April DECUServe Journal
Sharon Frey
The DECUServe Journal
=====================
April, 1994
This issue is short but sweet, folks. It features an interesting
article on recharging batteries, and one on wierd detached processes.
I'd like to take this opportunity to solicite help from the Internet
community. I need someone to take over the duties of Journal Editor for me.
Any individual or a small group that can work as a team is welcome to apply.
The only requirements are access to DECUServe (if you have Telnet, you have
the best (free) access!) (if you don't have an account on DECUServe, but
would like one, this is your chance!), and some editing software on a computer
of some type. Personally, I use Sled on a plain old 386 machine, and TPU
on the Vax.
The reason I need help is because I've just recently started a new
"real" job (editing the Journal is a volunteer "job"), that will require more
personal time from me than previous jobs. I'm going on 24-hour call next week,
and I'll be on call every other week from then on. Whew! "It's not a job,
it's an adventure!" :-)
Some of the niceties of being Journal Editor are full creative freedom;
a limitless supply of resource material (no kidding!); and of course, access
to DECUServe, that bastion of bizarre humor, generous intellects, and good
taste!
All interested parties please send me email. And on to the fun...
Table of Contents
-----------------
Letter fron the Editor .............................. 1
Technical Information ............................... 25
Articles
Wierd detached process login failure messages ... 2
Rechargable batteries ........................... 6
April, 1994 The DECUServe Journal Page: 2
Wierd detached process login failure messages
---------------------------------------------
The following article is an extract of the DECUServe VMS
conference topic 2284. The discussion occurred between
March 18, 1994 and March 28, 1994.
By Larry Stone, Jphn Briggs, Michael Spatz, Simon Maufe, Jean-Francois Mezei,
Pat Scopelliti, Joe Crum, Jonathan Prigot, Mark Hyde, Lee Gleason, Jay Thomas
(03/18/94 Stone)
----------------
I have a new VAXcluster just set up in the last few weeks (two 6000-410s
running VMS V5.4-3). Every 20 minutes (almost exactly), we get an OPCOM message
as follows:
%%%%%%%%%%% OPCOM 18-MAR-1994 05:17:02.20 %%%%%%%%%%%
Message from user AUDIT$SERVER on DNV201
Security alarm (SECURITY) and security audit (SECURITY) on DNV201,
system id: 31945
Auditable event: Detached process login failure
Event time: 18-MAR-1994 05:17:02.19
PID: 20200ACC
Username: <login> Parent username: SYSTEM
Status: %RMS-E-DNR, device not ready, not mounted, or
unavailable
This is always happens from the same node and as I said, almost exactly
every 20 minutes. I've seen this message before on other systems but never this
regularly. There's nothing useful in accounting.
Both systems run Multinet if that might be relevant.
Any ideas?
(03/18/94 Briggs: CRON, DECscheduler or the like?)
---------------------------------------------------
For what it's worth -- We got identical symptoms when we moved a
DECscheduler database to a new machine and had jobs scheduled to run at regular
intervals on now-nonexistant disks.
Perhaps you have a similar process scheduler type product that is
attempting to run a job at 20 minute intervals?
(03/18/94 Stone: Nope)
-----------------------
Nope. And the times don't correspond with when a batch job is supposed
to run.
April, 1994 The DECUServe Journal Page: 3
(03/18/94 Spatz: Accounting may give more informtion)
------------------------------------------------------
You may want to invesitgate the corresponding Accounting entry for further
information.
(03/18/94 Maufe)
----------------
if some product on the system is waking every 20 minutes and doing
this, perhaps you can search all .LOG files for one of the errant PIDs?
Typically nicely-behaved products log the PIDs so you can where they went to
(when did a detached process ever work right first time?)
since you mentioned a networking package in .0, perhaps NETSERVER.LOGs
are being generated at the same time as the login failures?
(03/18/94 Stone: Relevant accounting info)
-------------------------------------------
LOGIN FAILURE
-------------
Username: <login> UIC: [SYSTEM]
Account: <login> ...
... Elapsed time: 0 00:00:00.24
...
Final status test: %RMS-E-DNR, device not ready, not mounted, or
unavailable
(03/18/94 Stone: They stopped but why?)
----------------------------------------
I did this. No hits (other than in OPERATOR.LOG of course).
These have just as mysteriously stopped (after the one at 15:17). I
cannot find any thing "unusual" between 15:17 and 15:37 that could account for
this. Stranger and stranger.
(03/18/94 Mezei: Date of last login ?)
---------------------------------------
Perhaps you could search the UAF for a last login date matching that of
the OPCOM message. You have have one username whose default device is a logical
which doesn't exist on that node fo the cluster (or a disk which is not
mounted/CLUSTER)
(03/18/94 Scopelliti: Some thoughts)
-------------------------------------
>> Perhaps you could search the UAF for a last login date matching that of
I doubt that would help.. the accounting entry indicates the login
hadn't completed (Username = <Login>).
What the entry is saying is something tried to create a process and
LOGINOUT.EXE aborted due to the aforementioned error.
April, 1994 The DECUServe Journal Page: 4
The question to ask is what files is loginout trying to open upon
starting? Let's see..
SYSUAF, RIGHTSLIST, some RTLs, etc.
You don't have the special login password filters enabled do you? (Not
PWDDIC & PWDHIS, but the hooks wherein you can add additional password
filtering)
Were you able to login normally on the node where these messages were
originating?
(03/18/94 Mezei: Recurring batch job)
--------------------------------------
Another thing to check is for recurringbatch jobs that execute at every
20 minutes. That job may result in the attempted creation of a process.
(03/19/94 Stone)
----------------
These all appear to be OK.
> You don't have the special login password filters enabled do you? (Not
> PWDDIC & PWDHIS, but the hooks wherein you can add additional password
> filtering)
Nope.
> Were you able to login normally on the node where these messages were
> originating?
Everything else works normally.
(03/19/94 Stone)
----------------
> Another thing to check is for recurringbatch jobs that execute at every
> 20 minutes. That job may result in the attempted creation of a process.
That's not it either.
In any event, right now, I can't get into that system. Someone screwed
up a router between here (San Francisco) and where that system is
(Indianapolis). Son any furhter checking will need to wait until the router is
fixed or until Tuesday when I'll be in Indianapolis.
