I have a list of feature requests. Mostly to take the current box and
bring it up to more 'modern' PSAS standards.

None of this is to say that the LTC that people have been working on
over the last few years is bad! In fact it worked near perfect during
the last launch. But there are always improvements to be made. Since
we can get small embedded linux computers on a board for cheap now, a
lot of the hardware can be simplified.

What I would like to see:

 - Beagleboard Black -- they worked really, really well in the hot
desert this past summer.
 - Custom electronics as single PCB. Probably only 2 layer. Order-able
from OSH Park.
 - Hot Swap-able.  OSH Park will give us 3 boards. Build out 2. We
should have field backups for everything.
 - No Phidgets. It was worth a try at the time. But Beagleboard has easy GPIO.
 - Smaller box. COTS but we should probably be able to get away with
half the current size.
 - LiPo. The *whole* system should be LiPo. Currently we still carry
out ~15 kg of lead cells!
 - Better attachment. Just a hook and single hand tight bolt or
something very simple and light.
 - Better external connector design(?)
 - Readable text on external connectors and switches (laser cut high
contrast gravoply)
 - Thermal analysis/ventilation/heatsinks/active cooling/shading.
 - Use the same software infrastructure as the rocket (event driven
main code, command and telemetry extensions, TCP/IP interface)
 - Detect umbilical connection state!

External Interface:

 - Rocket umbilical
 - Away box
 - WiFi coax
 - Power + Ethernet spare (for subsystems we haven't thought of yet)
 - Debug Ethernet port
 - Camera triggers (GoPro, High Speed)

Things not to change:

 - Safety interlocks. We worked hard to figure this out right.
 - Solar charging. It makes sense.

That's all I can think of at the moment.


On Tue, Oct 7, 2014 at 8:57 PM, Paul Mullen <p...@nellump.net> wrote:
> The Launch Tower Computer III (LTC3) team has begun work on a design
> doc.  To get the ball rolling, I'd like to hear what people do and do
> not like about the current launch tower computer (LTC2).  Note that
> we're talking strictly about the hardware (the SBC and its
> peripherals).
> A very brief starter list (feel free to mangle):
> Good
> ----
> * Powerful (for our purposes) off-the-shelf SBC.
> * Most I/O travels over USB.  Software development
>   does not require an identical SBC, but rather can take place on any
>   system equipped with USB and a comparable Phidgets kit.
> Bad
> ---
> * Most I/O travels over USB.  Not-so-robust connectors; bulky cables.
> * Phidgets.  Expensive, fiddly, and require their own interface board
>   Their software stack is thick and troublesome to install.
> * So many cables and wires...  The insides of the enclosure are quite
>   busy.
> --
> Paul Mullen
> _______________________________________________
> psas-avionics mailing list
> psas-avionics@lists.psas.pdx.edu
> http://lists.psas.pdx.edu/mailman/listinfo/psas-avionics

psas-avionics mailing list

Reply via email to