction and their ability to shed a charge at a much higher rate
than a normal, one-shot, chemical battery.

Your observation about nicads jibes with most recomendations for thier
use. I nicad should be used to complete discharge before it is put back
in the charger. As opposed to constant topping-off. They are meant to be
"deep cycled". Thhis is the opposite of a car battery, which should always
be kept at top-charge and never deep cycled.

Er, change "I nicad" to "A nicad" above.

_I Nicad_? Wasn't that the Asimov book about the little battery that could? :-)

  Re #19, just kidding Rane...I thought you were doing this just out of
curiosity, which is the reason I'd watch a PO; didn't realize it's a
serious safety-related endeavor!
 
  I've heard of the full-discharging devices for powerbook batteries...I
wonder why they aren't used in regular NiCad chargers.  It seems like a
lot of my battery-consuming devices become unusuable prior to the batteries
being completely drained.  Should I hook up a resistor or something to
drain them more completely?
 
  Most of my uses are occasional...an electronic scale, flashlights, a smoke
detector, a mini tape recorder, that sort of stuff.  It *seems* like my
NiCads just wear out too quickly on those things.  Any battery advice for
those sorts of things?
 
  I use alkalines in my pager too, and get about 2-3 weeks out of them.
I start losing bits of pages a few days before my low battery indicator
goes off, at which point I'm losing big chunks of pages.  (An alpha-pager,
so it's not as critical as with a numeric).

Hmmm, that's interesting. My alpha pager continues to recieve
good pages *long* after the low-battery indicator has gone off.
In fact, almost dead batteries affect the display quality alot more than
they affect message content. I almost never lose characters with dead
batteries. Until, of course, it just stops working entirely.

I'm using a motorola advisor, what is your's?

I think a NiCad dischrager could be built easily from a resistor and
a Zener diode. Just as an off-the-cuff notion, pick the resistor to
produce a discharge current of (say) 0.15 C (i.e., if 600 ma-H, discharge
at 90 ma). The Zener limits the minimum voltage to which to discharge.
This is supposed to be 1 volt/cell, but doesn't seem critical (?). RS
has 1 watt Zeners for 5.1, 6.1 and 9.1 volts, so one would have to
discharge pakcs, or cells in series (say, 6 or more). For lower
voltage packs, like in telephones (e.g., 3.6 volts), you'd want to
discharge against a back voltage of (say) 2 volts. An LED might serve
for that purpose (though I haven't tried it). Oh yes - check power
dissipations, so as not to exceed either resistor, zener, or LED ratings.

  Re 24, Motorola advisor as well.  I've got a replacement on order, as
it's been periodically switching off/on, so maybe my next one will be
better.  My signal dropout depends on both battery age and on location;
I'm using AirTouch (formerly PacTel), mostly in SW Ann Arbor.  I used
PageNet a couple years ago, with the same problem, but it was much worse
in most areas, and also suffered seasonal quality fluctuations...AirTouch's
transmitters self-recalibrated, and PageNet's didn't, or something like
that.
 
  I just remembered a set of portable RF terminals I installed that had
a "discharge" button on the charger, and it did make a big difference.
Getting people to remember to push the button wasn't easy, but it soon
became clear why it was important!  It was the difference between 4 hrs
vs. 8 hrs use or thereabouts.
 
  Hm, why shouldn't the battery be discharged beyond 1V?  Is there a
discharge point of no return?
 
  Btw, on the subject of batteries, I most heartily recommend Sears
Dual-Start batteries...mine has saved me from needing jumps around five
times.  When the primary is dead, you flip the switch to the secondary,
drive around while it charges the primary, then flip it back...it cost
nearly twice as much as a regular cheap battery, but it's been worth it
to me!  I'm surprised it hasn't achieved more popularity.

Hmmm, my pager stills says "PacTel" on it. So we're using the same service.
I have always been amazed how *reliable* this damn thing is. You must
have a defective pager. Everybody in the company I work for has
these and they work really well.

I've been using an Advisor (from PageNet of Pittsburgh) for the past
two years or so.  About 1.5 years into lifetime of the first pager,
it developed what looked like a loose connection -- the pager would
periodically power-cycle.  Rapping the pager on a desktop (or
dropping it on the floor) could usually cause a power-cycle.  I
called PageNet and asked for a replacement pager, which they
supplied at no charge (the benefits of leasing a pager, rather than
owning one...), and which fixed the problem.

That's the problem I had until recent replacement, although I didn't even
need to rap it for it to turn on again.  I was talking to another person 
last week whose Advisor is doing the same thing.  Perhaps it's a design
feature to encourage people to rent rather than buy Advisors?  :)

My Powerbook batttery has died on me (again), so I forked over $50 for
another. But surely even an Apple battery can't be complicated, so I
did what we are cautioned against and opened the battery pack.

