DIY Solar Scintillation Seeing Monitor with FC support

[RTJoe]

Hello

Here is my first prototype of a DIY Solar Scintillation Seeing Monitor (based on https://www.ecu.edu/cs-cas/physics/uplo ... final2.pdf). It's currently build on an Arduino prototype board to check basic functionality.

IMG_1205.jpg (167.41 KiB) Viewed 23501 times

IMG_1206.jpg (186.33 KiB) Viewed 23501 times

The software is also just a first "quick and dirty" solution to display measured values. But - and I think that is an interesting feature - it has already the capability to start FireCapture recordings automatically when a specified seeing trigger level is reached.

ScreenShot01.jpg (58.51 KiB) Viewed 23501 times

The blue line on the left chart is the seeing value, the red line a moving average. The right chart displays the current input value (intensity) of the sensor.

I have just started to test the output. At least the results seem to be not completely unreasonable and the seeing measured this morning at cooler temperatures was clearly better than yesterday evening above the heated roofs.

There is a long ToDo list:

• - rebuild on perfboard, housing, …

• - validate results (compare measures with image quality)

• - user friendly software (improved display, configurable trigger for FireCapture, log file support, …)

• - maybe external display for standalone operation (site evaluation)

Joachim

[marktownley]

Interesting! Let us know how it develops. I've moved this to the mods section as it will get more attention there.

[Valery]

Very interesting development! Thanks for letting us know about this.

One important note:

As the FC has "start" and "pause" triggers, it will be better to make a following option:


User put the movie full duration (in seconds) and the seeing monitor starts and pause
and then restarts again, pause again, restarts again the record within this time movie duration.
This will cut the unwanted recording during poor seeing.

Say, user put the seeing maximal value, say, 1,2". Movie time limit - 40sec. So, within 40sec
the FC will record all frames when the seeing was better than 1,2".


Valery

[PEterW]

Just found this one by random surfing.... I have an arduino so will look to make my own one. Maybe add a display rather than just uploading to the serial output.

Cheers

PEter

[RTJoe]

Thank you for your comments.

Valery:This will be a very useful feature. I will implement it soon.

Joachim

[robertp]

Hello Joachim,

great to see that someone else is also working on the diy version. I also started to build it beginning of June. I've chosen the same approach (small breadboard and prototyping shield) and finished the electronics, but I'm still missing the IR-sensor. I ordered it more than 7 weeks ago from http://www.tme.eu but it still is backordered.
May I ask where you got your sensor from?


Best regards and steady skies,

Robert

[RTJoe]

Hi Robert

It’s really difficult to get the sensor used in the original project. I have currently replaced it by a BPW34 and also replaced the 220 ohm feedback resistor by a 2K variable resistor to get about 1V output.

Meanwhile I have switched to a soldered board and an Arduino Nano:

SSSM04.JPG (147 KiB) Viewed 23278 times

Currently I’m working on a sensor with longer cable because I want to attach it to the solar scope.

Good to know some other people working on such projects. Please let us know about your progress and results.

Joachim

[robertp]

Hello Joachim,

thanks for the hint with the BPW34, I will also try this one until the VTP4085H arrives. I ordered 3 of the VTP4085H, so if they ever arrive and you are interested to get one, I can send you one (I also live in Germany).

Robert

[RTJoe]

Hello Robert,

thank you for this kind offer, it would be great to compare the results of both photodiodes!

Joachim

[RTJoe]

Hello!

The DIY SSSM is making progress and before readers with less interest in the evaluation of the measurements stop reading, I would like to start with a question regarding Firecapture:

I have some problems when trying to resume a paused recording with keyboard commands. My software controls Firecapture by sending „key pressed“ commands to the FC window. In the FC settings I can assign the commands to start, stop and pause the recording. Everything works fine except that I’m not able to resume a paused recording; neither with a second „pause“ command nor with a new „start“ command. Does anybody have an idea how to resume a paused recording using the keyboard (not the mouse)?