(03/21/94 Crum: suggestions)
-----------------------------
If you haven't already, you might want to enable image accounting, and look
for any pattern before and after these events (if they occur again). A
possibility to consider is that someone may be trying to get into your system.
April, 1994 The DECUServe Journal Page: 5
(03/21/94 Stone: I think it's a non-issue now)
-----------------------------------------------
I think I sort of know what happenned. I believe something was brought
up in the wrong order resulting in an undefined logical name in the view of one
program.
(03/21/94 Prigot: Multiware, too?)
-----------------------------------
I know that you've said that you run Multinet. By any chance have you
recently loaded TGV's Multiware (Netware Services package)? If so, there is a
bug in the code where it goes looking for (other) Netware servers. Contact TGV
tech support for details. (The fix is a downloaded patch for the NFS code.)
(03/21/94 Stone: No Multiware)
-------------------------------
No multiware.
I think it was a sequence issue in how things were brought up concerning
one of our applications and an undefined logical name. Since this is a new
system, a lot of stuff wasn't in SYSTARTUP yet.
(03/22/94 Gaarder: Look at the default devices in SYSUAF.)
-----------------------------------------------------------
I had a similar problem where the default device wasn't changed when the
drive was changed. It doesn't have to be a logical that's hanging you up. Try
doing an AUTHORIZE LIST and scanning the devices.
(03/22/94 Hyde: another possibility)
-------------------------------------
There is at least one known bug in LOGINOUT.EXE that can cause
this problem also. It is apparently rare and is supposedly fixed in V6.0
(03/23/94 Gleason: periodicity)
--------------------------------
If something is occurring with extreme temporal regularity, a look see at
the TQEs can be very revealing.
(03/26/94 Thomas: I'm not sure what's the status here... )
-----------------------------------------------------------
> I think it was a sequence issue in how things were brought up
> concerning one of our applications and an undefined logical name. Since
> this is a new system, a lot of stuff wasn't in SYSTARTUP yet.
Does this mean you solved the problem? Or is it still happening?
April, 1994 The DECUServe Journal Page: 6
(03/28/94 Stone: Not solved but no longer a problem)
-----------------------------------------------------
> Does this mean you solved the problem? Or is it still happening?
No, I have not solved the problem. But it is not still happenning. So I
guess it's best to say that the sequence of events that cause the problem have
not recurred (and I consider it unlikely that they will occur again).
Rechargeable batteries
----------------------
The following article is an extract of the DECUServe
Hardware_Help conference topic 1608. The discussion
occurred between January 15, 1994 and March 19, 1994.
By Tah Fuh Chiam, Jorma Tapola, Jean-Francois Mezei, Jeff Killeen,
Jamie Hanrahan, Billy Youdelman, Linwood Ferguson, Barton Bruce, Shawn Allin,
Rob Brown, Bill Mayhew, Gerardo Razumney, Dale Coy, Joe Crum, Richard Norman
(01/15/94 Chiam)
----------------
I have had some inconsistent results in using rechargeable battery.
I have rechargeable batteries for my notebook PC as well as my cellular
phone. From time to time, I only get very low usage out of the battery that has
been fully charged. For example, the battery for the cellular phone should last
about 13 to 15 hours, after fully charged. Sometimes, I only get 7 hours out of
it. I will let it run down completely and fully recharge it and it will usually
work fine again, but sometimes I have to do it two or three times before it will
work as normal again.
So, what is the best way of dealing with this? Is it bad to "overcharge" a
rechargeable battery?
(01/15/94 Tapola: Some advice - go figure)
-------------------------------------------
I've seen lots of different advice, but I think they all had in
common that overcharge is bad.
To the knowledge I have gathered overcharging ages the batteries faster
than correct charging times.
Then there are advice how empty you should use batteries, and how to
store them. Here are examples what I've read:
- best results will be if you only use batteries to 1/3 of full charge
- Batteries should be sometimes fully discharged
- Batteries can be stored for years (no mention of charge)
- Batteries should be charged at least one a month
April, 1994 The DECUServe Journal Page: 7
- Batteries should be left empty even for couple of days
Well go figure.
And I was talking about Ni-Cd Batteries.
(01/15/94 Mezei: More myths and lore about Ni Cads)
----------------------------------------------------
To add to the myths of NiCads, i was once told that the charging
philosophy depends on how the battery is assembled:
If it is one big NiCad (as in my Powerbook 170), then you should
discharge it TOTALLY once in a while.
If it is a battery made up of many smaller cells, then you should never
discharge it completely.
As far as overcharging, aren't most chargers equipped with
"no-overcharge" logic ????? Yeah, sure....
I usually leave my Powerbook plugged in all the time, unless I am in
"mobile" mode. Yes, the battery does get warm, but I figure that these batteries
can take it. (Besides, since they are predicting -38 for tonight, I think I
will make sure my powerbook batteries stay warm :-(
(01/15/94 Tapola: Kind of no-overcharge logic ;-( )
----------------------------------------------------
That reminded me about charge of a friend of mine.
He doesn't understand technology at all. His sons RC-cars Ni-Cd
batteries ran out of power, and he went to put on charges. I asked him, does he
have quick charges or regular (6 * AA cells). With regular 4 battery and 50 mA
current it would have taken 30 hours to charge them. He said:"I got one hour
charger". Then I go really interested, and asked to see this ~1.5 A charger! At
the top it said "autoselect 25-180 mA", and in bottom:
When test light glows dimly charge 1-1.5 h
When test light doesn't glow charge 2 h
At least you don't OVERcharge with that charger!!!
(01/15/94 Killeen: Don't do a full discharge)
----------------------------------------------
> If it is one big NiCad (as in my Powerbook 170), then you should
> discharge it TOTALLY once in a while.
NO! Everyone I have read on this subject who knows what they are talking
about points out total discharge can damage the cell. I believe it has
something to do with causing its polarity to reverse. The recommendation is to
discharge down to the level at which your equipment cuts off and charge to the
level at which your cells are at full power.
