| 1 |
Jack Poulson wrote: |
| 2 |
> There flight tests proved the current system inefficient for all of the |
| 3 |
> planned computation and communication with the groundserver. They |
| 4 |
> claimed to be using a 416 Mhz xscale, which seems fairly beefy to me, |
| 5 |
> and I was under the impression that ramming floating point instructions |
| 6 |
> through a fixed-point register was grossly inefficient. |
| 7 |
> |
| 8 |
> However, I am very new to embedded systems, and your question |
| 9 |
> makes me think that the control algorithm itself is the culprit. |
| 10 |
|
| 11 |
In all likelihood you are emulating IEEE floating point in software, |
| 12 |
most likely because fixed point algorithms aren't all that well known |
| 13 |
outside of old fogeys like myself and people with a lust for squeezing |
| 14 |
the ultimate out of embedded hardware. |
| 15 |
|
| 16 |
If you're lucky, it's only 32-bit arithmetic. If you're *real* lucky, |
| 17 |
there are no logs, exponentials or trig functions in the code. If you're |
| 18 |
really really lucky, there aren't even any divides or square roots. :) |
| 19 |
|
| 20 |
By the way, it's been so long since I did any of this stuff that I don't |
| 21 |
know where one can find *current* references on the subject. But if |
| 22 |
you're willing to get your hands dirty in C and/or Forth code, there are |
| 23 |
some rewards waiting for you. You'd be amazed how much interesting stuff |
| 24 |
you can do with table lookup. |
| 25 |
-- |
| 26 |
gentoo-embedded@g.o mailing list |
Archives