DIY Solar Scintillation Seeing Monitor with FC support

GreatAttractor · Post by **GreatAttractor** » Fri Jun 23, 2017 10:02 pm

GreatAttractor wrote: ↑Sat Jun 17, 2017 3:35 am and an LM224AN op-amp; parameters similar to LMC6484

...not really, after a closer look (which also explains the price difference, I guess). Fortunately I found a supplier and got a few LMC6484AIN op-amps.

So far I've assembled and tested the "first stage" (sampling only the averaged light intensity), seems to work fine (I'm using Arduino Micro):

: stage1.JPG (237.49 KiB) Viewed 16881 times

Indeed, I had to put the cell phone's LED-flashlight very close to the photodiode to see some smallish increase in the sampled value. I should have some breaks in the clouds tomorrow for proper testing.

The diode is really tiny, here it is with a pendrive and a chunky 3.3 µF film capacitor:

: dc.JPG (174.43 KiB) Viewed 16881 times

Beside the current (1-second) seeing value and the 10- or 30-second average, I'm also going to display an "average of the best percentage", which I think is more meaningful for us imagers. For example, with my current setup I usually shoot a 15-second video and then stack 30% best frames. The expected image quality should correspond with the average of seeing values of the best 5 seconds out of 15 s.

Another change is using the timer interrupt to calculate the average (and update display/send data to PC) over exactly 1-second intervals ‒ and not by polling the ADC until 10 000 samples are collected. As you can see in the first photo, there were just 7754 read-outs in the last second.

IanL · Post by **IanL** » Sat Jun 24, 2017 11:37 am

Inspired by this thread, I built an SSSM a few months back with good results. I've written a detailed build guide here:

http://www.blackwaterskies.co.uk/2017/0 ... itor-sssm/

- Full strip-board layout diagram included for the less electronically inclined.
- Step-by-step build and test instructions.
- Refactored the SSSM sketch to make it more readable and modular (to support different display types for example).
- Plenty of use of #define and #ifdef blocks so that only required code is included for the functions enabled, minimising sketch size.
(https://github.com/IanLauwerys/SSM/blob ... Sketch.ino)
- Optional OLED display with graph functionality (you can pick up suitable modules from eBay for a few pounds/dollars).

Hope this is useful to someone and thanks to Joachim for bringing this to the masses and the great FireCapture plugin.

The OLED display didn't come out well in this photo, but it does work in practice!

PDB · Post by **PDB** » Sat Jun 24, 2017 12:03 pm

Ian,

this is a fantastic guide you provided. Also the Arduino code looks very good. (Think I will test it out as soon weather permits)

Regards,

Paul

GreatAttractor · Post by **GreatAttractor** » Sat Jun 24, 2017 12:31 pm

Thank you, Ian. As I've already finished my sketch today, I guess we'll have yet another on Github

UPDATE:

I managed to test today during a short break in the clouds. It was a hot day, I put the rig in front of an open window and got consistently 10-12", which seems too high. Maybe temperature difference from the outside was bigger than I thought... Will try again tomorrow (though it's going to be partly cloudy again).

Sunday Update:

There were still some super-thin uneven clouds today and the SSM always clocked more than 10". I hope it's just the clouds, not some problem with the circuit. I hope it won't be the traditional two cloudy weeks after getting a new toy until I can test properly.

RTJoe · Post by **RTJoe** » Tue Jun 27, 2017 8:46 am

GreatAttractor wrote: ↑Sat Jun 17, 2017 3:35 am
I'm wondering about filtering; Seykora's article mentions using filters which gave 220 nm FWHM at 510 nm. If the SSM is used during an Hα imaging session, shouldn't we e.g. slap a red filter on the diode to give more meaningful seeing values? Would there be any noticeable difference anyway?

Some months ago I have asked exactly this question to Mr. Seykora, here is his reply:

"Schott filters with a 220 nm passband centered at 510nm. Nothing is to be gained by using the filters. These filters used in the original Scintillation Seeing Monitor paper were used to relate seeing measured with Scintillations and Angle of arrival fluctuations. Scintillations are very independent of wavelength in the Vis. and near IR wavelengths, angle of arrival fluctuations are dependent on wavelength."

