Jim Stiles wrote:
> I am working on some fairly advanced rockets with a friend - we are
> currently using conventional high power rockets as test beds, but the goal
> is to get into much higher performance regimes (our wildest dreams have us
> building rotons). One of the threads of our efforts is to get some
> intelligence into the rockets. The normal way to do this would be to build
> custom PC boards for this, that, and the other thing. However, I am
> intrigued with the possibility of using a general purpose computer with A/D
> interfaces for sensors and controls.
>
> I know that launchable computers are not the primary focus of this group,
> but as a long-term lurker, it seems pretty clear to me that a lot of the
> same issues exist.
>
> The parameters for the space available is that it is as long as it needs to
> be, 6 inches in diameter, and the whole system will be subject to lots of
> G's (like in the low hundreds range, depending on motors, etc).
>
> Are there wearable systems that people have worked with that will stand up
> to this regime? I envision an OS that can run out of RAM (I would guess
> that a disk is unlikely to survive), and a PC board that will probably have
> to be potted to tolerate the forces. Compute power is presumably not a
> significant issue - I ran a TI micro system almost 20 years ago that had
> enough cycles to do this stuff (although the OS was BASIC, so presumably
> something real will require a little more horsepower).
>
> If anyone has suggestions (OS, processor, A/D interfaces, etc), I would
> love to hear them.
>
> Jim Stiles
I'll suggest a few things (Just went to a MicroChip PIC seminar, and
asked one gent there who'd done embedded computer technician work for
NASA earlier in his career, about how that worked!)
He commented that they tend to crimp the connectors going off the
boards (i.e. mechanically crimp the connectors firmly) as well as then
soldering them; This gives you good mechanical strength.
And said that nothing's wrong with "any way that works" for building
your circuitry - doesn't have to be PC boards even, just build it & pot
it in place (doesn't have to be crimped connectors -inside- the
potting.) (I believe Canada built their circuitry in 3-D in the early
space program days, and drove NASA nuts until they found that this
worked better at which time NASA quit being unhappy was what he said.)
If a MicroChip PIC ("Peripheral Interface Controller") IC would do
your job (or perhaps a Parallax Basic Stamp?) those are probably going
to be more robust than a laptop will be; The 17C line of PICs has chips
with 20 mA source and 35 mA sink per pin, each pin can be programmed as
an Input or an output dynamically, it does hardware multiply (in 1 CPU
cycle), 1 or 2 UART(s) on board (for telemetry?), up to 33 MHz clock
rate (basically all instructions run in 4 clocks) and up to 16k of
program memory, you can add in tons of EEProm data memory, hou have up
to 66 I/O ports (16 ADC 10-bit ports, 2 or 3 Pulse Width Modulation
outputs, etc.)
And PICs run off next to no battery power when idle, and little power
when running flat-out. You typically program the PIC in EPRom for the
17C model (and re-programming in-circuit is quite possible.)
There's a good mailing list for PICs (the "PicList"), the subscription
account to send "Help PicList" to for this is at
"L-Soft list server at MITVMA (1.8c)" <
>
The PicList, of course, has traffic resembling THIS list's traffic!
Worth it, though (if you want to just look up archives, ask me & I'll
point you the right way.)
I may well be wrong, but I am concerned about a laptop breaking under
G forces & vibration; if you pulled the momboard & potted it, maybe no
problem though. But - you don't have ANY analog inputs with a laptop.
You could, of course, mix & match, use a serial or parallel connection
between a PIC for the A/D interfacing, then a laptop for the
data-collection etc. duties.
Question: What is a Roton? (Haven't heard the term, out of touch with
the rocketry world though! Too many hobbies, too little time...)
Mark Willis, 
From Wear-Hard Mailing list Archive (WH)
Maintained by R. Paul McCarty
Archive created with babymail