[time-nuts] A real world project need for timing accuracy...

jimlux jimlux at earthlink.net
Tue Nov 2 14:02:19 UTC 2010


William H. Fite wrote:
> Exactly, jimlux, this is readily possible.
> 
> I find it interesting that no one has commented, one way or the other, on
> the uncontrollable environmental variables I mentioned.  Is this just about
> technology and not about validity?
>

That's sort of a separate issue.

In the case of imaging, there will be distortions, but they would be 
pretty uniform over the (small) field of view.  With some registration 
marks on the target, you could bound the uncertainty.

The real question, I think, is getting two pieces of information back:
1) how long it took the bullet to get down range
2) where the bullet hit the target

Acoustic pickups are certainly a great way to do this, especially if 
they are inexpensive (because inevitably, they're going to get damaged)

And, as you recall, the OP was looking for cheap.

so, there's a couple system architectures possible here..
a) record the time of impact from the acoustic sensors, and telemeter 
that back to the shooting position, then do the data reduction there
b) do the position calculation at the target, and just send that back.
c) send the live impact data back, and do the timing at the shooter end.

I think a and b are a wash.. either way, you need some minimal "smarts" 
at the target end AND you need some way to relate "time" at the target 
to "time" at the shooting end.

For "impact position finding" you need time to microseconds (in air, 1 
foot/millisecond, and you want 0.01 foot resolution, at least), but it's 
relative to each other.  For the "time of flight" you don't need that 
precision.  It would be interesting to put a series of measuring 
stations along the bullet path, since you could measure the bullet's 
trajectory (and see how closely it matches the theoretical).

Let's think about (c).. say each sensor modulated a (different) several 
MHz tone, which you could then send back to the shooter end, receive, 
demodulate, and time.  It just, gut feel, seems like it would be easier 
to do the pulse detection and relative timing at the target end.. It 
really is just a single chip micro of some sort..

So the problem becomes one of measuring the time when the target got hit 
relative to when the bullet was shot... the micro is going to have some 
sort of clock, and it seems you could fairly easily calibrate it (at the 
shooting end) by measuring, say, the bit rate of the telemetry message.




More information about the time-nuts mailing list