From Amsat Bulletin Board:


Initial attempts today at recovering the S2 beacon on AO-40 have so far
not been successful. However, these will be repeated many more times
from multiple locations in the days ahead. We have a good "fix" on
ALON/ALAT and we are spinning at 3.5 RPM, so we have quite a few
days/weeks before the squints become problematic due to the mystery
effect. Currently, the mystery effect is working for us to improve solar
angle. Attempts will also be made in the next few days to bring the K-Tx
on line and search for its signal. The command team has a series of
simple machine code commands that, if heard, should activate
transmitters, switch receivers, switch batteries, etc. without having to
load the flight software into the IHU-1.

By way of review, the command team had been watching the main battery for
a number of months and noting that its behavior, while not alarming, was
not fully as expected. In particular, there were voltage fluctuations
with spin under relatively light power loads. When panels 1-6 received
sunlight their 1/2 complement of solar cells would often lead to a
voltage dip, which would pop back up when the fully arrayed panels were
in the sunlight. This suggested that perhaps one cell in the 20 cell
main battery might be weakening. On Sunday, we noted that the battery
was not maintaining its set voltage and this led to the safety software
shutting off the passbands. Some minor adjustments were made to the
solar array voltage offsets and this helped some, but we stayed in
passband off mode to study the matter further. It appears that at this
time we may have acquired a second bad or shorted cell, though we are
analyzing this further. During eclipse on orbit 1487, at approximately
1930 utc on 2004-01-26 the battery voltage hit the extreme low voltage
trigger. This caused safety resets of the battery and solar array set
points and the auxiliary battery was tied in to the main battery. This
all appears to have happened benignly, and when I got AOS at MA=8, about
an hour and a half later on the following orbit, I noted that the extreme
low voltage flag had been set ,as well as the other changes noted.
Voltages on both batteries at this time were similar to other post-Sunday
values, as were the charge currents. At this voltage level, the
auxiliary battery was considerably undercharged. We appeared to be in a
stable, though not fully nominal mode, and I was in the process of noting
the status of things to the command team to decide on the next course of
action when, at 00:39:59 utc the battery voltage suddenly dropped from
its post-Sunday level of 25.5 to 26.5 volts, down to 18 volts. This was
accompanied by an expected increase in charge current. The safety
software shut off the S2 Tx at the next MA change, and I brought it back
up manually twice, noting that the voltage was relatively stable at about
14 volts with heavy charge current. After letting things sit for a few
minutes, a third attempt to start the S2 Tx was not successful.

Our current best understanding is that we suffered a catastrophic failure
of the main battery which is clamping the buss voltage at a low level.
Accordingly, we have been concentrating our efforts on trying to connect
the auxiliary battery to the main buss and disconnect the main battery,
placing it on trickle charge for further analysis. If we can manage just
10 volts on the main buss to activate the IHU-1 we should be able to
accomplish this with the simple machine codes noted above, and the
voltages should come up to levels where the S2 will function.

If you examine the telemetry during this time, 10 minutes after the low-
voltage "event", the temperature of heat pipe 4 + X + Y takes a 34.2
deg.C apparent increase in temperature over a span of 14 seconds (1 block
interval), and continues to rapidly increase almost to the top of the
scale. Given the timing of this change, 10 minutes after the low-voltage
"event", its rapidity, the thermal inertia of the large, circumferential
heat pipes, etc... it is difficult to believe that this is real. Most
importantly, this heat pipe sensor sits in a "benign" area at the apex of
panels 1 and 6, near the high-gain end of the spacecraft. There are no
batteries or BCR's in this region, only the controller for the liquid
rocket motor (which was powered off!), and the L1 receiver, which shows
no accompanying change in temperature. All of the BCR's and batteries
are much closer to heat pipes 1 and 2 at the omni end of the spacecraft.
Therefore, this may represent a failure of the software/hardware due to
low voltage, rather than a real temperature increase. However, this will
be studied further. No other telemetry anomalies have been noted but we
continue to look.

--W4SM for the AO-40 Command Team