Back to the DIY SSSM: for evaluation I have written some software to record the output of my ASI camera and log precisely the frame number and the corresponding SSSM seeing value. Then I processed the recorded AVI with AVIStack2, which is able to export the image quality data to a CSV file (usually I use AutoStakkert, but I have not found a possibility to export the quality graph).

The correlation coefficient of seeing and image quality was a negative value in all tests. That’s great, because larger seeing values mean lower quality. Here is the graph of one test run (685 frames):



Please note that the left y-axis (seeing values) and the right y-axis (image quality calculated by AVIStack2) have opposite directions.

I must admit that the graph is the one with the best correlation, about -0.5. The average correlation of my test runs was about -0.35. I guess that for a given seeing there is still a wide range of image quality possible. Therefore it might be not useful to start and stop recordings immediately based on the seeing values, but to wait for a longer period within some range.

Has anyone already tried to compare the results of the commercial SSM with the image quality? I would appreciate your reports and how you use the measured values for imaging.

BTW: During the SSSM test with an ASI120MM, I accidentally captured this:



Meanwhile a display is connected for usage without a laptop. Now only some housing is missing.



Joachim

[Valery]

Hi Joe,

Very interesting progress. Keep coming.

Hope you will find the best solutions and your SSM also will be available for us, solar nuts.


Thanks for sharing about your works progress.


Valery

[PDB]

Hello Joachim,

also started to build one, but have a question on whicht type of cable you used to connect the sensor. In the original doc a coax cable is recommended, but looks like you have used someting different. Suppose it is also a shielded cable?

Rgrds,

Paul

[RTJoe]

Hello Paul,

yes, I have used a shielded cable. The part you have seen in the pictures above was just a small piece from the board to an RCA socket at the housing.

I have just build another device and you can see the cable on the image:

SSSMon device

IMG_2127.JPG (592.68 KiB) Viewed 22234 times

From the RCA socket I use a standard video cable to attach the sensor.

BTW:
I'm currently working on a major update of the FireCapture plugin software. Solarchat user rxdeath has tested the AiryLab SSM support and it works fine. The next version will also include a live plot of the seeing measurements inside FireCapture, support for firmware with more samples per measurement and a log file with seeing values for each captured frame. Stay tuned...

Wish you success with your seeing monitor project!

Joachim

[PDB]

Hello Joachim,

making some progress, but still waiting for the LMC6484 to arrive. (Hopefully on monday). Meanwhile did some testing with the firmware loaded in the Arduino just putting some voltage (3,3V and 5V) to the analog pins. There is a small glitch in your modified sketch versus the original schema.

The schema connects to analog pins A0 and A1 (also reading them in the sketch). Your firmware is reading pin A1 and A2, so I wondered long time why the output in the SSSM program was not what I expected. Not a big deal, just changing the analogRead did the trick. (Already soldered my board to connect to pin A0 and A1, so that was difficult to change).

Kind regards,

Paul

[RTJoe]

Hello Paul,

sorry for the confusion caused by using A1 and A2 instead of A0 and A1. I should add a comment into the sketch. My intention was to keep A0 free, because the LCD keypad shield 1602 uses A0.

Wish you a lot of sunshine for testing!

Best regards,
Joachim

[PDB]

Thanks Joachim,

looking forward for your updated FC plugins as well. This should be a great help.
Hopefully I can start checking if I soldered all wires and components correctly on the print. Not having electronics background, that sometimes goes wrong. (Having written software since the the 70's that is more my business).

Kind regards,

Paul

[RTJoe]

Hi Paul,

same with me, I'm also in the software business. Fortunately embedded software...