This is a model MC-170. Inside are ten AA size nicads arranged as two
batteries of five cells each wired in parallel. There is no other
circuitry, except for a small metal cased unit wired in series within
each 5-cell battery, between cells 4 and 5 for on battery and cells 1 and 2
for the other. The unit is labeled "PEPI H". It reads zero resistance
(and zero volts). What is it - a fuse?

I was surprised by the parallel arrangement of the two batteries. I suppose
Ni-Cads even themselves out in this configuration, as if one drains faster,
its voltge falls, and it is recharged by the other. However if just one cell
falters - shorts internally, as they will - then the good side will drain
through the bad side, and the whole pack is kaput. I would think a diode
on each side would be useful to prevent this - except for the voltage drop
of a diode. 

In the case of this pack, the voltage is 6.13 (even though it is discharged),
which is OK for 5 NiCads. Each cell also tests close to its nominal 1.2 volts.
Yet, the pack won't hold a charge. Why do you think that is? (I exercise
my batteries by discharging them to 1 volt each - 5 for this pack - at
intervals, and then recharging, and of course not leaving them on charge
for over long.)

Now, I should be able to find 10 Nicads, and restore this pack for ca. $15
(plus several hours of time.....) instead of $50. The original cells are
spot welded together via tabs, but I'd have to solder - and there is
essentially no room for any thick solder. Any advice on soldering to the
tops and bottoms of NiCads (except, don't get them too hot...)?

The main reason you should never run a battery pack below 1 volt/cell is
that the cells don't discharge evenly.  If you run the pack down until it
reads 0 volts, some cells will be reverse-charged.  Then they'll *really*
heat up when you try to charge 'em.

 You will likely find just one bad cell in each battery.  So you could 
fix it even cheaper, except that other cells are likely to die soon 
anyway.

Each cell tests at nominal (1.2) volts (5x1.2 = 6.0, and the stack tests
at 6.1 volt). This is with the battery discharged to the point that it
would not let the Powerbook boot. When I recharge for the full charge period,
I get only ca. 15 minutes of operation. Yes, if just one cell in each
battery had negligible remaining *capacity*, it would discharge quickly, drop
the battery voltage to below ca. 5.3 volts, and the computer would shut down.
What is going on in a NiCd such that its reads 1.2 volts "open", but has
very little remaining capacity? [I agree with you that replacing just one
cell in each battery, if I could find the culprits, would not be a long
term fix.] 

Hmmmm...what if I discharged each cell *separately* to 1.0 volts, and then
recharge the pack? Could just one cell in each battery get into "memory
effect" mode? That would explain what I observe.

 My experience with individual NiCd cells in regular groups (I always 
keep the same cells together) is that one cell eventually dies well 
ahead of the others.  It still has normal voltage, but a very short 
charge life.  Replacing it with a fresh NiCd cell restores the group to 
full capacity.

PEPI makes thermal switches.  This way, if your battery pack gets too
hot, they will open and keep the batery pack from melting down and
doing all sorts of inappropriate pyrotechnics.
 
Rane, you need to check your NiCd fully charged and under
a working load.  Checking batteries under no-load conditions will tell
you little about their health.  You will also want to monitor the voltage
of a battery under test for about a minute before you decide it's good.
Even a bad NiCd cell can look OK for a few seconds.  I do my cell 
testing with a flashlight bulb.  The bulb is a reasonable load and the
brightness of the buld is a relative indicator of cell voltage.



 
If you want to restore the pack to peak preformance and life, replace
all the NiCd cells.  Otherwise determine the bad ones and replace just
those.  To solder to them, I use liquid acid flux.  Just slightly wet the area
you want to solder to and then touch a hot iron, with solder already molten
to the tip, to the wetted area for a second or less.  If you did
everything right, the end of the cell will be tinned.

?????acid flux???? ummmm, what's wrong with this picture?

I would like to be able to change batteries in some instruments *from
outside*, so the cases do not have to be opened. I've looked for
battery holders that are made to do this, but I haven't found any
in catalogs. I have in mind something like large fuse holders that
will hold (say) two AA batteries, and one just has to unscrew a lid
and exchange batteries. They could even be made as paired holders
for four AAs. Does anyone know where such holders are available and,
if not, does anyone have any ideas for building such? I have considered
cutting big rectangular openings in cases and mount (say) flat 4X AA
holders, and close the hole with a plate that requires only two screws,
but those flat 4X holders are flimsy and not made to be readily mounted
against a chassis (though one could make-do with brackets that one makes).
Any ideas?

I have some aluminum AA holders you'd be welcome to.  They are designed
to be screwed to a chassis or board.  I think Purchase has them in stock
too.

I presume that they mount with the closed side against the chassis? I
would need the opposite, so they would face a (largish) window cut
into the chassis. But it does sound like they'd be strong enough to
rearrange the mounts. I'd like to see one. Will you be at the next
ARROW meeting? (You better be, since you are giving the program....)
How about tube types (like fuse holders)?

- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss