Dark Sky Sensor Working

[Rollo]

As mentioned, I got the Dark Sky Sensor ( for amateur astronomers ) working. It now accurately measures the period of a square pulse coming off a light sensor. Readings show that it is measuring the dark_noise of 0.66 Hz almost dead on, with varinaces of 0.02 Hz. Great! And it is doing this by measuring the period of only one pulse.

But I wanted to mention, in case other read these messages and want to do some light sensor work. Beware the kind of light you use! Daylight or DC light is quite stable. But it turns out that the variable readings I was getting were probably causes by the 60Hz flickering of my monitor!

Monitor flickers at 60Hz - too fast for the eye to see, but what about a light sensor that measures a pulse with at 15kHz?
I was getting readings varying by 100\% When this idea tweaked to me, I instead used a flashlight and the readings were now perfectly steady. Went back to the monitor -- readings varied.

So... the moral is... if you're using a sophisticated light sensor ( like the TSL230BR chip ), you probably need to watch out what kind of light source you're using to test it...

That problem had me scratching my head for days. Thanks to those who offered so much help in a variety of ways to try to solve the problem. Maybe the problem was a good thing after all, considering everything I got to learn about interrupts, pulsewidths, external clocks, etc. etc.

Cheers
Rol
___________________________
This message was ported from CCS's old forum
Original Post ID: 13298

[R.J.Hamlett]

:=As mentioned, I got the Dark Sky Sensor ( for amateur astronomers ) working. It now accurately measures the period of a square pulse coming off a light sensor. Readings show that it is measuring the dark_noise of 0.66 Hz almost dead on, with varinaces of 0.02 Hz. Great! And it is doing this by measuring the period of only one pulse.
:=
:=But I wanted to mention, in case other read these messages and want to do some light sensor work. Beware the kind of light you use! Daylight or DC light is quite stable. But it turns out that the variable readings I was getting were probably causes by the 60Hz flickering of my monitor!
:=
:=Monitor flickers at 60Hz - too fast for the eye to see, but what about a light sensor that measures a pulse with at 15kHz?
:=I was getting readings varying by 100\% When this idea tweaked to me, I instead used a flashlight and the readings were now perfectly steady. Went back to the monitor -- readings varied.
:=
:=So... the moral is... if you're using a sophisticated light sensor ( like the TSL230BR chip ), you probably need to watch out what kind of light source you're using to test it...
:=
:=That problem had me scratching my head for days. Thanks to those who offered so much help in a variety of ways to try to solve the problem. Maybe the problem was a good thing after all, considering everything I got to learn about interrupts, pulsewidths, external clocks, etc. etc.
:=
:=Cheers
Brilliant.
At least the readings are really good now.
One thing you could do, is use one of the averaging methods mentioned, and smooth the reading over a couple of seconds. For measuring the sky, this would not matter at all, and it would remove this problem (and things like the flash of a cars headlights), from causing a problem in the real unit.

Best Wishes
___________________________
This message was ported from CCS's old forum
Original Post ID: 13313

[Rollo]

:=Brilliant.

I guess that for work on a light sensor system, I had that pun coming. ;-)

:=One thing you could do, is use one of the averaging methods mentioned, and smooth the reading over a couple of seconds.

Good idea. A friend of mine, working on the Laptop-based version of this project, is already doing that and also recommended that I do the same thing. It's a good way of averaging out these variations.

Thanks,
Rol
___________________________
This message was ported from CCS's old forum
Original Post ID: 13324

[R.J.Hamlett]

:=
:=:=Brilliant.
:=
:=I guess that for work on a light sensor system, I had that pun coming. ;-)
:=
When you consider most of my code is dealing with various organic/inorganic pollutants in water, you can perhaps work out the versions I get...

:=:=One thing you could do, is use one of the averaging methods mentioned, and smooth the reading over a couple of seconds.
:=
:=Good idea. A friend of mine, working on the Laptop-based version of this project, is already doing that and also recommended that I do the same thing. It's a good way of averaging out these variations.
:=
:=Thanks,
:=Rol
What sort of other astronomical stuff are you interested in?. I have done a couple of projects for telescope accessories, such as a 'servo' focusser, which behaves like the RoboFocus unit, but gives much greater speed between positions (it gets to 120RPM, yet gives a final position to better than 1/10th degree). There is at least one other person on this group at times, who is an astronomer.

Best Wishes
___________________________
This message was ported from CCS's old forum
Original Post ID: 13346

[Rollo]

I'm part of a group of observers here in Ottawa, Ontario. We do much observing/imaging from darker sites. A couple of us are working on this dark sky sensor project in order to make:

"a unit that can give amateur observers user-independent quantifiable and repeatable readings of the relative darkness of observing sites."

