Solar Tracker Circuit Is Giving Me Issues - Appreciate Anyone's Help

[marktinkerman]

cost is the criterion for me as I need at least 10 tracking assemblies for what I want to do during the summer which is power providing at festivals ... sometimes a big combined array for a stage sometimes small separate arrays for cafe, stall or poets corner etc. so at least 10 arrays but at on the positive side its only single axis control I need as I'm setting up at new festie each week or so. Following your development with interest mark

 

[rs14smith]

cost is the criterion for me as I need at least 10 tracking assemblies for what I want to do during the summer which is power providing at festivals ... sometimes a big combined array for a stage sometimes small separate arrays for cafe, stall or poets corner etc. so at least 10 arrays but at on the positive side its only single axis control I need as I'm setting up at new festie each week or so. Following your development with interest mark

Yeah cost was a big issue for me as well, as all I'm building is a mere "prototype" (was suppose to be for a college project), as far as the construction of the solar tracker (what it's built out of, how big it will be, etc), but I want the actual circuit to be able to handle a general size motor (12v for example) that you could use in the real world if you wanted to.

I finished both video test now, so all I have to do now is import it and upload it shortly.

 

[rs14smith]

Solar Tracker Test 2:
[video=vimeo;10312475]http://vimeo.com/10312475[/video]

Solar Tracker Test 2-2:
[video=vimeo;10313222]http://vimeo.com/10313222[/video]

Also, just a note: The N-Channel FETs can handle up to 175 Degrees C (347 Degrees F), so in the test where the heat of the FET is rising, it was still below what it could handle which I did not know. Just something good to know

 

[Dragon Tamer]

You need to check the current consumption of your linear actuator. I'm still convinced that your circuit design is poor and that it could be what causes the heat itself. In my opinion, if the circuit is made properly, then the little 12V motor and actuator should not cause any kind of heat generation. Please post a more organized schematic of your circuit.

I'm going to "poke" at your schematic a little, I think that U2 C and D are a redundancy. I don't understand why you have the 2 2.2k resistors. Also I think that you should get rid of the voltage divider on U2 A and B. Why not use a standard H-Bridge, why do you have to use the MOSFET's in that particular configuration?

I've included an attachment of a circuit I made several years ago, it's a simple circuit (yes this circuit works) and it can track fairly accurately. The motor should be as low of a power as possible.

 

[rs14smith]

You need to check the current consumption of your linear actuator. I'm still convinced that your circuit design is poor and that it could be what causes the heat itself. In my opinion, if the circuit is made properly, then the little 12V motor and actuator should not cause any kind of heat generation. Please post a more organized schematic of your circuit.

I'm going to "poke" at your schematic a little, I think that U2 C and D are a redundancy. I don't understand why you have the 2 2.2k resistors. Also I think that you should get rid of the voltage divider on U2 A and B. Why not use a standard H-Bridge, why do you have to use the MOSFET's in that particular configuration?

I've included an attachment of a circuit I made several years ago, it's a simple circuit (yes this circuit works) and it can track fairly accurately. The motor should be as low of a power as possible.

The current consumption of the linear actuator with no load is 2A.

Dragon Tamer: "Please post a more organized schematic of your circuit."
I thought the previous schematic I directed you to was quite organized, compared to for example sketches of circuits, similar to the one's you posted previously. I'm not sure how much more organized one could make the schematic, so if you can please kindly describe to me what exactly isn't organized so I can organize that for you that would be great .
I'm using Multisim for my schematics, so making changes is quite easy.

As far as the hookup for the comparator and resistors in the schematic, most of this was recommended/designed by mneary as you can see on the previous pages of this thread, so I'll allow him to explain that part of your question.

As far as the H-Bridge is concerned, that was a lengthy discussion as well in this thread, and it was concluded from others that it's easier/better to create an H-Bridge with P-FETs and N-FETs (p channel FETs are used to SOURCE current (top side) and n channel FETs are used to SINK current (bottom side)).

 

[mneary]

The voltage divider on U2A/B is intended to provide a region where tiny changes in light don't cause the motor to move. Without it, the motor always runs, reversing several times a second. If you want the motors to run all the time accomplishing nothing, take it out.

U2 C/D shut the motors down when it is dark. You cannot depend on U2A/B to be "off" when both inputs are 0V. If you want your solar array to hunt for sunshine in the darkness, they can be removed.

The 2.2K resistors are required. U2 is an open collector comparator and has no internal pull ups. Take them out if you get tired of seeing it work.

A standard H bridge is a possibility if you want to spend the money.

 

[Dragon Tamer]

Thank you for explaining that (I've been asking about that for a while now). On another note, I noticed that you didn't include limit switches... why? Since you are using FETs then it should be easy to put them into the design. Instead of turning the motor off at night, why don't you use U2 C/D to make the circuit think that the sun was in the easern sky?

