Sol Guider using Solar tracker circuit

Mr RB said:

Their solar tracking rate might not be that accurate, but mine is.

Even driven from a xtal you can tune it to one PPM accuracy, which is about "X" / 1000000 per adjustment step. Alternatively running from a GPS lock it gets about a hundred times more accurate than that!

Anyway if you prefer a closed loop system to really accurate clockwork tracking that's up to you. In that case I think sensor design, reliability and cleaning will be critical.

Click to expand...

Hi Roman,

Well their Solar tracking is quite accurate but the most problems with amateur solar imagers is the perfect alignment of the mounts. Do not think it is that easy. Also a lot more para meters do play a role here., especially refraction. The closer the object ois to the horizon its speed decreases. See here for example the Special King Rate which adjust the speed continously from Zenith to horizon, etc, etc, etc.

What I intend to do is to just add a guider to an already existing tracking device of the Sun. I have read your page. Very interesting. Will keep getting ideas from your Website too.

I made some tests this weekend and the idea I am develipong seems to go in the right way. As you said precisión here is the Key word.

The interesting thing of this project is what I have learned in the last weeks and days.

Hi,

On the weekend I made a simple hardware in order to get a differential voltage fom a Pair of sensor LEDs. Today I had the opportunity to test it. THe sensitivity to an angle change is quite good. This I could test using different tracking speeds which I can actuate manually in all 4 directions. The 4 red LEDs tell me what is going on.

It works great. Still lot to do until I get the desired sensitivity

Here is an image of my todays setup.

I will need to integrate another circuit to shut down the guiding function in the moment it gets cloudy or the Sun is covered by half or any non acceptable amount of clouds.

regards Rainer

rsfoto said:

Hi Roman,
Well their Solar tracking is quite accurate but the most problems with amateur solar imagers is the perfect alignment of the mounts. Do not think it is that easy. Also a lot more para meters do play a role here., especially refraction. The closer the object ois to the horizon its speed decreases. See here for example the Special King Rate which adjust the speed continously from Zenith to horizon, etc, etc, etc.
...

Click to expand...

Thanks for the info, I did not know that and appreciate it. My (limited) experience with telescope tracking has been with helping a few people set up Linistepper drivers on telescopes for star photography where they tend to avoid the sky near the horizon.

rsfoto said:

...
What I intend to do is to just add a guider to an already existing tracking device of the Sun. I have read your page. Very interesting. Will keep getting ideas from your Website too.
...

Click to expand...

I haven't made a solar sensor but have a few ideas to throw in the mix... Making it robust will be important. A flat glass lens at the top that can be easily cleaned with one wipe. Not plastic (as that will craze and go opaque with UV) and with no crevices where bugs/dirt etc could build up.

Then the sensors themselves need to be robust. Most opto devices will age and deteriorate over time, and age worse when exposed to strong light especially high power sunlight full of UV etc. Of course heat makes it worse.

If the sensors are at the bottom of the tube, I would add a glass filter in the tube to reduce the UV to zero of possible, and the light levels to many times lower. You might be able to use one or more glass lenses from sunglasses?

Or maybe you could use a pinhole followed by a glass lens to focus 1sqmm of light onto an area the size of the tube, to reduce the light power level by hundreds of times. That might also give benefits as to how angular sensitive it is.

Also remember the UV will destroy all common plastics pretty quickly, so the final design should really be metal, painted white to reduce the heat.

Hi Roman,

Thanks for the excellent addings to the idea for the Solar tracker.

We can get so called UV/IR filter which cutoff from 400nm to 700nm the light from the Sun.

About dimming this is worth doing some and see what influence it has on the sensitivity. There are so called ND = Neutral density filter for Photography which dimm the light and do NOT change the color of itself.

I still have a lot of trials to do with different ways of defining the light fall onto the sensors with different types of slits etc. Still a long way to go in this regard. Once I am 90% convinced this all will be wither built an Aluminum housing be it s tube or square case.

Again thanks for the comments.

The sensitivity depends a lot of the length from where the sensors are located and the entrance holes for the collimated ray of the Sun. The one you see in the photo has per each entrance hole 2 sensor LED which are onle covered by let me say mayube 40% of their surface. When both LED get the same light I am centered on the Sun. The slightest movement makes a change of the shining surface onto the sensor LED by turnin from a circle into an ellipse and immediately one sensor LED get much more light and activates the switch for the correcting movement.

I have made some drawings in order to calculate the surface change in regard to angle change, etc, etc,etc.