Best regards,
Joachim

Spectral Joe · Post by **Spectral Joe** » Tue Jun 27, 2017 3:46 pm

With regard to high, thin clouds, these indicate high altitude wind shear that would cause poor seeing. Sounds like you are getting the right number. As for the filters, this combination would give a photopic response in combination with the photodiode spectral sensitivity. This type of detector / filter combination is commonly used in light measurement. As Joachim says, not needed in this application. I've built a slightly different version of this device, since I'm an old timer I prefer watching a needle move on a meter movement so I used digital pins 5 and 6 (on an Uno) for analog (PWM) outputs that represent seeing and level. A switch selects which one is displayed.

Joe

GreatAttractor · Post by **GreatAttractor** » Thu Jun 29, 2017 3:47 pm

Good news - it turned out I forgot to connect 2.5 V reference voltage to pin 5 of the op-amp, so the analog input 2 was returning some garbage values. After fixing this I'm getting the expected seeing levels (0.8"-2.5" at the moment, during breaks in passing high clouds).

I've started assembling the final "box" version.

GreatAttractor · Post by **GreatAttractor** » Sun Jul 09, 2017 5:22 pm

The box is almost finished; the circuit board didn't turn out much tidier than the test version...:

: ft2.jpg (140.12 KiB) Viewed 16503 times

The sensor assembly is made out of sawed-off RCA connector:

: cz1.jpg (77.04 KiB) Viewed 16503 times

: cz2.jpg (60.35 KiB) Viewed 16503 times

Field test:

: ft1.jpg (100.36 KiB) Viewed 16503 times

It's attached with a 1 meter long shielded RCA cable, to be mounted on the telescope. I'll make a second, directly pluggable sensor on a right-angle RCA plug for operation in the field (I'd like to take some readings from my favourite mountain in good weather, and maybe from a lake too; probably save them in the Arduino's EEPROM for later retrieval).

I've uploaded my Arduino code to: https://github.com/GreatAttractor/ssm
It uses timer interrupt (via TimerThree library) to sample over predetermined duration, and includes comments for the 4.46 "magic number".

PDB · Post by **PDB** » Mon Jul 10, 2017 7:42 am

Looks Good Filip.
Which Arduino are you using? I don't get your code compiled on the Nano. (It does for the ATMega) Some problem in the Timer file definitions.
Also tested Ian's code, but since use I2C versions of the LCD and OLED display had to do a few modifications in the display settings.
Works fine on the Mega, but displaying 2 graphs on the Nano is just too much and becomes unstable (not enough memory?) If I comment out one of the 2 graphs it works on the Nano as well. (Could see the Micro has slighly more memory, so that could help?)

Regards,
Paul

IanL · Post by **IanL** » Mon Jul 10, 2017 8:03 am

Paul,

Yes the code is pretty tight on memory. Another option worth trying might be to reduce the number of points stored and displayed for the graphs as they use a fair amount of memory.

GreatAttractor · Post by **GreatAttractor** » Mon Jul 10, 2017 8:30 am

Paul,

I have Arduino Micro. If TimerThree causes problems, one can comment out the whole timer thing and just use a fixed number of samples as in the original.

GreatAttractor · Post by **GreatAttractor** » Mon Jul 17, 2017 7:31 pm

The box is a hack job, but at least it's not in pieces anymore

The input-regulating variable resistor is accessible via the central hole. The knob is for LCD backlighting (the good thing about black-on-green LCD is that it's also perfectly legible in full sunlight with backlight off).

: bo2.jpg (146.88 KiB) Viewed 16348 times

: bo1.jpg (106.91 KiB) Viewed 16348 times

: bo3.jpg (71.2 KiB) Viewed 16348 times

I've also put together (github repository) a simple client (text-mode) to display and store SSM data in CSV format for later analysis, e.g. in LibreOffice Calc. Some data from today's golden evening (timestamps are in UTC, add 2 hours for my local time):

: ssm_log.png (299.61 KiB) Viewed 16348 times

Not bad, considering that the Sun was at barely 17° down to 8° altitude.

Update – histogram of seeing values:

: histogram.png (18.39 KiB) Viewed 16340 times

I'm already thinking of rebuilding into nicer and smaller version 2; already ordered one of those nice small OLED displays like Ian's, I'll probably have a PCB made too.

