Giles,
 
this e-mail is in html format and you should be able to see all the feature using outlook or saving it as an html page that could be directly attached to the acbar web page for your convenience and future reference. However, it should be possible to see it by text-mode ordinary mail client, also if some features will be disabled. Let me know if you like this way.
 
I performed 2 different tests:
 
1) system restarting
2) "martin" heater pc shutdown
 
I hope the pictures will help all of you to better understand the problem.
 
1) system restarting
as expected the system drive a lot of current at startup (Tambient = -67 F) as lot of heat is needed to reheat the system:
 
time       V*     I
hh:mm
00:00    61 V   70 A
00:20    61 V   47 A
01:00    61 V   10 A
*voltage directly measured on the chassis outlet with the fluke
 
After 1 hour the current stabilizes into a narrow range accordingly to the heat requirements of the various subsystems. This is a picture of power supply monitor at startup (not very exciting, just to let you understand what I'm talking about and for the record...):
 
Fig. 1
 
The awg 4 silicon cables (+ and gnd) connecting the chassis with the power supply overheat a little bit in this condition. The cable is an AWG 4, with two conductors for each polarity.
 
Fig.
 
I esteemed about 105-110 F (40-45 C) in this conditions. However, as you can see the cable is safe till 194 F (90 C).
 
The outlet distributing the power heats a little bit more. However, after tightening all the screws encircled in red, the overheating reduced. This confirm the suspect that there was a loose connection.
 
Fig. 3
The fan is pretty effective in improving the cooling of the area in "critical" situations.
 
It is puzzling why the cyan and white cables, probably awg 14-16 or close (I can't really tell), visible in next pictures
 
Fig. 4
 
are not overheat much worst. Maybe the insulation is a much worst thermal conductor.
Note the "insulation" I added (a sheet of rubber) to avoid possible contact between the PC case and the cable in case of further overheating of the shrinking tube. This has been, I think, the worst risk. To take the pictures I removed some more insulation from the cable ("shrinking tube [...]" that I added after the original shrinking tube peeled off after the overheating.
 
1) "Martin" heater PC shutdown
As soon as the heater PC is switched off (and not before) the system get apparently completely out of control.
The leds situation is as follows:
Fig. 5
not corresponding to the normal condition as all the green lights slowly switch off, while in standard conditions only a few of them switch off and on time by time. This means in my humble opinion that the electronics is still measuring the correct temperatures and that seea the various subsystem overheating. The red lights remain as they were. Note the cable that I removed from the watchdog connector on the left side of the picture as for Martin instructions.
 
However, this is the power supply condition at same time:
Fig. 6
 
The current readout is 80 A and apparently remains like that forever. In this condition, the silicon awg 4 cable get hotter, how much is difficult to say, however well into the allowed temperature range. I haven't kept the system in this conditions for longer than 10-15 minutes, but in the past I experienced temperature really hot, probably of the order of 140 F at the level of the outlet. After tightening the screws I feel like the overheating of the outlet is reduced, but the silicon cable is still warm. Again, the awg 14-16 cable are still apparently cold. Consider I couldn't touch the cable when I discovered the problem.
 
Note that the system restart to work correctly as soon as I switch on the "Martin" heater pc. I mean that the labview program, or even the operating system are not needed to keep the heater system working correctly.
 
For what concerns the Mac, it is off and disconnected from the rest of the heater system. The system is not affected in any way by powering on or off the mac.
 
In my opinion there are no big problems in running the instrument in this conditions. The overheating of the cable does not represent a fire hazard. I carefully inspected and checked the temperature of any cable not felling any hazardous condition. There is a potential hazard if "martin" heater pc is switched off by a failure, but the fan helps and I constantly check the situation on the back of the chassis. It would be interesting to see what happen after physically unpluggin all the connections between the chassis and the heater PC.
 
Further questions and suggestion welcome.
 
Paolo