This is an interesting exercise. There is already a protoype from a company dedicated to astronomy accesories but the price still has not come out. Look here

https://digiborg.files.wordpress.com/2013/05/hinode1.jpg

I might be reinventing the wheel, but just as an exercise and keep my brain active I am working on my solution. As you see it is already very sofisticated. Let me see if it is possible to doit in a simpler way

Thanks for the link, I'll check that out.

Well a ew days later I have tested the solar guider everytime I had a bit of Sun but it has gone rare here in Mexico and so I decided to develop a few other circuits for control purposes, shut down of the guding and changing polarity for morning and afternoon imaging sessions. This is necessary as the equatorial mounts change the corrections before the meridian and after the meridian.

The first circuit is an automatic shut off of the guiding if the Sun is covered by clouds in order to avoid that the mounts searches like crazy for the Sun. I am using an LM393 and with a LED as sensor I measure the intensity of the Sun light directly into the Sun. The shut On and Off is ver clean with no flickering.



The next circuit is for changing the guiding corrections between morning and afternoon. I used a 555 in Bistable mode and a DPDT relay switch with a coil inside



So I am still working hard for the hardware to focus the Sun light to the guider sensor LEDs in order to get a precise guding below 0.5 arcminutes or even less.

regards Rainer

MikeMl said:

It is hard to tell from your photo, but do the optical axes of the two LED diverge by about 20 to 40 degrees? If not, try that. It might give you a smoother crossover in the center...

As long as the wire to the LED sensor array is shielded, 5m shouldn't be a problem.

If you run into a problem where an LED flickers on-off as the sun angle changes, write back, and I will add a little hysteresis to the comparator switching points.

Click to expand...

Hi Mike,

I am coming back to your offer in regards to Hysteresis as I have noticed while changeover from one LED to the other they are slightly flickering, eg. my switching would not be clean. I already tried with a 10K Pullup resistor between the output and +V as a 1MΩ resitor between output and + input of each comparatator but something is running wrong and it does not work as it should.

Is this due to the fact that there is the connection of the + input of one comparator with the negative input of the paired comparator ?

Thanks and regards Rainer

Another polarity switch using this circuit

https://forum.allaboutcircuits.com/sh...ad.php?t=40717

and, of course being a fanatic of optocouplers , I used 4 of them for making a DPDT switch.

I found this one more elegant as I just need 1 toggle switch in orcer to change between 2 modus.

Instead of using a 1μF Tantalum capacitor I used a 0.1μF and with a quick push on the toggle button it switches over very fast.

Just breadboarded the above circuit and played with it. Now I will add 4x 4N25 with some load resistors and see what comes out of the optocouplers.

When applying power the red LED goes on first which I will use for AM modus (Telescope on the WEST side of the mount) and then with the toogle switch go to the green LED as PM modus (Telescope on the EAST side of the mount).

As I am not switching more then 0.2mA coming from the Inputs for reversing polarity why use a huge DPDT realy with a coil.

So in total I have now 3 separate circuit blocks for my Solar guider.

The guider itself using an LM339, an automatic ON-OFF circuit using a LM393 in case of clouds and this one for changing AM-PM modus with the NE555N as Bistable Multivibrator.

Next step will be to get some hardware for the Sun ray projecting onto the sensor LEDS as accurate as possible and then solder the first prototype on a pre drilled plate. I hope to finish before August first because then I have one week time in the Observatory for making the real tests with a Sun inging set up.

Thanks for reading and looking and if you find a bug please report and THANKS A LOT for the support I have got here

Well.

All this started here on 15 of June 2013 and today finally I was able to do a longer real life test on an Equatorial mount and a camera in order to see how it works.

I am really satisfied how far I got in this 8-9 weeks ...

Without your help I would not have got so far HUGE THANKS especially to Mike for the circuit with the inverted LEDs as sensors. I made some measurements and those are really sensible to a tiny change of illumintaed surface or not.

I have to analize now the video I made 80 minutes in total to get the sensibility and then work on more hardware.

Tomorrow more videos will be shot.

Look at the image here

Hi Mike,

Long time no hear from you

The circuit is working quite well.

Here you can see a video of a 1hour 45 min guiding of an equatorial mount on the Sun. The RA axis error is at about 60 arcseconds which is better then a few days ago but still not very satisfactory.

**broken link removed**

Yesterday night with a test device I was playing with the Deadband resistor. The above movie was done using 100Ω . I then added in parallel another 100Ω resistor and I saw that the deadband was about half of what it was before. So I decided to solder into the circuit I am testing a second 100Ω resistor and so work with a narrower deadband.