PDB · Post by **PDB** » Sun Jul 23, 2017 7:41 am

Hi,

first version of the Linux version of a real-time Solar Scintillation Monitor. (on Ubuntu 16.04) Still a long way to go ... Error handling and decent graphs (need to learn more about qcustomplot libraries, but first results are promising. It should work with different versions of the Arduino code (Original, RTJoe, Ian and GreatAttractor) Since there was no sunshine for real testing, I also build in a simulator device generating (erratic) random seeing levels and fixed 0.95 input value. (Thanks GreatAttractor for the coding example

)
Code is not in a "publishing" state yet, but hope to do that very soon ..

here is a screenshot

: Screenshot.png (44.14 KiB) Viewed 16182 times

Regards,

Paul

GreatAttractor · Post by **GreatAttractor** » Sun Aug 20, 2017 6:19 pm

As was to be expected, I haven't had imaging opportunities since finishing version 1 of my SSM

In the meantime, I bought a 1.3" mono OLED display and learned to use it; also, version 2 of the board is almost built, and new Arduino code is ready. Not that having a seeing graph is particularly useful, but it was fun to implement:

: oled.jpg (38.22 KiB) Viewed 15823 times

The thicker black line is the central running average (30 s). Using Polish style decimal separator for the screenshot, but it can be redefined to period. As I'm not attempting to keep a copy of frame buffer in SRAM (that would be 128x64 bits = 1 KiB), there's still plenty of memory left:

Code: Select all

AVR Memory Usage
----------------
Device: atmega32u4

Program:   13824 bytes (42.2% Full)
(.text + .data + .bootloader)

Data:        702 bytes (27.4% Full)
(.data + .bss + .noinit)

I think the code is cleanly split into components, so feel free to have a look/reuse: https://github.com/GreatAttractor/ssm/tree/dev/v2/src

GreatAttractor · Post by **GreatAttractor** » Sun Sep 03, 2017 5:05 pm

Finally! Two nice cloudless days, and I could at last verify the values reported by SSM. After attaching the sensor so it would follow the Sun:

: moco1.jpg (187.9 KiB) Viewed 15569 times

: moco2.jpg (216.09 KiB) Viewed 15569 times

: moco3.jpg (99.53 KiB) Viewed 15569 times

i started data collection, and captured a 50-second video at 20 fps. It would seem the SSM works correctly, because on a joint graph of the seeing and the (exported from Stackistry) frame quality, I got:

: seeing_vs_quality.png (193 KiB) Viewed 15569 times

Frame quality on the right-hand vertical axis (0 = best, 1 = worst). Correlation seems to be quite good.

Quality estimation

Stackistry uses a simple method of determining the quality of an image A; it's the sum of absolute differences of pixel values between A and a blurred version of A:

q = Σ |A − Ablur|

In other words, the total brightness of the “high frequency component”. If A is already blurred, then the differences, and therefore q, will be small. If A contains a lot of sharp details, they will remain after subtracting the blurred version, and the sum will be higher.

This quality measure is not affected by simple geometric distortion (“rubber membrane”) of images, which might explain imperfect correlation at some moments.

PDB · Post by **PDB** » Mon Dec 11, 2017 9:41 pm

Hi Rainer,

if it will run on the UNO depends on which Arduino code you have downloaded. There are several versions now and some are not compatible with all Arduinos. (Size of data and program for the graphics display, use of timer routines). The most simple version without display, should run on every Arduino.

Mine is running on an cheap Nano, but tested on and UNO and a Mega. If you use a display, then it depends which one you choose. I have been creating some code for a graphical version and small memory footprint but haven't published it yet.

Regards,
Paul

Sorry: can't answer your first question (too little real electronics knowledge)

rsfoto · Post by **rsfoto** » Tue Dec 12, 2017 1:04 am

PDB wrote: ↑Mon Dec 11, 2017 9:41 pm Hi Rainer,

if it will run on the UNO depends on which Arduino code you have downloaded. There are several versions now and some are not compatible with all Arduinos. (Size of data and program for the graphics display, use of timer routines). The most simple version without display, should run on every Arduino.

Mine is running on an cheap Nano, but tested on and UNO and a Mega. If you use a display, then it depends which one you choose. I have been creating some code for a graphical version and small memory footprint but haven't published it yet.

Regards,
Paul

Sorry: can't answer your first question (too little real electronics knowledge)

Hi Paul,

Thanks for the answer. Ordered also some LMC6484.