April, 1994 The DECUServe Journal Page: 8
(01/15/94 Chiam)
----------------
Hmm ... if memory serves me right, the instructions for characterizing
the battery for the Digital notebook PC is that you charge it fully, then turn
on "characterize battery" and then let it run down COMPLETELY and then recharge
it fully.
Now my next question is, how do you know that the battery has been
charged to the level at which the cells are at full power?
(01/15/94 Hanrahan)
-------------------
>NO! Everyone I have read on this subject who knows what they are talking about
>points out total discharge can damage the cell. I believe it has something to
>do with causing its polarity to reverse.
Correct. And it does not matter if it is "one big battery", as JF put it,
because unless your gadget runs on 1.2 volts, there is no such thing. NiCad
cells put out just 1.2 volts per, regardless of physical size.
(Size of an individual cell affects capacity (measured in amp-hours) and
internal resistance, hence maximum discharge rate. But the voltage is cast in
stone by the cell's chemistry. There is a term from Chem 101 that I can't
remember, but basically it's a measure, in volts, of the eagerness with which
an atom with a less-than-filled outer electron shell wants to either acquire
some electrons to fill the shell or lose some to empty it completely. For the
two atoms that are involved in NiCads, this sums to about 1.25 volts, and all
the atom pairs in a given cell are effectively in parallel, so it doesn't
matter how many of them there are -- voltages in parallel don't add, currents
do. But then, why does the cell's voltage drop as it is discharged, if the
voltage is fixed by the chemistry? Well, as the pairs of atoms react with each
other, there are fewer of them in parallel, hence higher internal resistance in
the battery, hence lower voltage *under load*. This is why you have to test
cells under load to find out their state of charge; the no-load voltage at the
cell's terminals doesn't tell you much, unless the cell is deeply discharged.)
So that "one big battery" has discrete cells inside it, connected in series.
If you discharge any series battery "completely", the cells inevitably will have
slightly different discharge curves. You can get to a point where the remaining
voltage in one cell is less than that of the rest of the pack combined. At that
point the low cell has current forced through it in the wrong direction, and
indeed starts to charge in reverse. The only way to deal with a battery in this
condition is to disassemble it and replace the bad cell, if possible. You can't
fix it from the outside. You may be able to "charge" the pack but the
once-reversed cell will never regain its former capacity.
hAs for the need for deep discharge, The Word from several battery
manufacturers' literature is:
1. The so-called "NiCad memory effect" is about 99% myth. The remaining 1%
arose from test situations where the battery was discharged to a particular
point, and then recharged, repeatedly, using very close to the same usage cycle
and the same discharge point each time. It is EXTREMELY unlikely to happen in
actual use in a consumer device.
April, 1994 The DECUServe Journal Page: 9
1a. Battery manufacturers have long since re-engineered their NiCads to
eliminate most of the remaining 1%.
2. The advice to "discharge the pack completely" is based on a desire to cure
the "memory". But since the "memory effect" is almost 100% myth, don't bother.
3. What DOES happen with NiCads in lots of consumer gear is that they are
overheated during charging. This causes a more or less permanent, moderate
drop in the pack's output voltage, even when it's at full charge. The charger
doesn't try to bring the pack up to it's expected full-charge voltage because
there is some expected variation in that voltage between packs; chargers have
to rely on sensing when the pack voltage stops rising, or when the charge
current drops off. So the pack comes out of the charger with a full charge in
terms of amp-hours, but its voltage will be a bit on the low side for the
entire discharge curve. This in turn causes the voltage-sensitive cutoff
circuit in the equipment being powered to think that the battery has less
charge in it than it does, and shut down before the pack's usable capacity is
really exhausted. It mimics the memory effect, but it isn't the same.
> Hmm ... if memory serves me right, the instructions for characterizing
> the battery for the Digital notebook PC is that you charge it fully,
> then turn on "characterize battery" and then let it run down COMPLETELY
> and then recharge it fully.
this doesn't really run it down completely, only as far as the voltage
sensor in the PC allows. Presumably on the "characterize" setting this allows
it to go a little lower than the normal cutoff point.
> Now my next question is, how do you know that the battery has been
> charged to the level at which the cells are at full power?
you don't. Only the charger knows, and it may be fooled by a "depressed"
battery.
The bottom line, imho, is that NiCads are really much too fussy for use in
consumer applications that need to use more than about 40% of the battery's
rated capacity. At least, not without fairly sophisticated charging circuits
(they need to sense the temperature of the battery being charged, for one
thing). Even a ten-hour rate trickle charger can easily overheat a NiCad pack
if left connected too long.
(01/15/94 Youdelman)
--------------------
I have a PROM programmer that stops charging after a 5 degree C rise in
battery temperature. This seems to work well, though it's only been a couple
years and probably too soon to tell.
(01/15/94 Chiam: Can I leave it fully charged and not use it?)
---------------------------------------------------------------
Ok, another question. If I have two battery pack for a notebook
computer, one basically for spare. Should I alternate the use of them or can I
April, 1994 The DECUServe Journal Page: 10
keep using one and keep the one spare (after it has been fully charged) for
emergency use? In other words, will it harm the battery if it is not used after
a full recharge?
(01/15/94 Ferguson: Some manufacturers apparently say discharge)
-----------------------------------------------------------------
I fly where a commuter does maintenance on metro airplanes. These have
batteries which are composed of numerous (like 20?) good-sized nicad cells
assembled into a stainless steel container about the size of an auto battery.
As part of their required maintenance, they bring these in and service
them periodically, and part of that involves a complete discharge. And I mean
complete -- after they are discharged on a special machine (which regulates the
speed of the discharge), they attach metal shorts on each elements' terminal and
leave them there overnight.
The mechanics simply say "because it says this is how we do it", but if
the "memory" aspects are a myth, then why would they do this? These are _very_
expensive batteries and equipment with very precise instructions from
manufacturers. That doesn't make them sensible of course, but it leans that
way.
(01/15/94 Hanrahan)
-------------------
Re alternating batteries - I don't know. My feeling is that you should
alternate them. They will self-discharge fairly rapidly anyway. ie if you
charge one and then leave it on the shelf, it won't have much of a charge left
after a month.