Where do you think is the limit in this circuit for the deadband resistor ? By the way, even with flickering commands to the autoguider port, the mount behaves quite well. That would mean that a hysteresis is not necessary. I was able to measure the maximum and minimum voltage and it was between 6.8mV and 24.1mV that means the deadband ? is about 17mV ?

Since today in the morning I am testing again the circuit with a 50Ω deadband resistor and it looks better then before. I will have to analize the videos when they are ready. I am taking every 6 seconds one image of the Sun and so I get a nice time lapse movie which was the reason for building this solar guider.

How can I calculate the limit for the deadband resistor ? What would happen if I have NO deadband resistor ?

Later this day I will post the movies from today with the 50Ω resistor.

Again I appreciate your great help with this circuit which is exactly what I was looking for.

Sorry, but your video is huge, and I am on a slow DSL line.

It occurs to me that you are not utilizing the deadband at all. The deadband was required in the original solar tracker so that at sunup the next day, the mount would slew (backwards) to the Eastern horizon to align itself with the rising sun, and then slew forward as the sun travels from East to West. The deadband is only needed to prevent the motor from "hunting", where it overshoots and then actually reverses to run the stage in reverse.

I still think that in your application you can get by with just a single comparator (bang-bang, thermostat-like) control system. Set the motor drive up so that it runs just slightly faster than would be required to track the sun movement. Use the comparator to start/stop the motor to keep the LED voltage at the comparator trip point. If the motor gets ahead of itself, it just needs to sit there for a short time and the sun will "catch-up". You will have to manually set the system up for each observation. The comparator needs to have just a small amount of hysteresis to prevent "chattering".

A much better system would be a PID controller based on an error voltage derived from the LED output minus a set point, where the motor always runs, but its speed is modulated to keep the error voltage constant. That would prevent any jerks from effecting your scope mount.

MikeMl said:

Sorry, but your video is huge, and I am on a slow DSL line.

It occurs to me that you are not utilizing the deadband at all. The deadband was required in the original solar tracker so that at sunup the next day, the mount would slew (backwards) to the Eastern horizon to align itself with the rising sun, and then slew forward as the sun travels from East to West. The deadband is only needed to prevent the motor from "hunting", where it overshoots and then actually reverses to run the stage in reverse.

I still think that in your application you can get by with just a single comparator (bang-bang, thermostat-like) control system. Set the motor drive up so that it runs just slightly faster than would be required to track the sun movement. Use the comparator to start/stop the motor to keep the LED voltage at the comparator trip point. If the motor gets ahead of itself, it just needs to sit there for a short time and the sun will "catch-up". You will have to manually set the system up for each observation. The comparator needs to have just a small amount of hysteresis to prevent "chattering".

A much better system would be a PID controller based on an error voltage derived from the LED output minus a set point, where the motor always runs, but its speed is modulated to keep the error voltage constant. That would prevent any jerks from effecting your scope mount.

Click to expand...

HI Mike,

Thanks for the explanation. Let me digest it and I will come back later.

One of the problems is that some Equatorial mounts can be adjusted in the tracking rate and some others not.

I think just as the circuit is now I can use it and play aroiound with the deadband resistor. The last test I used a, if I remember well, 29Ω resistor and the voltage between lighting up onw LED and the other was aropund 10mV only and that is quite good with no lighting up of both LEDs at the same time. So I will change the existing resitor in the soldered board and make new tests as soon as I have Sun again.

A friend build me test device where I can read on a micrometer 0.01mm and that is how I am checking the deadband by moving the hole which lets through the light.

rsfoto said:

HI Mike,

Thanks for the explanation. Let me digest it and I will come back later.

One of the problems is that some Equatorial mounts can be adjusted in the tracking rate and some others not.

I think just as the circuit is now I can use it and play aroiound with the deadband resistor. The last test I used a, if I remember well, 29Ω resistor and the voltage between lighting up onw LED and the other was aropund 10mV only and that is quite good with no lighting up of both LEDs at the same time. So I will change the existing resitor in the soldered board and make new tests as soon as I have Sun again.

A friend build me test device where I can read on a micrometer 0.01mm and that is how I am checking the deadband by moving the hole which lets through the light.

Click to expand...

Hi Mike,

Sorry for coming back so late. The problem here is that some mountd do allow a speed regulation and some do not. Now perhaps you think that I want to build this commercially but you can be rest assured that I am not going to do a business with it.

I will present it at various astronomy forums and if the people are interested then they can get the circuit and they built it by themselves.

The last trials I have made by using different resistors as deadband have been quite positive and I was able to get a clean switching with a ~10mV window which is quite good.

My next step is to try a longer tube for the sensor LEDs which means that my sensitivity will increase compared to a 100mm long sensor tube.

Thanks again