Will start with the original sketch in the paper of E.J.Seykora and then work upwards.

I just read a lot about all this and I see that we need 2 values and not just one wn the input is different as they camoe from 2 different pins of the LMC6484.

So I will set up the board as in the original paper.

BTW I found an interesting software package for displaying values on a PC. Easy to implement together with the Arduinos. It is called MegunoLink

regards Rainer

rsfoto · Post by **rsfoto** » Wed Dec 13, 2017 8:13 pm

Hi,

I am waiting to get the Photo diodes and the OP amps to start with the building.

What did draw a bit my attention is the follwong in the paper of Dr. Seykora

Care should be taken that the cable is grounded only at the circuit board and not at the detector.

A coax cable has only 2 lines for data transmission, the inner core and the outer mesh ...

So if I do not connect the outer mesh to the cathode of the BPW34 then I will get no signal :-)

: CB-COAX-Cable.jpg (95.67 KiB) Viewed 14990 times

GreatAttractor · Post by **GreatAttractor** » Thu Dec 14, 2017 8:59 pm

It seems clear to me: "not grounded at detector" = "don't try to attach the outer mesh at the detector to an external ground".

rsfoto · Post by **rsfoto** » Fri Dec 15, 2017 1:01 am

GreatAttractor wrote: ↑Thu Dec 14, 2017 8:59 pm It seems clear to me: "not grounded at detector" = "don't try to attach the outer mesh at the detector to an external ground".

Hi,

Thanks for that. Now if I understood correctly that menas that if I have a metal case for the detector then I should not ground it to that metal case ?

Again thanks

BTW. I got my BPW34 today and made some mesurements having connected it to an OpAmp. WOW this little buggers are really sensitive

I think I can use them for my ongoing project of the Solar guider I am building.

GreatAttractor · Post by **GreatAttractor** » Fri Dec 15, 2017 3:26 pm

Now if I understood correctly that menas that if I have a metal case for the detector then I should not ground it to that metal case ?

Right, that's my understanding too.

I think I can use them for my ongoing project of the Solar guider I am building.

That's cool! Also keep an eye on the Planetary Imager – automatic tracking (self-guiding) is in the works.

rsfoto · Post by **rsfoto** » Fri Dec 15, 2017 3:43 pm

GreatAttractor wrote: ↑Fri Dec 15, 2017 3:26 pm
Now if I understood correctly that menas that if I have a metal case for the detector then I should not ground it to that metal case ?
Right, that's my understanding too.

I think I can use them for my ongoing project of the Solar guider I am building.
That's cool! Also keep an eye on the Planetary Imager – automatic tracking (self-guiding) is in the works.

Hi,

Thanks. Will take a look at the Planetary Imager Software.

rsfoto · Post by **rsfoto** » Fri Dec 15, 2017 7:06 pm

Hi,

In regard to the Scintillation Monitor ...

Has anyone tried to diminish the angle of entrance ? So far I think everybody is using the full angle of the sensor is active.

Would it mean that using it recessed some values would heve to be changed ?

Looks like my LMC6484 will arrive today.

GreatAttractor · Post by **GreatAttractor** » Fri Dec 15, 2017 8:51 pm

As long as you don't use any optics in front of the naked diode, I think recessing won't make a difference. The Sun is thousands of times brighter than any other light source you'll encounter on a typical day, so no "background light shielding" should be needed.

rsfoto · Post by **rsfoto** » Sat Dec 16, 2017 1:49 am

GreatAttractor wrote: ↑Fri Dec 15, 2017 8:51 pm As long as you don't use any optics in front of the naked diode, I think recessing won't make a difference. The Sun is thousands of times brighter than any other light source you'll encounter on a typical day, so no "background light shielding" should be needed.

Hi,

Thanks for the input. Well the circuit is ready on the bread board and now I have to see how I do get the sensor into the sun. I guess I will prepare a long cable from my desk to the window

Also making up my mind how to eventually integrate this BPW34 photo diode into my solar guider project.

