1 Axis Solar-Tracker

[rieders]

Hello
I would like to build a solar tracker for 2 solar panels each driven by an actuator.
The whole should be on a flat roof.
The engines are to be powered by an L298n. The voltage should be 12 v to use the voltage converter from the L298n for an ESP32.
The angle of the solar modules is to be monitored with an MPU5060, one per module. In addition, an anemometer should be connected via analog port.
I would like the modules to fold up according to the angle of the sun and fold down in the dark and in strong winds.
I've seen a lot of trackers, most of them have a light sensor.
I think if there is snow or dirt on the sensor the system will stop working.
Hence the query via the gyro/position sensor.

The question is whether this can be implemented with ESP-Easy.

Kind regards

Andre

[TD-er]

In theory it could, however you really should think about how the whole tracker will return to a safe position in case of some failure.

For example, what if there was some power outage, the ESP has no idea about the time/date, etc. and it is really terrible weather.
What would be the safe position and how can you guarantee this?

Also how will you detect when the system is blocked from moving?
For example if the entire entrapment is covered in snow, or there is something (or someone) blocking the tracker.

How will you decide whether it is too windy for a specific orientation of the panels?

[rieders]

Many thanks for the quick response.

The position is to be monitored by means of the gyro sensor.
The wind should be monitored with an anemometer.

If the power fails, the system is powered by a battery that is charged by solar.

If the mechanics are blocked, this is monitored by a gyro

[TD-er]

I've been playing a bit with such sensors, but I'm not that enthusiastic about their stability regarding the absolute orientation.
Sure they can detect gradual changes, but when having them laying still, you will see some drift.
Also they show some drift due to temperature.

With the gyro sensor in phones, you also sometimes need to rotate them around a few times to have a proper general idea of orientation again.

I think it is more useful to have some marking on the tracker axis and consider that marking as reference point.
For example in the center of the range of motion in both axis.

Also you need some end stops which will actually cut the power to the motors.
Maybe you can also attach some tacho disc and optical sensor on the axis of the motor to detect motion when moving.
These are all supported as "inputs" for the "PWM motor controller".

I also suggest to connect some shunt on the designated pins of the L298n to make sure you won't overload the H-bridge and/or motor.
Those L298N H-bridges are not really capable of a lot of current and moving a large solar panel tracker can take quite some power... or take a lot of time to move again to the desired orientation.


Do you plan on following a specific track based on the longitude/latitude and time/date?
Or use 4 LDRs + some small shield to make the tracker follow the sun automatically? (not sure whether this may give false "directions" when some clouds move in front of the sun)

[rieders]

TD - thanks for the help

I have 2 of these linear actuators.
https://www.vevor.de/linearantrieb-c_11 ... f4f7936e.3

I just want to flip the modules.
A shunt to protect against overload is good.
The drives each have 2 limit switches.

I would like to determine the optimal angle using position data/time and module direction.
After sunrise the module should move into the area of the angle. It doesn't have to be very fast. A query from the angle could be made every 5 minutes and corrected.
At winds above x m/s the modules should lay flat.
After sunset, the modules should also lie flat.
A possibility to control the modules manually via web interface for maintenance or cleaning would also be good.

[TD-er]

mit DC 12V Ausgangsspannung und 8-9A Ausgangsstrom zu verwenden.

I think you need to look for other H-bridge drivers.
I know those can be set in parallel, but still it is next to impossible to equally divide the current over all drivers.
So that will be one huge disappointment as they will for sure burn out.

[rieders]

I'll test tomorrow what kind of current is actually flowing.
Without burden and with burden.

[TD-er]

The start current will often be much higher compared to the "running" current.
Thus it probably doesn't really matter if you have a load connected.

[chromo23]

mit DC 12V Ausgangsspannung und 8-9A Ausgangsstrom zu verwenden.

As unfortunately to often with this kind of things, the specs are not right.
Rated current seems to be, a bit more believable, 3A
They provide a kind of datasheet here: https://d2v0huudrf11kh.cloudfront.net/v ... 956022.jpg

[TD-er]

Still 3A is way more than the L298N can handle.

[chromo23]

And since this thing has limit switches just use a relay board with two relays for back and forth and like Td-er mentioned some kind of decoder.
A magnet on the rotor and a hall sensor for counting pulses would be sufficient.
I don´t think that stalling will be an issue but you could put some ptc in between the motor and the powersource and limit the time the relays are switched apart from counting the pulses...

[chromo23]

A possibility to control the modules manually via web interface for maintenance or cleaning would also be good.

Then maybe easyfetch is something for you. Follow the link in my signature.

[rieders]

Good day
I will use this motor driver for the project.
DollaTek 7A 160W L298 PWM Speed Controller Dual DC Motor Drive Module
https://www.amazon.de/DollaTek-Speed Controller-Motor-Drive-Modul/dp/B099DRSL45/ref=sr_1_14?keywords=L298+Motor+Controller&qid=1674129927&sr=8-14
Alternatively, one could also take the other parallel.
This increases the current from 2 to 4A.

[TD-er]

The L298 datasheet mentions a max. of 3.5A when set parallel if I'm not mistaken.
Also you should add a small resistor (in resistance, it will get hot, so make it a 'beefy' one) of like 0.33Ohm or something like that in series with each separate H-bridge (on both sides) output before joining them together.
This way you will even the load on each bridge part as much as possible.

[rieders]

Hello.
I saw that atom.io is no longer supported.
now i wanted to activate some unofficial plugins with vsc in esp32.
since I did it under atom.io a long time ago, I don't remember how it worked.
I hope someone can help me.

Regards

[TD-er]

Have you seen this page in the docs? https://espeasy.readthedocs.io/en/lates ... ith-vscode

[Ton_vN]

Fixed Solarpanels only are correctly/optimally positioned once per day.
Therefore your approach for a 1-axistracker makes interesting reading,
keeping the center azimuth of the panels in line with the sun's position, optimally exploiting the sun's radiation,
while avoiding the complexity of an elevation drive.

To check on your realisation:
for 'aiming' of the pedestal, are you now in practise using a lighttracker or using another method?

The sun is well known to have a predefined trajectory over the day and over the year:
perhaps easiest and most reliable to have a processor look at the trajectory-data and use that info for aiming of the pedestal.