 

[rs14smith]

Thank you for explaining that (I've been asking about that for a while now). On another note, I noticed that you didn't include limit switches... why? Since you are using FETs then it should be easy to put them into the design. Instead of turning the motor off at night, why don't you use U2 C/D to make the circuit think that the sun was in the easern sky?

That's what I was thinking about since I will be using a "linear actuator" in this design as when the actuator reaches it's full stroke length or when the stroke is pulled all the way in, it locks the stroke. Unlike the Hobby Motor, it would just continue to turn east all night long wasting power, which isn't ideal at all. So if we go this route, it'll be like we are excluding any other motor from being used.

Another idea would be to use some type of sensor that we could attach to the solar array/somewhere on the frame, and once that sensor meets back up with another sensor that would be located in the far east location, the motor would stop turning east and the motor would turn off.
A better example would be, if you have an alarm system in your home, you know the alarm system guys place sensors inside the doors so that when the door opens, it breaks the signal, and when the door closes, it re-activates the signal.

Just an idea

 

[mneary]

We were building this circuit a step at a time and limit switches were next, one each to be placed at the bottom of the 2.2k resistors. The limit switch on the Eastern extreme of the motor excursion would probably make the darkness comparator on the East photodiode unnecessary.

But Dragon Tamer entered in message #116 saying that we scrap it and start over.

Your call.

 

[rs14smith]

We were building this circuit a step at a time and limit switches were next, one each to be placed at the bottom of the 2.2k resistors. The limit switch on the Eastern extreme of the motor excursion would probably make the darkness comparator on the East photodiode unnecessary.

But Dragon Tamer entered in message #116 saying that we scrap it and start over.

Your call.

Ah okay, I didn't know we had already established what method we would use next to implement the return back to east function.

As for starting over, lol I was never on board with Dragon with that idea, as we are talking about wasted time/money, so I'm not really focused too much on simplicity, but on getting the tracker to work. Afterwards, we can determine what can be done to simplify it, but yeah, my goal is just to get the tracker to work, and understand what is taking place in the circuit, rather than just taking someone else's circuit online and guessing what is taking place.

As far as the limit switches is concerned, I can't remember if I've ever used those or not (I probably have but have forgotten). So I'm not quite sure how to implement that at the moment in the circuit, so I'll surely need help there

 

[Dragon Tamer]

I had an idea for the return east function, but seeing as how you have that taken care of, I'm going to stay out of it. I do however have another idea, since Q4 is getting hot, why not make a circuit that will cool it? Really simple to make and can be part of a different circuit. I'm under the assumtion that this is going to be outside (if not I'm going to have to drive all the way down to Tennessee just to smack some sense into you ). You'll obviously need to make a housing for your circuit. The fan can just sit at the bottom of the housing to cool (so rain does not get in). I can have the circuit ready for you today if you want.

Also, If you still want to make the "North" and "South" functions, then my idea for the return to east will be really simple (and good).

 

[rs14smith]

I had an idea for the return east function, but seeing as how you have that taken care of, I'm going to stay out of it. I do however have another idea, since Q4 is getting hot, why not make a circuit that will cool it? Really simple to make and can be part of a different circuit. I'm under the assumtion that this is going to be outside (if not I'm going to have to drive all the way down to Tennessee just to smack some sense into you ). You'll obviously need to make a housing for your circuit. The fan can just sit at the bottom of the housing to cool (so rain does not get in). I can have the circuit ready for you today if you want.

Also, If you still want to make the "North" and "South" functions, then my idea for the return to east will be really simple (and good).

Well as mneary was saying, we are trying to take this 1 step at a time, which allows you to accomplish things a lot quicker than jumping around here and there. In the real world project, it would be easy to allow others (1 team) to be working on 1 part of the circuit, and another (team) working on the other. But here in the forum, it's hard to do that as it just makes things confusing on what exactly is taking place, espically for those who may read this long thread 3 months from now.

So at this moment, what we are working on is the return to east direction. Once that is completed, we can offically say we have created a single axis solar tracker which is soemthing "mark" is looking for. But of course, we need to create an enclosure as you were stating, but 1 thing to keep in mind about the heat issue is, the FET that is becoming hot as we "rapidly" switch polarities has a max temperature of 175 C, and in the video test, I only got up to around 60 C. Another thing is, what most people recommend using for FETs is simply a heat sink. So I think, a heat sink combined with the fan will be more than enough, as another thing to remember is, in the test, I performed logics that are unlikely to happen if you were to sit the tracker outside. The sun isn't going to be going back and forth that fast, at least I hope not lol

So the enclosure (heat sink/cooling circuit) part is something we would possibly work on after we implement the return back to east function.

Hope that makes sense, it's just nice to have a linear layout/plan.