PDB · Post by **PDB** » Sat Dec 16, 2017 9:41 am

Hi,

I agree with GreatAttractor, it will probably make little difference recessing the sensor. Maybe you need to aim a little bit better for the Sun

(If you place a trimmer instead of the 220K resistor ove U1, you can adjust setting very easy)

I have added an Arduino sketch on my github ssMon pages (https://github.com/paul-db/ssMon )
Currently I am reworking the ssMon program, now only working on Linux, so that it will also run on Windows. There will also be an interface to Planetary Imager (already in the V.02.01 branch, but not officially released) and when I have time

a connection to the newly released Linux version of FireCapture.

With regards,

Paul

rsfoto · Post by **rsfoto** » Sat Dec 16, 2017 5:29 pm

Hi Paul,

Maybe you need to aim a little bit better for the Sun

I will try

I asked as I thought it would be better to have it a bit protected. As written I will try to get a better aiming of my telescope to the Sun

Will take a look at your Arduino sketch. Does it run on an Arduino UNO ?

Thanks

rsfoto · Post by **rsfoto** » Sat Dec 16, 2017 5:40 pm

rsfoto wrote: ↑Sat Dec 16, 2017 5:29 pm Hi Paul,

Maybe you need to aim a little bit better for the Sun
I will try I asked as I thought it would be better to have it a bit protected. As written I will try to get a better aiming of my telescope to the Sun

Will take a look at your Arduino sketch. Does it run on an Arduino UNO ?

Thanks

Hi Paul,

Can not open the tar.gz file as I have Windows :-) Would you please make a zip file ? Thank you very much.

PDB · Post by **PDB** » Sat Dec 16, 2017 5:49 pm

Now in ZIP format as well.

P.

rsfoto · Post by **rsfoto** » Sat Dec 16, 2017 6:27 pm

PDB wrote: ↑Sat Dec 16, 2017 5:49 pm Now in ZIP format as well.

P.

Hi Paul,

Thanks a lot.

rsfoto · Post by **rsfoto** » Sat Dec 16, 2017 9:35 pm

Hi,

Well. Here it is and testing.

Does not look nice but seems to work. Did not get 20M Ohm resistor so I used 2x 10M Ohm and as well did not find a 3.3uF capacitor so I am used 3 1uF capacitors in Parallel.

I used a 500 Ohm Trimpot for adjusting the Sun intensity.

Some images

Set up sitting on a chairs back rest in order to get it close to my telescope.

: IMG_2712_C1.jpg (422.18 KiB) Viewed 14136 times

View of the front of the telescope with a white light solar film from Baader. Many years old

The black box is the sensor of my Solar guider and the second image is the control box of my Solar guider.

: IMG_2713_C1.jpg (454.5 KiB) Viewed 14136 times

: IMG_2719_C1.jpg (280.41 KiB) Viewed 14136 times

Screen shot of the Laptop screen and the written values

: IMG_2717_C1.jpg (345.18 KiB) Viewed 14136 times

rsfoto · Post by **rsfoto** » Thu Dec 21, 2017 8:15 pm

Hi,

Well I am in my Observatory, 420 km North of Mexico and Mr Murphy too

:seesaw

NO full sun shine, very high heavy clouds and can not calibrate it.

By the way, Does anybody have a graphic of how the values are when clouds are present ? Would be interesting for me to see how the Sun intensity is and the seeing value

Is it possible that with clouds the scintillation of the sun is less ?

This is my Sunshine at the moment

: IMG_2720_C1.jpg (390.99 KiB) Viewed 14086 times

GreatAttractor · Post by **GreatAttractor** » Fri Dec 22, 2017 9:19 am

Is it possible that with clouds the scintillation of the sun is less ?

I sometimes get high, thin clouds that don't interfere much with recording (judging visually by the raw video and by successful stacking). However, I'd think SSM output through the clouds is meaningless, as the brightness changes due to clouds will completely swamp and distort the scintillation caused by seeing.

rsfoto · Post by **rsfoto** » Fri Dec 22, 2017 5:05 pm

GreatAttractor wrote: ↑Fri Dec 22, 2017 9:19 am
Is it possible that with clouds the scintillation of the sun is less ?
I sometimes get high, thin clouds that don't interfere much with recording (judging visually by the raw video and by successful stacking). However, I'd think SSM output through the clouds is meaningless, as the brightness changes due to clouds will completely swamp and distort the scintillation caused by seeing.

Hi and thanks.

I guess the clouds act as a filter and dampen the scintillation of the sun. I continue chasing the sun and well nature is nature and hopefull some of these days I get a cloudless day

... distort the scintillation caused by seeing.