I do know that mfrs deliberately ship NiCads in a not-charged state. I
remember reading somewhere that leaving a charged Ni|Cad pack to self-discharge
on the shelf (or in shipping) isn't the best thing for them, but the source
wasn't authoritative.
(01/15/94 Hanrahan)
-------------------
> And I mean complete -- after they are discharged on a special machine
> (which regulates the speed of the discharge), they attach metal shorts
> on each elements' terminal and leave them there overnight.
Most likely the "special machine" brings them down to 1.1 V or so per cell,
then the metal shorts kill each cell individually.
> The mechanics simply say "because it says this is how we do it", but if
> the "memory" aspects are a myth, then why would they do this? These
> are _very_ expensive batteries and equipment with very precise
> instructions from manufacturers. That doesn't make them sensible of
> course, but it leans that way.
the reports of the "memory effect" got a lot of publicity when they were first
circulated -- but without the qualifications as to test conditions. Then lots
of people said "yes, that's what we've seen too" when they were really seeing
April, 1994 The DECUServe Journal Page: 11
voltage depression through overheating. (and it doesn't take much in the way
of excess temperature to do that, just a small overtemp over a period of days
will do it, and that's what a lot of people do to NiCads with simple chargers.)
Even some well-respected sources, like Horowitz and Hill's _The Art of
Electronics_ (probably the best book ever written on electronics for those for
whom electronics isn't their main job), reports the "memory effect" as
something everyone needs to worry about with NiCads.
Discharging the individual cells to zero won't hurt them, and it will help
prevent any "memory" -- but the instructions to do this were likely written
without knowledge of just how rare this effect really is.
(01/16/94 Bruce)
----------------
Some while ago (years even) there was a PA based (?) HAM selling
a custom NiCad recharger that did a super fast job of recharging and was
supposed to undo any memory effects. Supposedly a battery in pretty bad shape
could simply be plugged in and cleaned up and fully charged - FAST.
It was several HUNDRED dollars and needed a plugin thing to tailor it
for each battery size it was to be used for. The claim was that the charging was
done with very short high amperage pulses and then a rest (and maybe look at
what happened) period of hundreds of milliseconds.
I recall something about 'zapping' whiskers that may have grown and that
limit capacity.
I have seen trade press references over the last year or so to custom
recharger chips designed to automate fast recharging, and assumed they
implemented some or all of whatever was valid in that HAM's custom product.
There certainly ARE some expensive fast rechargers for the cordless tool
users that cost MORE than most of the tools, and are offered BY some of the
tool makers themselves. I assume they are based on the new chips.
(I should have circled some of those bingo cards when I saw the chips
first announced.)
(01/16/94 Brown: Battery conditioners?)
----------------------------------------
So the NiCad memory story is a myth eh?
Just like everybody else, I thought that this was what I was running
into the the batteries for (Sony) camcorders.
The Sony Store will sell me a battery conditioner ($50-$100 price
range)which is supposed to extend the life of the batteries and maybe even help
a battery that isn't as good as it once was. I know you can get battery
conditioners for other NICad batteries too.
What is a "battery conditioner" likely to do? Might it be of any
benefit to a NiCad user?
(01/16/94 Hanrahan)
-------------------
> I recall something about 'zapping' whiskers that may have grown and that
> limit capacity.
April, 1994 The DECUServe Journal Page: 12
yes - this is one of the things that happens to cells that are reverse
charged. (Either nickel or cadmium gets deposited on the wrong surface,
and grows these "whiskers" reaching for the opposite electrode.) A time-
honored way to zap these is to charge a hefty capacitor, say a few tenths
of a farad at about 10x to 20x the cell voltage, and discharge it into the
cell. (Not a multi-cell battery!) (Kids, don't try this at home...)
(01/16/94 Hanrahan)
-------------------
> So the NiCad memory story is a myth eh?
Look -- at some point these discussions devolve into quoting authorities,
anecdotes, and so on at each other. The sources for the info I posted are what
I consider authoritative. I have not done the experiments myself and I'm not
about to.
(but with Radio Shack's handy-dandy $130 multimeter with EIA232 output, and
some creative, low cost ways to get computer control of some medium-power
relays, it wouldn't be THAT difficult to set up; this would make a nifty
science fair project for some bright kid.)
> The Sony Store will sell me a battery conditioner ($50-$100 price
> range)which is supposed to extend the life of the batteries and maybe
> even help a battery that isn't as good as it once was. I know you can
> get battery conditioners for other NICad batteries too.
> What is a "battery conditioner" likely to do? Might it be of any
> benefit to a NiCad user?
I don't know. I won't say they're of no benefit. I will say that, based
on what I've read and experienced, elimination of the "memory effect" is
unlikely to be the mechanism for any benefit.
I will also note that Sony is no less willing than any other consumer
electronics giant to sell folks what they want to buy, regardless of whether
it's what they really need; and that Sony has designed and sold an
embarrassingly large number of turkeys in the last ten years or so; their
reputation for "pricey, but worth it" is not justified by anything like all of
their products, not even their high-priced products. (Ask me in the audio/video
equipment in HOBBIES about their current line of receivers...)
(01/16/94 Ferguson: How about lead/acid?)
------------------------------------------
While we're on the subject, how about lead/acid batteries? Whether car
or speciality.
I've always heard these should be kept at a high level of charge for
longest life, and deep cycles harm them. Basically the opposite of what you
commonly hear (and now I'm confused about) for nicads.
Is this truth or myth?
April, 1994 The DECUServe Journal Page: 13
(01/16/94 Allin: "Myth" is a myth?)
------------------------------------
Anecdote or not, experience at work tells me Ni-Cad memory is not rare.
We have a pool of laptops that have obviously changed over the years, but
started with Toshiba T1000/1200/1600's and Compaq SLT's that would be signed out
for employee's use. After 4 or 5 months, everyone one of them started showing
signs of "memory" in the guise of short battery life. And, in each case, a
total discharge (let laptop completely die, followed by shorting out battery
overnight) would bring them back to health. The same holds for my AA's in my
Walkman at home.
(01/16/94 Mayhew: OK, tell me what to do :-) )
-----------------------------------------------
So let's get down to brass tacks, here. :-)
I have a laptop that I bought about 3 years ago with 3 battery packs.