Some hints:
- It's really necessary to have full(!) sunlight for testing. I was very surprised of the very small change of voltage at pin 1 of LMC6484 when holding the "flashlight" led of a smartphone 1-2cm in front of the sensor.
- You can already check the voltage at pin 5 of LMC6484, should be 2.5V
- Without sunlight, you should measure about 2.5V at pin 1 of LMC6484 and the ADC of Arduino
- Use a variable resistor instead of the fixed 220 Ohm resistor to adjust input level
- If you have the possibility to use an oscilloscope, this would be vary useful in case of problems

Wish you a lot of success,
Joachim

[PDB]

Waiting for the sun, but did some measurements and testing. Discovered that I soldered 1 wire to the wrong place; after correcting that something started to work. Holding the smartphone led almost in contact with the BPW34 photodiode I was able to generate input values even higher than 1. Seeing values also showed some variation then. So looks like now it is just waiting for real life test.

Regards,

Paul

[PDB]

Hi,

with a bit of sun and passing clouds did some quick tests this morning. The board seems to be working but I get different results if I connect it to the Nano or the Arduino Mega. Results from the Mega seems more what I expect. The only difference I see between the 2 setups is the supply voltage. The Nano has a 0.3 voltage drop when power is given over USB. I measure 4.7 instead of 5. On the ATMega I measure 4.9V. Can that explain the difference?

SSSM with Nano (all seeing values with sun around 4)

SSSM.JPG (70.83 KiB) Viewed 22058 times

SSSM results with Mega (much more variation in seeing values)

SSSMMega.JPG (83.64 KiB) Viewed 22058 times

The test board (before the LMC6484 arrived ...) Still planning to make a smaller 1 but need to solve this issue first before I solder a Nano to the print.

20161110_143949_resized.jpg (183.31 KiB) Viewed 22058 times

Regards,

Paul

[RTJoe]

Hi Paul,

you are right, the results of the Nano version are strange while the results of the Mega version seem to be reasonable.

I also have discovered the lowered voltage for USB powered Nano, but this should not make a (big) difference because the measures are normalized to the measured intensity. You can use the more stable voltage by using external power of 7-12V on the Nano. But my devices are working with 4.7V USB power or external power without problems.

According to the input plot you had a lot of trouble with clouds. Maybe you should try again with clear sky (I know it's hard to wait...) and recheck the connections to the ADC inputs of the Nano.

One more remark: Is the big capacitor on your test board a standard electrolytic capacitor? Note that the schematic requires a non-polarized or bipolar one:

NPCapacitor.JPG (9.82 KiB) Viewed 22038 times

Best regards,
Joachim

[PDB]

Hi Joachim,

The capacitor is of NP type so that should be ok. When directly displaying the values from the A0 pin, I see they are slightly lower with the Nano. There might just be some interference on the circuit, but to find that out i will need to find a scope. Could be from my USB ports.
(Had already been thinking of connecting a 9V battery to the VIN pin and GND to power the Nano and circuit board.)

Kind Regards,

Paul

[GreatAttractor]

Great reading, guys! I'm thinking of building one myself.

Hello!

Then I processed the recorded AVI with AVIStack2, which is able to export the image quality data to a CSV file (usually I use AutoStakkert, but I have not found a possibility to export the quality graph).Joachim

Good point; I'll add it to Stackistry in the next release (coming soon).

[PDB]

some response to our thread on "Cloudynights" http://www.cloudynights.com/topic/55698 ... eshooting/

Paul

[solarchat]

please be VERY careful with any posts from “Apollo".

He is one of two members that have ever been removed from this forum because of the content of their posts…believe me it takes an EXTREME occurrence for me to do that.

[RTJoe]

Hello!

Then I processed the recorded AVI with AVIStack2, which is able to export the image quality data to a CSV file (usually I use AutoStakkert, but I have not found a possibility to export the quality graph).Joachim

Good point; I'll add it to Stackistry in the next release (coming soon).

Great idea! The next version of the plugin can write log files with seeing values for each captured frame. By comparing the seeing measures with the image quality we should be able to analyze possible improvements.

Looking froward to the next Stackistry release,
Joachim