I think what you wrote there is a bit upside down. From what I know seeing is defined as quality of the sky and in this case the quality of the capability of seeing is the scintillation of the Sun eg. the less the Sun scintillates the better the seeing.

Seeing is a product of the scintillation of the Sun which is caused by air turbulence passing in front of the Light emitted by the Sun.

At least that is what I understand what seeing and scintillation is.

Happy festivities and regards

rsfoto · Post by **rsfoto** » Wed Dec 27, 2017 5:43 pm

Hi,

Well, weather improved and I was able to play around with my SSM.

Had a few glitches which I have to observe but nothing deadly

It is interesting to observe in real time the images coming from my camera and compare it to the recorded values of the SSM. Afterwards when one processes the images one can corroborate that the image quality or at least the analysis AviStack presented in the diagrams are much in concordance with the values of the SSM.

So far I have not seen any wave tendency of the scintillation where I have my Observatory. At the moment I do not see a practical use for it apart from beeing a nice addition so one does not get bored while doing time lapses which take hours trying to get a flare or see something interesting.

Below some images of my setup and telescopes

: IMG_2736_C1.jpg (938 KiB) Viewed 14004 times

@ Paul, is the sensor well placed ?

: IMG_2738_C1.jpg (1.04 MiB) Viewed 14004 times

: IMG_2740_C1_solarchat_a.jpg (263.12 KiB) Viewed 14004 times

: IMG_2742_C1.jpg (1.22 MiB) Viewed 14004 times

: IMG_2751_C1.jpg (1.5 MiB) Viewed 14004 times

Below 2 graphs of the session of yesterday over on hour

: RS-eeing_20171226_1h.JPG (203.45 KiB) Viewed 14004 times

: RS-eeing_20171226_1h_tendency.JPG (211.64 KiB) Viewed 14004 times

and here the one hour video taken at 1260 mm focal length of Sun Spot 2692. I hope the next one is around the corner so I can continue with the tests and keep learning.

https://vimeo.com/248812555

Thanks for looking

PDB · Post by **PDB** » Wed Dec 27, 2017 6:23 pm

Nice setup Rainer.
Seems you had decent seeing, which shows in the video as well.

Looking at the green graph some small clouds?
Maybe you also need to tune the intensity value. Sometimes it seems to go over 1 (if i read the graph correct)

Regards,

P.

rsfoto · Post by **rsfoto** » Wed Dec 27, 2017 6:38 pm

PDB wrote: ↑Wed Dec 27, 2017 6:23 pm Nice setup Rainer.
Seems you had decent seeing, which shows in the video as well.

Looking at the green graph some small clouds?
Maybe you also need to tune the intensity value. Sometimes it seems to go over 1 (if i read the graph correct)

Regards,

P.

Hi Paul,

Thank you. Yes and that is my problem. Today in the morning without having changed anything yesterday y started up everything and the RS-eeing Monitor was delivering values under 0.5 and even further down which was not logic for me as the sky is less cloudy then yesterday or could I be mistaken and the whole enchilada RS-eeing Monitor is working correct ?

Could the telephone cable be a problem ? I am using a 6 strand telephone cable and 3 of them are wired together so the resistance of the cable is more then low. Couls it be that is is an EMI = Eletromageetic Interference problem and should I go for a Coaxial cable ?

I just measured the 4m long cable and I have 2.8 Ohm on both triple strands.

The sensor is delivering about 2400uA ~ 2.4 mAmp at 0.5 Volt.

I think I will get a few meters of Coaxial cable and built a second sensor to see how it performs.

BTW now the Sun intensity is jumping around 1.0 and 0.7 ¿?

PDB · Post by **PDB** » Wed Dec 27, 2017 6:56 pm

I use a coaxial cable with a length of 2.5 meter. I do notice some strange values sometimes but there I suspect some bad soldering on the cable ends. (I don't have a fixed setup and have to take scope, computer, cables out every time, reconnect everything, which is not good for the soldered connectors, have to redo these)
The fluctuations can be fairly high between days (transparancy of the air) and between seasons (air-mass?) I try to keep the input level between 0.5 and 0.9 . Unfortunately almost no sun in the last month(s) for testing, and the quark back to the US for repairs ....