All of them are essentially dead (except I've found that I can jury-rig, in an
emergency, two of them in *series* and get enough oomph to run the computer with
the AC adapter plugged in). (The PC is a Toshiba T1000SE and the AC
adapter/charger doesn't provide the right juice to run the computer unless the
batteries are at "some" level of health. I don't know how/why this works...)
Recently I went out and bought a new battery pack so I could use the
thing sanely. (Ignoring the fact that it is, after all, a PC, so sanity doesn't
*really* apply.)
I use it irregularly and infrequently. Should I:
a. Leave it unplugged, and with the battery in it, needing to recharge
it for several hours each time I want to use it;
b. Plug it in to the wall, with the battery in it, presumably leaving
it in constant-charging state;
c. Plug it in to the wall, with the battery in it, and turn it on.
I don't think there are any other useful options... and what I want, of
course, is whatever will be the least hassle while preserving the battery life.
(I'm guessing that if I were to open up the battery pack, I might be
able to cannibalize enough "good" cells out of the 3 dead packs to produce 1 or
2 good ones, too. But it's not amenable to easy opening.) c.
(01/16/94 Hanrahan)
-------------------
> Anecdote or not, experience at work tells me Ni-Cad memory is not rare.
Sounds to me like you're describing voltage depression. Remember, you can't
see the mmeory effect unless you repeatedly discharge to the exact same point.
This is extremely unlikely in any real-world situation.
> And, in each case, a total discharge (let laptop completely die,
> followed by shorting out battery overnight) would bring them back to
> health.
This does cure voltage depression, IF you don't end up reverse-charging a
April, 1994 The DECUServe Journal Page: 14
cell. and the risk of that is fairly high.
(01/16/94 Hanrahan: Copied from Usenet... )
--------------------------------------------
X-News: cmkrnl sci.electronics:24119
From: ga...@ke4zv.uucp (Gary Coffman)
Subject:Re: Nicad Memory Effect-Fact or Myth?
Date: Sun, 21 Mar 1993 15:51:08 GMT
Message-ID:<1993Mar21.1...@ke4zv.uucp>
In article <1993Mar21....@ssc.com> t...@ssc.com (Tad Cook) writes:
>
>Do the performance of NiCad batteries suffer when they are repeatedly
>only slightly discharged? The story goes that one should do a deep
>discharge every time, before doing a complete charge. Otherwise
>after a number of shallow charge-discharge cycles the battery cannot
>do a deep discharge anymore.
>
>Some years ago I read a letter in the Technical Correspondence column
>in QST from a couple of engineers at Gould saying that this was a myth,
>and that the nicad memory effect could only be demonstrated in the
>lab when the battery is repeatedly discharged many times to precisely
>the same level.
>
>Last year QST magazine ran an article on nicads that mentioned the
>memory effect, and said that it was a myth, or at least vastly
>overstated.
The correct term is "voltage depression". It will occur only under
certain special conditions. The more important consideration is
how much *overcharge* you put on your batteries, and at what rate
and for how long. All batteries, NiCad or otherwise, require an
overcharge to reach full charge. Depending on battery construction
and battery chemistry, not all NiCads are created equal, different
amounts and rates of overcharge are required for optimum battery
life. Certain batteries designed for standby use can tolerate a
persistent moderate overcharge. Most HT and personal electronics
battery packs cannot.
If you routinely subject your HT battery to a full 14 hour charge
at the 10 hour rate when it only *requires* 2 hours of charge, then
you'll have a ruined battery in short order. Subjecting the typical
NiCad battery to a persistent charge at greater than the 100 hour
rate is asking for trouble. *Some* rapid chargers can sense onset
of full charge and switch off, but most don't. A rapid excessive
overcharge is almost certain to quickly ruin a battery.
Now on the other hand, you run the risk of *reversing* a cell
if you overly discharge a pack, so the dischargers you see on
the market are dangerous tools that must be used with extreme
care. If you suspect that a battery is suffering from voltage
April, 1994 The DECUServe Journal Page: 15
depression, you should *disassemble the pack* and discharge
each cell *individually*. You can't hurt the cells this way,
even if you take them to zero voltage. It's also true that NiCad
batteries will last for about 500 complete charge/discharge cycles
while they will last for a thousand or more partial charge/discharge
cycles. But *only* if the rules about proper overcharge procedures
are followed. This can be a real pain to do in practice.
What do I do? I use a pack until the HT complains of a low
battery, then I swap on a fresh pack and recharge the other
in a proper voltage droop sensing charger. This type of charger
is more reliable than the temperature sensing chargers. Regardless
of charge state, terminal voltage will droop when the cell reaches
full charge. With the thermal sensing chargers, you're already into
excessive overcharge before temperature rise shuts the charger down.
With timed chargers, you're likely to cook the battery, especially
if a momentary power line glitch resets the timer.
Gary
--
Gary Coffman KE4ZV | You make it, | gatech!wa4mei!ke4zv!gary
Destructive Testing Systems | we break it. | uunet!rsiatl!ke4zv!gary
534 Shannon Way | Guaranteed! | emory!kd4nc!ke4zv!gary
Lawrenceville, GA 30244 | |
X-News: cmkrnl sci.electronics:31110
From: rd...@cbnewsf.cb.att.com (richard.b.dell)
Subject:Re: Nicad memory effect: Any DEFINITIVE answer?
Date: 7 Jul 93 16:18:12 GMT
Message-ID:<C9syM...@cbfsb.cb.att.com>
OK, someone should do this, so here goes my first draft
attempt at a NICAD charging/discharging FAQ. Any additions or
comments via email please to cut down on bandwidth.
Most of this is verified or validated or inspired by what
might be considered the definitive work:
Rechargeable Batteries Application Handbook
by Technical Marketing Staff, of Gates Energy
Products, Inc.
Butterworth Heineman, 1992.
ISBN 0-7506-9227-0 (note -- correct ISBN,
incorrect one is listed inside book, correct
one on back cover, per Butterworth Heineman)
$49.95 from
Butterworth Heineman
80 Montvale Avenue
Stoneham, MA 02180-2422
1-617-438-8464 voice
1-617-438-1479 fax
April, 1994 The DECUServe Journal Page: 16
Other suggestions for sources of information welcome.