In other words, this standalone Microchip Dark Sky Sensor unit that I've ported over with CCS would permit me to tell you exactly "how dark" it is at the TSP compared to one of our local dark sites, at specific points in time. In my case, I'll also be able to output values in visual magnitudes that match my eyes.

I'm curious to hear about your focusing device - do you have a site address where I could learn more, by any chance?

Cheers,
Rollo


:=:=
:=:=:=Brilliant.
:=:=
:=:=I guess that for work on a light sensor system, I had that pun coming. ;-)
:=:=
:=When you consider most of my code is dealing with various organic/inorganic pollutants in water, you can perhaps work out the versions I get...
:=
:=:=:=One thing you could do, is use one of the averaging methods mentioned, and smooth the reading over a couple of seconds.
:=:=
:=:=Good idea. A friend of mine, working on the Laptop-based version of this project, is already doing that and also recommended that I do the same thing. It's a good way of averaging out these variations.
:=:=
:=:=Thanks,
:=:=Rol
:=What sort of other astronomical stuff are you interested in?. I have done a couple of projects for telescope accessories, such as a 'servo' focusser, which behaves like the RoboFocus unit, but gives much greater speed between positions (it gets to 120RPM, yet gives a final position to better than 1/10th degree). There is at least one other person on this group at times, who is an astronomer.
:=
:=Best Wishes
___________________________
This message was ported from CCS's old forum
Original Post ID: 13356

[R.J.Hamlett]

:=I'm part of a group of observers here in Ottawa, Ontario. We do much observing/imaging from darker sites. A couple of us are working on this dark sky sensor project in order to make:
:=
:="a unit that can give amateur observers user-independent quantifiable and repeatable readings of the relative darkness of observing sites."
:=
:=In other words, this standalone Microchip Dark Sky Sensor unit that I've ported over with CSS would permit me to tell you exactly "how dark" it is at the TSP compared to one of our local dark sites, at specific points in time. In my case, I'll also be able to output values in visual magnitudes that match my eyes.
:=
:=I'm curious to hear about your focusing device - do you have a site address where I could learn more, by any chance?
:=
:=Cheers,
:=Rollo
:=
:=
:=:=:=
:=:=:=:=Brilliant.
:=:=:=
:=:=:=I guess that for work on a light sensor system, I had that pun coming. ;-)
:=:=:=
:=:=When you consider most of my code is dealing with various organic/inorganic pollutants in water, you can perhaps work out the versions I get...
:=:=
:=:=:=:=One thing you could do, is use one of the averaging methods mentioned, and smooth the reading over a couple of seconds.
:=:=:=
:=:=:=Good idea. A friend of mine, working on the Laptop-based version of this project, is already doing that and also recommended that I do the same thing. It's a good way of averaging out these variations.
:=:=:=
:=:=:=Thanks,
:=:=:=Rol
:=:=What sort of other astronomical stuff are you interested in?. I have done a couple of projects for telescope accessories, such as a 'servo' focusser, which behaves like the RoboFocus unit, but gives much greater speed between positions (it gets to 120RPM, yet gives a final position to better than 1/10th degree). There is at least one other person on this group at times, who is an astronomer.
:=:=
:=:=Best Wishes
I haven't put anything together about it yet. I'll have to generate a .pdf sometime.
It is a little board, about 1.5 * 2.5, holding a PIC18F252, a RS232 transceiver, temperature sensor, voltage regulator, and a HIP4020 driver (together with the support discretes, fuse etc.). It has a single header, connecting to a small Maxon servo motor, that is 5/8" in diameter, by about 2" long (including a magnetic shaft encoder, and a reduction gearbox in this length). The board is designed to accept 12v power, and the same serial commands as the RoboFocus unit, and emulates it's returns (except that RoboFocus sends a character for every 'step', but steps at only a (very) few hundred point/sec, but since I get to about 8000 points/second, when at full speed, I only send a 'guide' character to show the movement). The processor implements a standard servo algorithm, to move the shaft accurately to the requested position, and hold it there. As with the RoboFocus, it implements 'backlash' correction, allways approaching the requested point from one side. But unlike the RoboFocus, it does this even when asked to move manually (I have a couple of connections for input buttons).
I had to build it, following two problems with the RoboFocus system. The first was that it is not powerful enough to focus one scope I have, when the CCD camera is fitted, and the second is that is it so slow!...
If you want some details, email me directly (ttelmah@ntlworld.com), and I'll take a couple of pictures of the board, and send them to you with the board details.

Best Wishes
___________________________
This message was ported from CCS's old forum
Original Post ID: 13364