P.

rsfoto · Post by **rsfoto** » Wed Dec 27, 2017 7:00 pm

PDB wrote: ↑Wed Dec 27, 2017 6:56 pm I use a coaxial cable with a length of 2.5 meter. I do notice some strange values sometimes but there I suspect some bad soldering on the cable ends. (I don't have a fixed setup and have to take scope, computer, cables out every time, reconnect everything, which is not good for the soldered connectors, have to redo these)
The fluctuations can be fairly high between days (transparency of the air) and between seasons (air-mass?) I try to keep the input level between 0.5 and 0.9 . Unfortunately almost no sun in the last month(s) for testing, and the quark back to the US for repairs ....

P.

Hi Paul,

Thanks. Will experiment and observe in the next days. While I am writing this very high thin clouds are passing so no chance to do a calibration which would make sense.

I just took out the arduino to the cold room in order to avoid the EMI of my 3 PC Monitors I have here.

rsfoto · Post by **rsfoto** » Wed Dec 27, 2017 7:37 pm

Hi,

As usual Mr. Murphy is always there where he should not be

:seesaw

View to the South

: South view; IMG_2762_C1.jpg (1.13 MiB) Viewed 13992 times

and view to the North. As when you build a house they say to put the kitchen window to North as it is the most uniform light

: North view; IMG_2763_C1.jpg (755.49 KiB) Viewed 13992 times

rsfoto · Post by **rsfoto** » Fri Dec 29, 2017 10:15 pm

Hi,

I have been having fun with the SSM.

Below a graph of measuring during 1.5 hours in order to see if there is a periodic wave form of the seeing which would allow us to plan imaging but I do not see any

So a SSM is a nice addition but so far not a big help for planning ...

: RS-eeing_test_20171228_test_1.5h.JPG (181.52 KiB) Viewed 13966 times

GreatAttractor · Post by **GreatAttractor** » Sat Dec 30, 2017 10:15 am

I like your results! Good correlation with the quality graph, so the device is working fine.

So a SSM is a nice addition but so far not a big help for planning ...

It seems to me that (apart from being a fun project) it's mainly useful for big-aperture people to trigger video recording during short moments of good seeing.

rsfoto · Post by **rsfoto** » Sat Dec 30, 2017 4:14 pm

GreatAttractor wrote: ↑Sat Dec 30, 2017 10:15 am I like your results! Good correlation with the quality graph, so the device is working fine.

So a SSM is a nice addition but so far not a big help for planning ...
It seems to me that (apart from being a fun project) it's mainly useful for big-aperture people to trigger video recording during short moments of good seeing.

Hi,

Yes, could be. It is a question then of where to set the upper seeing limit as best quality measurement and a need for really fast cameras so it makes sense to trigger an acquisition.

I have been dealing with very high clouds all these days ...

I just finished building a second sensor and will test it.

: IMG_2765_C1.jpg (795.32 KiB) Viewed 13950 times

: IMG_2766_C1.jpg (700.75 KiB) Viewed 13950 times

rsfoto · Post by **rsfoto** » Sat Dec 30, 2017 4:48 pm

GreatAttractor wrote: ↑Sat Dec 30, 2017 10:15 am I like your results! Good correlation with the quality graph, so the device is working fine.

So a SSM is a nice addition but so far not a big help for planning ...
It seems to me that (apart from being a fun project) it's mainly useful for big-aperture people to trigger video recording during short moments of good seeing.

Hi,

The above post made me think about how to decide what are good seeing values for the set up one has.

I guess it all depends of focal length and resolution per pixel. Am I correct in assuming that the less resolution ( more arcseconds per pixel) requires a better seeing the high resolution (less arcseconds per pixel) or is it the way around ?

¿ has anyone any info on how to correlate this ?

The original program takes 10000 measurements for one second values. ¿ has anybody tried to synchronize the seeing value to the speed of the camera ?

eg. My camera takes at an image size of 1632 x 1216 pixels 12 frames per second in average. So this could mean that if I go down and take 833 measurements the SSM would give me the seeing for each frame in my movie, more or less.

Will make some tests with this and then compare the SSM seeing graph with the graph AviStack delivers me about pixel movement. There should be then similar graphs. I suppose

Well I think I have found some use for the SSM.

The next step is to add a high speed ADC converter to the Arduino in order to have more values per iteration and frame. Have been looking at some yesterday.