What is not supported by the above comes from various posts
here and other memories.
A. Overcharge Effect or Voltage depression (also incorrectly
refered to as memory effect).
1. The _real_ memory effect is caused by repeatedly
discharging to _exactly_ (or close to) the same point
in the discharge cycle, and has been eliminated.
2. "Cells exposed to overcharge for extended periods of
time, _particularly_ at elevated cell temperatures, may
also exhibit a phenomenon called voltage depression.
This results in the cell voltage being depressed
approximately 150 mV below the normally calulated
values..." (emphasis added)
3. "Voltage depression initially appears on the discharge
curve near the end of the discharge. With the
extension of the overcharge time (non-discharge) of the
cell, this depression progresses slowly towards the
mid-point and beyond."
4. "Voltage depression is an electrically reversible
condition and disappears when the cell is completely
discharged and recharged (sometimes refered to as
conditioning)."
5. "An end-of-discharge voltage on the _minimum_ cell of,
for example, 75 per cent of its MPV or less will
accomplish that restoration with each discharge."
(MPV -- mid-point voltage, where 50% of charge is
withdrawn, typically 1.20 volts)
6. This is, by my calculation, .90 volts.
B. Cell Reversal
1. Due to imbalance of the various cells in a series
connected battery, one cell may be driven into a
April, 1994 The DECUServe Journal Page: 17
reversed voltage before the other cells are completely
depleted of charge.
2. This effect _only_ occurs in batteries, not in single
cells, and is more likely to occur as the battery
becomes higher volatge, ie -- contains more cells.
3. Ignoring this effect causes (personal experience)
shorted cells, cells with short shelf life (one cell I
was trying to rejuvenate on the bench had a 3 hour
rather warm shelf life, but did seem to have full
capacity if discharged quickly), and cells which will
not take a charge.
4. Shorted cells _may_ be partially recovered by _zapping_
them with a high current attempting to _burn_ away the
short. I generally use a current limited supply of
about .5 amps, and set the voltage to about 20 volts,
charging a capacitor of around 10000 uF or so. The
cells rarely have much shelf life after being unshorted
though.
5. The book has a formula which leads to these minimum
recommended discharge voltages:
Num. of Nominal Minimum
Cells Voltage Voltage
1 1.20 -0.20 (0)
2 2.40 .85
4 4.80 2.95
5 6.00 4
8 9.60 7.15
10 12.00 9.25
12 14.40 11.35
6. "...the damaging effects of cell reversal depend on the
cumulative amount (amp-hours magnitude) of capacity in
reverse and each reversal may utilize only a fraction
of the total tolerance of the cell..."
7. Note that this is below .9 volts per cell until there
are ten cells (or a 12 volt nominal battery).
8. So, for any battery up to 12 volt nominal, a minimum
discharge voltage of .9 volts per cell should be
April, 1994 The DECUServe Journal Page: 18
adequate to recover from the voltage depression state
without unduly damaging the battery.
C. How to charge _properly_ and quickly.
1. Not all cells are _designed_ to be quick charged
(charged at a rate greater than .1 C where C is
capacity in amp-hours).
Type of Maximum Charge
cell recommended Control
rate
Standard .1 C NO
Quick .33 C NO
Fast 1-4 C YES
2. There are several types of charge control. Somewhat in
reverse order of desirability (ie--least stress to
battery) using a split rate constant current, they are:
a. inflection point cutoff -- the slope of the
voltage rise with time starts decreasing after the
initial rise. This method requires a backup
method.
b. rate-of-voltage change -- designed to detect the
rapidly rising voltage as the battery approaches
full charge. This method requires a backup
method.
c. voltage-decrement cutoff -- detects the voltage
drop as the battery passes into the overcharged
state. (typically equivalent in stress to a delta
Temp. cutoff of 5 - 10 degrees C)
d. incremental or delta Temperature cutoff -- fast
charge is terminated when a selected cell (with
temperature sensor) reaches a fixed increase from
the ambient temperature.
e. temperature cutoff -- fast charge is terminated
when a specific temperature is reached. If this
setpoint is automatically reset, a very
undesirable temperature and chargerate cycling
occurs, leaving average charge rate high, and
temperature high.
f. fixed time cutoff -- fast charge is terminated
after a set period of time. assumes that battery
April, 1994 The DECUServe Journal Page: 19
was completely used.
g. other more sophisticated and simple methods also
exist, but I suspect these cover most cases at
present.
h. backup methods may be any that come later on the
charging curve, with the simpliest being the fixed
time.
i. some very sophisticated pulse charging techniques
attempt to fast charge a battery which is not
designed to be fast charged. Reports on these
charges seem generally good. Some of these
chargers discharge the batteries first, removing
any voltage depression.
3. Take notice that the temperature cutoff and fixed time
cutoff are particularly susceptable to over charge
possibility, particularly temperature cutoff with an
automatic reset, which is very cost effective for the
manufacturer. If the battery stays very warm after
being in the charger for long enough to be on standby,
then take it off of charge.
4. A safe trickle charge to _maintain_ the charge on a
fully charged battery is from .02 to .1 times C.
D. Opinions and Conclusions
1. I personally would not trust any nicad battery on long
term standby charge at greater than .1 C, and would be
happier with .05 C if the cells are not Quick Charge
types or better. Warmth is my primary indication of
comfort, if the battery stays too warm, I am
uncomfortable.
2. If the battery is discharged to .75 MPV, it seems that
all types can mostly be safely be quick charged at
least, as long as the charge is terminated when the
cells reach full charge.
3. Temperature seems to be the enemy here.
--
Richard Dell
X-News: cmkrnl sci.electronics:31342
From: my...@hpfcso.FC.HP.COM (Bob Myers)
Subject:Re: Nicad memory effect: Any DEFINITIVE answer?
Date: Fri, 9 Jul 1993 19:18:29 GMT
Message-ID:<748...@hpfcso.FC.HP.COM>
April, 1994 The DECUServe Journal Page: 20
re: Conditions required for NiCd memory.
OK.
