portuesi@tweezers.esd.sgi.com (Michael Portuesi) (07/06/90)
From the toshiba BBS
**** SysOp Note ****
I have released this very nice and informative file on the T1000's NiCad
battery with reservation. The author talks about opening up the machine and
making an adjustment on the T1000. While opening the machine does not void the
warranty, any modifications or adjustments that cause damage will void the
warranty. However, I think that you users are wise enough (and deserve) to
learn from another's discovery. In the words of a famous sergeant "...lets be
careful out there..." applies to this situation.
T1000 Battery/Charger Notes
by
Jim Nichol
I have been looking into the battery and charging circuit on
the T1000. First the battery is a package of 4x 1.3 amp-hour nicad cells
for a voltage of 4.8 volts. With my T1000 with the 768k memory board I
measured a battery drain of .25 amp with just the computer running and .6
amp with the floppy drive in operation. I also ran the batteries down
until the computer stopped running which occurred at about 4.4 volts.
Plugging this data into the battery curves, the battery pack would run for
4.6 hours with the computer running and 1.8 hours with the computer and
floppy both running. The obvious point is we should minimize the use of the
floppy drive and use the RAM disk instead for maximum battery life.
These figures are still somewhat theoretical since I found my T1000 seemed to
have a shorter life than these figures if I stopped when the power LED changed
from green to red, also the computer would run for a long time after the
LED changed to red. I measured the battery voltage as the battery drained
down with occasional floppy disk drive operations and found that the
battery voltage started around 5.4 volts and the LED changed from green
to red at 4.8 volts which is a very conservative setting for this
threshold considering the computer does not actually stop until the
battery voltage drops to 4.4 volts. The LED green/red threshold may be
adjusted downward from the 4.8 volt factory setting by turning the
trimpot next to the battery counter clockwise while measuring the battery
voltage as the battery voltage crosses the 4.6 or even 4.5 volt level
during discharge. The LED will stay green all the way down to 4.6 volts
and will allow operation down to the point where 20 or 30 minutes of operation
are left before the computer stops. Actually I found that the computer would
probably run even further below 4.4 volts but the first floppy drive access
would be the stopper and this floppy access may be the one to save your files,
so it is best to have your data in RAM disk and just turn off the computer when
the power LED turns red.
It is possible to determine whether your T1000 requires the above
trimpot adjustment by charging up your batteries, running the T1000 until the
power LED turns red then running the T1000 until the computer stops. Calculate
the amount of time remaining after the LED turns red to see how much time you
actually have after the LED changes. If you find a considerable time remaining
you may adjust the trimpot counter clockwise one notch and rerun the above
procedure to see how much time after the LED turns red you have with the new
trimpot setting. Using this method is time consuming, but no test instruments
are required and you will also erase any battery memory by running the tests.
I tried an external battery for the T1000 and started with
the measurement of the T1000 power LED characteristics. I ran down the
T1000 internal battery until the LED changed to red then connected an
external power source to see what was needed to make the power LED happy
(after that is what I am here for). The LED can be returned to green
after the internal battery is depleted by connecting a voltage greater
than 8.2 volts DC to the T1000 power connector which means seven (or even
better eight) nicad cells in series. Using the "C" size cells commonly
available will give an additional amount of time equal to the T1000
internal cells, again based on the .25 amp drain for the computer alone
and .6 amp drain with floppy drive operation. Unfortunately using the "D"
cell size does not give a greater battery life since the battery makers
have taken the "C" cell and added filler material to make it the same
size as the "D" cell, buyer beware.
As for using a 12 volt source I guess one could place a 2.7 volt 3
watt zener diode in series with the 12 volt to drop the voltage down to the 9
volts required by the T1000. Since the external battery is charging the
internal battery when it is plugged in, the external battery should only be
plugged into the T1000 after the power LED has turned red indicating the
internal battery is low. This technique will allow the external battery to take
over after the internal battery is depleted so the external battery will be
used to run the computer rather than recharging the internal battery.
The question of overcharging the T1000 battery reminds me of
a problem I encountered with the rechargeable batteries on my HP-25
calculator which used to have a nicad battery failure about once a year. I
eventually got around to looking into the circuit and found that the
charging circuit was overcharging the batteries excessively when the
calculator was turned off. I lowered the charging current and have not
had the batteries fail since; but also this lowered the charging current
to the point where the batteries would discharge if the calculator was
left on overnight. But back to the T1000. The normal recharging
current for a 14 hour overnight recharge is .1 of the amp-hour rating of
the nicad cell, so the 1.3 amp hour T1000 battery should be charged at a
.13 amp rate. I measured the charging current of the T1000 with the Toshiba
power module and found that the battery was charged with a .1 amp current
with the computer turned on and .22 amp with the computer switched off.
So the batteries are being charged normally with the computer on, but
would be overcharged if the recharging took place over a long period of
time with the computer plugged into the charger and turned off. For the
record I measured the charging current with the computer off which resulted
from various charging voltages if you want to connect your own external
charger:
External Voltage Charging Current
6 Volts .03 Amp
7 .07
8 .11
9 .18
10.9 .22 <---Toshiba Charger(Computer
off)
Date: 07-10-88 (06:15) Number: 473
To: ALL Refer#: NONE
From: JIM NICHOL Read: (N/A)
Subj: T1000 BATTERIES Status: PUBLIC MESSAGE
With respect to my eagerly awaited Battery/Charger textfile I uploaded
it as T1000bat.txt which is presumably somewhere in Toshibaland being
reviewed by Jeff, sorry about the pun it looks like a batch file. Seeing
the questions raised on the battery life, I ran two more tests this
weekend on my T1000 with the RAM card installed. First I loaded many BBS
message files into my RAM drive along with my E88 editor and commenced
deleting lines and editing the files while running on the RAM disk with
NO floppy disk accessing. After a very boooooorrrring 4 hours and 15
minutes the battery LED turned red and after 4 hours and 45 minutes the
computer finally quit. The contents of the RAM disk were intact after I
plugged the unit back into the charger. I recharged overnight and got
out a box of new floppies which needed to be formatted and started
formatting continuously to see what the time would be with constant
floppy disk drive operation. After an even mooooorrrre boooorrrring 2
hours and 40 minutes the LED turned red and after 2 hours and 45 minutes
the computer stopped.
So the answer to battery life is somewhere in between the two
times depending on how much you use the floppy disk drive. And if you
really want long battery life load both your program and data files into
RAM disc and run on RAM disk the whole time. One approach is to use
small programs which can be stored in the RAM disk such as PC-Write
2.55(not the bigger 2.71),XTREE, DK109 keyboard macro, Laplink etc so
that everything can run on RAM disk including the data files. One final
thing I observed that when doing long tests of this type, as then end of
the battery life approached with the LED red the computer would finally
stop dead when a floppy disk drive access was started....and this could
be the last disk save operation you were trying to do to save your
results which would be lost. Saving to RAM disk would have no such
problem.
Well I think this discussion will continue.....
Till next time
Jim
--
__
\/ Michael Portuesi Silicon Graphics, Inc. portuesi@sgi.com
"man, this is weak."