Vivante SE DC power
- Since my Vivante SE battery is down to about 2/3 to 1/2 the capacity it
once had, I've been thinking of getting an external battery pack or two to
extend my runtime, which would simply plug into the DC connection in back
where the power brick usually plugs in. The SE power brick is about 19V,
and you can find a 19.2V 3000 mAH NiMH battery pack with charger and
connectors from a place like batteryspace.com for about $37. However, since
batteries are not constant-voltage devices it is important for me to check
the voltage tolerance and behavior of the SE's external DC connection.
A 16-cell battery pack like the one above will start about 22.5V when fully
charged, operate near 19-20V for most of the cycle, and end up near 16V when
mostly discharged, so a similar tolerance at the Vivante's DC input is
important. I thought I'd pass the results along.
With a spare DC connector I powered my Vivante SE from a DC power supply in
our electronics lab, and slowly turned down the voltage from 20V, watching
the current along the way. While booting I observed a current draw of
0.7 to 1.3 Amps. For the test I started a CPU-intensive program to keep the
load high. Above 90% CPU the SE drew 1.2 Amp, compared to 0.7-0.9 Amp at less
than 5% CPU. I turned down the the voltage by 1 volt every few minutes.
I saw the SE switch to the internal battery at just below 16.5V. I turned
the voltage up and down around this point to check the hysteresis, and it
switched near the same voltage in each direction. Interestingly, the current
draw from the external power supply did not immediately switch off.
Instead it ramped down from 1.2 Amp at 16.5V to less than 0.1 Amp at 15V.
(The internal battery was not charging at any point during this test, and I
did not try the test without the internal battery installed)
This looks like a very good result. As expected for a well-engineered
device, the Vivante SE has a large input voltage tolerance, and it appears
that the input voltage range is very compatible with a 16-cell NiMH
external battery pack. When a NiMH cell is mostly exhausted under normal
load its voltage will droop below 1.1V. You don't want to discharge a NiMH
cell much below about 0.9V, or it can be damaged. The Vivante SE internal
battery kicks in at 16.5V, or 1.03V per cell for a balanced pack of 16 cells,
and this means that the NiMH pack will be mostly exhausted when the Vivante
switches to the internal battery. In addition, the switch is graceful, with
a slow decrease of the external current load down to near zero at 15V, or
0.94V per cell. This feature is important to prevent oscillation, since the
NiMH voltage will rise as the current load is reduced. The maximum current
draw I observed was about 1.3 Amp, which is about 0.43C discharge rate for a
3000 mAH pack. The minimum draw was averaging 0.8 Amp, or about 0.27C.
This is a moderately high, but quite acceptable discharge rate. Batteries
are usually tested at 0.2C, which means that the 3000 mAH rating was
determined at a somewhat lower load than I will be using. Considering that,
I should probably expect about 3.5 hours at idle, and maybe a bit over 2
hours at full load.
It looks like I will pick up one of these battery packs to get my
Vivante SE back above 4 hours battery runtime. (5-6 hours, I hope.)
If it doesn't work out for some strange reason I can always use the
battery pack for my new bike headlight project instead....