From General Electric's tech. note regarding memory (posted w/o permission,
and any errors are my own):
"Among the many users of batteries in both the industrial and consumer
sectors, the idea of a memory phenomenon in nickel-cadmium batteries has
been widely misused and understood. The term 'memory' has become a catch-all
'buzzword' that is used to describe a raft of application problems, being most
often confused with simple voltage depression.
To the well informed, however, 'memory' is a term applied to a specific
phenomenon encountered VERY INFREQUENTLY [emphasis mine - RLM] in field
applications. Specifically, the term 'memory' came from an aerospace
nickel-cadmium application in which the cells were repeatedly discharged to
25% of available capacity (plus or minus 1%) by exacting computer control,
then recharged to 100% capacity WITHOUT OVERCHARGE [emphasis in the original].
This long term, repetitive cycle regime, with no provisions for overcharge,
resulted in a loss of capacity beyond the 25% discharge point. Hence the
birth of a "memory" phenomenon, whereby nickel-cadmium batteries purportedly
lose capacity if repeatedly discharged to a specific level of capacity.
The 'memory' phenomenon observed in this original aerospace application
was eliminated by simply reprogramming the computer to allow for overcharging.
[Note that no mention is made of adding an intentional *discharge* to clear
the problem - RLM] In fact, 'memory' is always a completely reversible
condition; even in those rare cases where 'memory' cannot be avoided, it
can easily be erased. Unfortunately, the idea of memory-related loss of
capacity has been with us since. Realistically, however, ' memory' cannot
exist if any one of the following conditions holds:
A. Batteries achieve full overcharge.
B. Discharge is not exactly the same each cycle - plus or minus 2-3%
C. Discharge is to less than 1.0 volt per cell.
Remember, the existence of any ONE of these conditions eliminates the
possibility of 'memory'. GE has not verified true 'memory' in any field
application with the single exception of the satellite application noted
above. Lack of empirical evidence notwithstanding, 'memory' is still blamed
regularly for poor battery performance that is caused by a number of simple,
correctable application problems."
[End of quote from GE tech. note]
This note goes on to list the following as the most common causes of
application problems wrongly attributed to 'memory':
1. Cutoff voltage too high - basically, since NiCds have such a flat
voltage vs. discharge characteristic, using voltage sensing to determine
April, 1994 The DECUServe Journal Page: 21
when the battery is nearly empty can be tricky; an improper setting coupled
with a slight voltage depression can cause many products to call a battery
"dead" even when nearly the full capacity remains usable (albeit at a
slightly reduced voltage).
2. High temperature conditions - NiCds suffer under high-temp conditions; such
environments reduce both the charge that will be accepted by the cells
when charging, and the voltage across the battery when charged (and the
latter, of course, ties back into the above problem).
3. Voltage depression due to long-term overcharge - Self-explanatory. NiCds
can drop 0.1-0.15 V/cell if exposed to a long-term (i.e., a period of
months) overcharge. Such an overcharge is not unheard-of in consumer
gear, esp. if the user gets in the habit of leaving the unit in a charger
of simplistic design (but which was intended to provide enough current for
a relatively rapid charge). As a precaution, I do NOT leave any of my
NiCd gear on a charger longer than the recommended time UNLESS the charger
is specifically designed for long-term "trickle charging", and explictly
identified as such by the manufacturer.
4. There are a number of other possible causes listed in a "miscellaneous"
category; these include -
- Operation below 0 deg. C
- High discharge rates (above 5C) in a battery not specifically
designed for such use
- Inadequate charging time or a defective charger
- One or more defective or worn-out cells (NiCds DO have a
finite life; they won't keep charging and discharging
FOREVER no matter how well we baby them.)
To close with one more quote from the GE note:
"To recap, we can say that true 'memory' is exceedingly rare. When we see
poor battery performance attributed to 'memory', it is almost always certain
to be a correctable application problem. Of the...problems noted above,
Voltage Depression is the one most often mistaken for 'memory'.....
This information should dispel many of the myths that exaggerate the idea of
a 'memory' phenomenon."
And THAT is the last I'm gonna say on the subject for quite a while! :-)
Bob Myers KC0EW Hewlett-Packard Co. |Opinions expressed here are not
Systems Technology Div. |those of my employer or any other
my...@fc.hp.com Fort Collins, Colorado |sentient life-form on this planet.
(01/17/94 Brown: An apology)
-----------------------------
>Look -- at some point these discussions devolve into quoting authorities,
April, 1994 The DECUServe Journal Page: 22
>anecdotes, and so on at each other. The sources for the info I posted are what
>I consider authoritative. I have not done the experiments myself and I'm not
>about to.
Sorry.
I was not trying to be argumentative. I was not even disagreeing. It
was only a (very) poor attempt at humour.
Thank you for the information.
(01/18/94 Razumney: E.M.F.)
----------------------------
> stone by the cell's chemistry. There is a term from Chem 101 that I can't
> remember, but basically it's a measure, in volts, of the eagerness with which
> an atom with a less-than-filled outer electron shell wants to either acquire
> some electrons to fill the shell or lose some to empty it completely. For the
It's the electromotive force, usually referred to as "e m f".
(01/18/94 Razumney: Correction)
--------------------------------
Actually, I think you were referring to the 'oxidation potential'
of a 'half cell'. The EMF is the technical term for the theoretical voltage you
get when two such 'half cells' are put together to form a cell. The oxidation
potential is somewhat of a convention: since you can never measure the voltage
of a half cell, it is defined as the EMF of a cell composed of the half cell in
question vs. a standard half cell -- a hydrogen cell in a clearly defined
condition.
(01/19/94 Hanrahan)
-------------------
> Actually, I think you were referring to the 'oxidation potential'
> of a 'half cell'.
That's it!
(01/19/94 Hanrahan)
-------------------
> -< An apology >-
Well, I didn't take it as personally as I probably indicated.
It's just that a lot of discussions will deterioriate into people quoting
authorities and/or experience at each other, which when you think about it,
is largely a waste of time.
