[eagle] Re: CAN-Do-Too! ??????????
greencl at mindspring.com
Tue Jul 10 09:02:34 PDT 2007
I've collected a few topics from various peoples comments.
The CAN-Do! was designed to be mounted to the front wall of the module
box. Since the connector could do this with the greatest simplicity,
that's the way we went. Part of this had to do with minimizing the
module volume used and part with heat conduction for the module power
switch transistor. With the new mechanical configuration I have in
mind, that will not be sufficient. Mounting spacers to the module PCB
will likely be used. But this is all preliminary thinking and any
specific suggestions would be appreciated.
It has been suggested that a DE-9 would be sufficient. Technically,
this is correct. But it would require the elimination of five of the
six user pins. Maybe "require" is too strong a word. Note that there
are two CAN-H, two CAN-L, two power, and two ground (well, we ended up
with three grounds). Technically, only one of each is needed. But two
each were were provided to simplify the construction of the wiring
harness (no small issue). The six user pins (one of them is named EB)
were for special needs a module might have. Six may be overkill? The
naming of one "EB" is a legacy from P3D where the engineering beacon was
so pervasive throughout the satellite that it was decided to actually
name one of the user pins as such. Part of this is also driven by
P3E/P5A perceived requirements. P3E has used some of the user pins to
input structure thermistor signals to modules that do not otherwise use
all of their analog input pins on the 40 pin connector.
The sub-miniature D connector series has served us well. If anyone has
*experience* with something they think might be a better choice, we'd
love to hear about it.
The switching frequency of approximately 5 kHz observed is not the
natural frequency of the existing supply. It is likely the supply
running in an idle mode since the actual power requirement of the widget
is relatively small. To verify this, Juan might try putting an
additional load on it. This can be done at the 40 pin connector. In any
event, this points to the desirability for an experienced power supply
designer re-designing the power supply.
It has been assumed, and enforced by the widget design, that a high
power module would have a separate power input connector. If switching
such a module on/off by the IHU is desired, the widget switched power
would be used as a control line to switch the module power. The widget
is capable of supplying several amps to the module but the current limit
is specified at under an amp as a "normal" module can only dissipate a
limited amount of heat anyway. We also scaled the upper end of the
current measuring circuit at the specified current limit. And we added
a two amp fuse.
More information about the Eagle