Received: from hpoclrf.sj.hp.com by opus.hpl.hp.com with SMTP (1.37.109.8/15.5+ECS 3.3+HPL1.1) id AA04990; Thu, 29 Sep 1994 15:05:27 -0700 Return-Path: <kirk@hpmsd3.sj.hp.com-DeleteThis> Received: by hpmsd3.sj.hp.com (1.37.109.8/15.5+IOS 3.22) id AA13889; Thu, 29 Sep 1994 15:07:37 -0700 Date: Thu, 29 Sep 1994 15:07:37 -0700 From: Kirk Lindstrom <kirk@hpmsd3.sj.hp.com-DeleteThis> Message-Id: <9409292207.AA13889@hpmsd3.sj.hp.com-DeleteThis> To: poulton@zonker.hpl.hp.com-DeleteThis, wind_talk@opus.hpl.hp.com-DeleteThis Subject: Re: Re: Re: 3rd ave sensor design
KenP writes:
> The point is that even a defocused low-power laser is much brighter from
> a mile away (the distance at 3rd) than a white sunlit card. To get a
> good signal from a sunlit card, you need to view *only* on the part that
> gets obscured by the spinning cups. Otherwise, the background
> contributes such a high ambient light that you cannot get your signal
> out. (Hitting a six inch target at one mile means aiming to better
> than .01 degree. This sounds hard/expensive.)
>
Well, you use filters to remove all but the freq. of interest.
Noise over such a narrow, low freq band is nearly zero and I bet you
really only need nW of optical power....
Another idea.....why not tie the optput of the spinner to a legal Rf
source. Modulate the carrier by the spinn rate and detect it from
shore with a simple antenna. Build a simple FM detector and you then
get the speed off of that. Maybe a CB radio set to broadcast once
per hour the modulated signal....maybe 1Khz with 2Hz/MPH Modulation
.......Hmmmmm.......... In fact, I bet weather buoys do something like
this....
> With a laser, the apparent brightness at the wavelength of interest can
> be much brighter than the background, so the focusing/aiming requirement
> may be relaxed. system, you need a single photodetector device (no
> CCDs).
>
Still with the sexy laser.....labs guys.....never economical.....8-)
> My friend says we need to recieve around 1 uW. If we assume a 5 mW
> laser (pretty common), and assume a cheapy 3" telescope to receive with,
> we need 1 uW in 7 in^2. This says we can let our 5 mW spread out
> over 35000 in^2, or a 17 foot diameter. At 2 miles roundtrip, that
> allows a .1 degree spread (and aiming). He thinks that is tight, but
> not very hard to do.
>
Doing a quick calc......1nW should do for spinnrates up to 2KHz....
For 100Hz HPF, get....250pW optical power. If 4 cycles/MPH and wind
above 25 MPH, do we really need more accuracy at the cost of
sensitivity?
Assume SNR of 1 and sample the same data many times.....10pA/rtHz
HPF, 2pole LC at 1Hz.......... Last I looked DC is <<<< 0.0000001Hz.....
0.4A/W Si PIN.
Should be very easy to do.
Of course, this might have to be encased in a metal box, but the
bandwidth is so low that you really don't need much light. Way I
figure it, if you can see it, we can measure it...8-)
Now, how big a scope do you need to see a spinner 3.6 miles from shore?
Or, how big does that spinner have to be to see with binoculars? I can
seldom see the far channel marker.......
> By comparison, a sunlit card 6" on a side might reflect 11 W, but
> diffusely. With 1/r^2 radiation, I think we get 98 nW/ft^2 at one mile.
> Then to collect 1 uW, we need a 3.6 foot diameter telescope (and .005
> degree aiming). Maybe there's one on Mt Hamilton we can use for this
> during the day. :-)
>
good points. again, we don't need to see IT, just get the beam close so
it comes back.
> Sorry for the english units, and there are plenty of losses I have
> omitted, but I think these are good order-of-magnitude numbers.
> I welcome further refinements by anyone with more experience.
>
> (Kirk, I kept this idea to myself for six months before deciding *I* was
> never going to have time to build this thing. If Windsight is willing
> to put up the $$ to build this it and the time to get it sited, and
> the $$ to run the thing, they deserve to be paid for it. From my view,
> it will benefit us all if we can get better wind data, even for money.
> We may as well see if we can sketch out the design for them. Otherwise,
> it will probably never happen.)
>
Sure I agree. I wonder how good HP's equipment would be if we were to
rely on "free advice" rather than hire experts? Probably not that good
and we'd be doing something else.
"Job enrichment has been around for sixty years. It's been successful
every time it has been tried, but industry is not interested."
- Peter Drucker
This archive was generated by hypermail 2b29 : Mon Dec 10 2001 - 02:28:13 PST