Also, I note that I stated that overcharge-induced voltage depression
isn't curable. I said this because I had understood that it was curable
only by discharging individual cells to zero (something that can't be done
from the outside of the typical "battery pack"). I've since seen a quote
from a battery mfr indicating that discharge to about 1 volt per cell should
April, 1994 The DECUServe Journal Page: 23
be enough to cure voltage depression, and that this is a safe discharge level
to go to in packs of up to six cells (which means packs of up to 7.2 volts).
Another interesting note: If you're fed up by laptop makers charging around
$100 for a lousy six-cell nicad battery pack, you may be able to make your own.
Digi-Key (1-800-DIGI-KEY) sells Panasonic NiCads in many variants (high
cpacity, quick charge, etc., etc,.) They aren't cheap but they're a heck of a
lot less than the $15/cell that some of the laptop makers are getting. If you
look at the table and measure your pack you may well find that your battery
pack is made of a number of these standard-sized cells. Do-it-yourselfers will
order the individual cells (solder tab versions) and make up packs of their
own. Others can turn to the next page in the catalog, where you'll find that
they can provide the cells in ready-to-go battery packs in many standard
configurations.
(02/07/94 Mezei: Warning, "empty" batteries aren't empty)
----------------------------------------------------------
I know that this topic is a bit quiet, but I figured I would share my
experience.
I figures I woudl try to completely discharge my Powerbook 170 battery.
So, I ran the laptop until it shutdown, then tried to boot it until it wouldn't
boot at all. Though the battery was dead ??? NO WAY !
Wow, did I get a surprise when I tried to short the battery !
I ended up connecting a 6 volt headlight (from my bike) on the
battery, and it stay completely lit for about 45 minutes before fading!
So, when you laptop says your battery is dead, it is far from being
dead and can still cause fires if shorted.
(02/07/94 Coy: I hope you didn't hurt yourself)
------------------------------------------------
WARNING - *NEVER* directly short any battery - and particularly a
NICAD.
(02/08/94 Crum: Direct short could sure burn your fingers...:-) )
------------------------------------------------------------------
I recently tried the same thing on a camcorder battery. It also had
lots of "juice" left - I did completely discharge it, but using some large
resistors. They were ceramic resistors, and they got kind of hot.
What's the danger, Dale - explosion?
(02/08/94 Coy: It's just not very nice... )
--------------------------------------------
The danger is, primarily, in what YOU will do when whatever you are
using to short out the battery
gets very hot
melts
etc...
You also could damage the battery, or have it catch fire.
April, 1994 The DECUServe Journal Page: 24
(02/08/94 Hanrahan)
-------------------
And even if you don't short it out... to repeat something that has been
said many times in this thread... you are NOT repeat NOT supposed to drain
NiCad batteries until they're dead. If you suspect voltage depression,
you might try draining it (with an external load, such as JF's motorcycle
headlamp) until the voltage under load drops an additional tenth of a volt
per cell from where it was when the camcorder or whatever declined to use
it any more... NO lower.
With his lamp, JF was in no danger of short-induced hazards, but he may have
drained the battery to the point where cell reversal is a possibility.
(02/13/94 Mezei: Second life after battery is dead ?)
------------------------------------------------------
While we are on the subject of rechargeable batteries...
I have noticd that when rechargeable batteries go dead, if you turn off
your device for a few moments, you can then fire it up and get a few seconds of
power again, and you can repeat this for quite a while.
Is there any harm in doing so ? What exactly happens to get some power
available again ?
(02/14/94 Coy: Chemical action)
--------------------------------
There is no particular harm in doing that. If you are going to
reverse-charge one or more cells in a multi-cell pack, then you have already
done that (and already damaged the pack).
The chemical reaction in the batteries will do a very small amount of
"recharging". [This is not something that continues - that is, you can't wait
"forever" and expect the battery to totally recharge]
(02/22/94 Norman)
-----------------
side note of possible interest:
According to Sunday paper a company in Atlanta has developed a
fast charging system for lead-acid batteries. It works by using a pulsing
technique which includes a brief discharge cycle which they claim helps to mix
the chemicals evenly. To prove it works they just set the record for most miles
traveled in 24 hours by an electric vehicle. Drawback: prototype charger costs
$50,000.
(02/22/94 Hanrahan: Not just prior art, *ancient* art.)
--------------------------------------------------------
Re .-1 -- this has been done for YEARS!!! It doesn't cost any $50K, either.
I sure hope someone at the Patent Office does more than rubber-stamp this
one.
April, 1994 The DECUServe Journal Page: 25
(02/28/94 Norman: Speed??)
---------------------------
Perhaps my synopsis was a little off. I was doing it from memory.
Ok check that, it was most likely way off ;-)
I think the primary point of the article was the "speed" with which
they could recharge was the break through. Perhaps a subtle twist on an old
technique?? They did list the __prototype__ as $50,000. Misprint? Guess I'll
have to fish it outta the bird cage ;-)
The DECUServe Journal
=====================
Publication Information
-----------------------
Topic threads in the DECUServe VAXNotes conferences are selected for
publication on the basis of strong technical content and/or interest to a wide
audience. They are submitted to the editor by a team of Contributing Editors
who are DECUServe Moderators, Executive Committee members and other volunteers.
Articles in the DECUServe Journal are downloaded to a 286 PC and formatted
using a standard text editor.
Editorial Content Disclaimer
----------------------------
Opinions and information presented here belong to each individual author.
No inferences should be made that the authors are expressing the policies of
their employers unless specifically stated. Content has been deemed acceptable
by the Moderators of DECUServe according to the DECUServe Canons of Conduct.
How can I use DECUServe?
------------------------
DECUServe is available 24 hours a day, 7 days a week, except for the last
Thursday of each month at 5pm Eastern time for backups. Membership is by
individual subscription only, and subscriptions are sold on a yearly basis
currently for $75.00. Subscription forms are always included in DECUS New
Member packets and Symposia registration packets. Six-month subscription
forms are often distributed at symposia as well.
Foreign DECUS chapter members are invited to subscribe as well, although
they must do so through their chapter office. Forms and price details can be
obtained from the chapter office.
Also subscription forms can be downloaded directly from DECUServe by dialing
1-800-521-8950 with the username INFORMATION. Set your modem to 2400 or 9600,
1 stop bit, 8 bits, no parity. Other access modes are Tymnet, PC pursuit,
and Internet.