Would a higher resolution then 10 bit give us an advantage ?

saludos

rsfoto · Post by **rsfoto** » Sat Dec 30, 2017 7:15 pm

Hi,

I have a question in regard to the setting of the following offsets which I guess do affect the values of Sun Intensity and Seeing

...
IValueB=IValueB/Number+.0; // + or – small dc off-set
...
ValueB=ValueB-.05; // + or – small dc off-set
...

How can I determine the offset ?

Thanks and

rsfoto · Post by **rsfoto** » Sat Dec 30, 2017 7:42 pm

rsfoto wrote: ↑Sat Dec 30, 2017 7:15 pm Hi,

I have a question in regard to the setting of the following offsets which I guess do affect the values of Sun Intensity and Seeing

...
IValueB=IValueB/Number+.0; // + or – small dc off-set
...
ValueB=ValueB-.05; // + or – small dc off-set
...
How can I determine the offset ?

Thanks and

Hi,

Well after posting this message I read again the paper and the offset comes from the difference the voltage divider is delivering to pin 3 and pin 5.

I just measured the voltage there and my divider delivers 2.455 Volt and so I now added to the sketch the correction value of 0.045 in order to get the 2.500 V. In both cases I added it in the lines.

¿ was this correct ?

Sorry if this is becoming a monologue

rsfoto · Post by **rsfoto** » Sat Dec 30, 2017 11:51 pm

Hi,

Interesting day today while playing around with the SSM. I think a seeing of about 0.14 ia a record here

Below some videos and graphs

900 mm focal length

https://vimeo.com/249199990

: 2017-12-30_12-44-14_graph.JPG (117.38 KiB) Viewed 13930 times

1260 mm

https://vimeo.com/249200078

: 2017-12-30_12-49-17_graph.JPG (113.15 KiB) Viewed 13930 times

1800 mm focal length which is more then the absolute limit for my setup using a 4x extender setup which gives me an f/21 and here I have to crank up the gain too much in order to have reasonable frames per second rates. I would say 1260 mm is more then enough for this.

https://vimeo.com/249200037

: 2017-12-30_12-56-17_graph.JPG (99.38 KiB) Viewed 13930 times

The next move could be to use my TOA 130 but moving everything there is too much work

OK, let me add that in regard to the values you should add the offset of 0.045 to the seeing value as because I later found out ¿? how to add the offset to the Arduino sketch.

Interesting is that after many hours of measuring the sky the values start to diminish to unreasonable values and only umpluggin the sensor cable or restarting the Arduino brings it back to normality ¿ have you experienced that too ?

Tomorrow another session.

PDB · Post by **PDB** » Sun Dec 31, 2017 11:46 am

Hi Rainer,

seems I missed a few additions to this post. (Did not get any e-mails to warn me)
Never really thought about the DC offset, but what I noticed is that my seeing was much more optimistic with higher values. Maybe need to take my multimeter again and do some measurements.

I guess it all depends of focal length and resolution per pixel. Am I correct in assuming that the less resolution ( more arcseconds per pixel) requires a better seeing the high resolution (less arcseconds per pixel) or is it the way around ?

Well to my mind the limiting factor is your scope's maximum resolutiion when you use a camera that is not undersampled. In that case it will be the resolution of the camera which will be the limiting factor. Eg for my 80mm and ASI-120 @ f/24 I set the trigger usually around 1.2'' (scope = 1.5 theoretically)

I then use the SSM to trigger the start of a recording when seeind drops below that seeing level (hoping I will have some seconds of good seeing).

Regards,

Paul

rsfoto · Post by **rsfoto** » Sun Dec 31, 2017 3:00 pm

Hi Paul,

Maybe need to take my multimeter again and do some measurements.

Measure from ground to pin 3 and 5 which are the ones which get the 2.5 Volt from the voltage divider with the 2x 10K resistors abd rest that value from 2.5 and you get the offset.

Have a Happy New Year 2018 ( while writing this still 15 hours to go )

Saludos

rsfoto · Post by **rsfoto** » Sun Dec 31, 2017 6:43 pm

Hi,

Sitting in the observatory and making some tests I decided to connect an additional power supply to the Arduino and now the voltage at the 10K/10K divider is 2.488 Volt and so the offset is only 0.012 for the program.

Looks like I will have a small hole in the cloud cover :-) Let us see ...