 

[mneary]

Does Q4 get hot/warm when the system is idling or just when you go East a lot?

 

[rs14smith]

Does Q4 get hot/warm when the system is idling or just when you go East a lot?

Well in the video test I performed, the temperature only went up when the switching between polarities was going on "continuously". The temperature goes back down when it's idle or in other words, doing nothing.

So the heat issue didn't really have an issue with east/west, it's when the polarities continue to switch back and forth continuously. Watch the linear actuator video test (test 2-2) and you'll see what I mean if that's not clear enough.

 

[Dragon Tamer]

I was going to say that you could add a second comparator for the up and down rotation (I'll call it U1), and just make U2 C/D work with U1 C/D then just put a limit switch (PBNC) on the V+ pin of the 339.

I have made the fan for the circuit if your still interested in it. It doesn't even interact with the solar tracker circuit, it does it's own specific duties (with relative efficiency). You can even add a simple solar activated circuit to turn the fan and the IC's off at night so they do not waste power. The circuit can be arranged for several different configurations, but this is the best because it will cut the power to the fan if the fan is not needed. It will differentiate between the temp of the heat sink on Q4 and the ambient temperature. R2 is to adjust the speed of the fan. You can use any op-amp with this circuit, but it must be an op-amp. R12 can be replaced with a higher resistance value, but the gain will not be as high as it is currently set to (a gain of 11). If you have any questions just ask.

 

[rs14smith]

I was going to say that you could add a second comparator for the up and down rotation (I'll call it U1), and just make U2 C/D work with U1 C/D then just put a limit switch (PBNC) on the V+ pin of the 339.

I have made the fan for the circuit if your still interested in it. It doesn't even interact with the solar tracker circuit, it does it's own specific duties (with relative efficiency). You can even add a simple solar activated circuit to turn the fan and the IC's off at night so they do not waste power. The circuit can be arranged for several different configurations, but this is the best because it will cut the power to the fan if the fan is not needed. It will differentiate between the temp of the heat sink on Q4 and the ambient temperature. R2 is to adjust the speed of the fan. You can use any op-amp with this circuit, but it must be an op-amp. R12 can be replaced with a higher resistance value, but the gain will not be as high as it is currently set to (a gain of 11). If you have any questions just ask.

Great. Can you by chance draw up a example circuit (very simple version) of a limit switch so I can possibly start playing around with the current circuit to try to implement the return back to east? That way while mneary and others are trying o figure it out, I can do the same, but at the moment I can't quite remember what a limit switch looks like in a circuit...I know I've used one before, but can't remember.

I know in one of the schematics (solar tracker) you posted had one in it, however, I'm looking for a straight forward example of just a "limit switch (1 limit switch)" circuit so I can get the idea of how it works...again...

If you or no one else have the time, I'll just continue to google, but that would be helpful, at least for me

 

[Dragon Tamer]

Here is a link to where you can buy a limit switch, and **broken link removed** is an example of a limit switch in a circuit. (copmpliments of google)

 

[nickelflippr]

Just thinking out loud on the return to East and other functions:

1. Use a "dark activated" 555 one shot circuit, to run the motor/actuator back to the East side limit switch. I think this could be the low output from the "light activated" circuit described in item 3.

2. Use a 556 and have the option of a PWM function to keep Fets and motors cool, and also allow for differing gear ratios on the rotating drives.

2. Use a dedicated photo diode/cell "light activated" circuit, with the diode/cell on the East side of the project box. This could serve double duty as a "universal" minimum light level for the comparator inputs, and the "dark activated" return to East function.

 

[marktinkerman]

here is a link to a tracker with limit switches that I found maybe it helps **broken link removed** mark

 

[rs14smith]

Well it seems from my understanding of "limit switches" they consist of some type of spring that has to be "pressed inward" similar to a "push button" in order to activate or deactivate a component/electrical action. So if my understanding of that is correct which it probably is not , that would mean that the only way to effectively use this limit switch is if someone "manually" pressed the spring (switch) when it's night time, or if you constructed your solar tracker so that when the angle/tilt reached the maximum it could go "west", the limit switch would have to be made to bump into something to signal that it was night time?

But just because my solar tracker is facing the "west" direction and cannot turn anymore, doesn't necessary mean it's night time (completely dark), so I'm hoping my logic behind it is wrong and it's not something that has to be manually pressed in order to activate/deactivate a component/electrical action, as we will possibly loose a bit of accuracy there if you see what I mean?

My idea or logic behind it would be, implement something that see's when both photo transistors are 0v, it begins to turn back east, and "then" implement a limit switch to stop the motor from running once the solar tracker (the actual construction build of the solar array) returns to it's starting position (east); it would basically bump something and trigger the switch. Of course that will only work if my understanding of limit switches are correct or not.