Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Solar controller optimisation ?

Status
Not open for further replies.

mikkoj

New Member
Use of the solar panel charger’s surplus energy
I have built the solar panel battery shunt regulator described in this thread. The way I use it is not so urban or sexy. I have a summer cottage in southern Finland. The cottage has a WC and all the sewage water from it goes to a container, which must be emptied at least once a year. In order to reduce its load I have built a separate dry toilet. Those, as some of you may know, can create a bad smell, if you do not ventilate them well. In order to reduce the risk of smell I have started to construct a solar panel-driven ventilator, which starts to run as soon as I open the lid of the toilet seat.
In summer time in southern Finland the sun is about 18 to 20 hours above the horizon. However, because the solar panel will be in a fixed position, the maximum power will be got only in the middle of the day, and at that time on a sunny day there probably will be a surplus of energy (the panel produces 12 V 5 W maximum). The surplus could be used for extra ventilation

1A2ym9D

Following the outlines discussed on this thread I have built a controller described in the adjoined schematics. As a shunt I used a low-power P-FET BSS92 with a load resistor of 150 Ohms. This resistor restricts the charger’s maximal shunt current to about 100 mA. When the voltage at the drain pin of the FET raises to the up-going threshold voltage (about 10 V, the shunt current is then about 66 mA) of CD40106 Schmitt trigger inverter 1, the capacitor at inverter 2 is discharged, inverter 2 output goes high and the transistor BC337 conducts. Now the fan (from a computer power supply) will run as long as the voltage at the input of inverter 1 is above the down-going threshold voltage (about 5 V). When the voltage goes under the low-threshold, the fan continues to run for the delay time determined by the RC-combination at inverter 2 input. The fan can be started and the delay activated manually any time by opening the toilette seat lid (which closes a normally open door switch). The power for the ventilator comes directly from the lead acid battery, which means that the fan always runs at full power. This can reduce the battery voltage, if the solar panel cannot deliver enough current to compensate it (the fan takes about 120 mA). As soon as the battery voltage is reduced, the FET is closed. The fan still runs for the delay period and then shuts down. Now, if the sun still shines, the battery voltage starts to rise, and when it is fully loaded, the FET will be activated and the cycle starts again.
One problem is that the total current consumption does not reach 300 mA, which maximum value is given in the specs of the solar panel. The fact that there is the ON- delay and that the fan always runs at full power may solve the problem. The shunt current can be increased by reducing the value of R1, but then a bigger portion of the energy will be wasted. The optimal resistor value can only roughly be estimated without knowing the real current production of the panel and frequency of the use of the toilet.
I appreciate all improvements and hints you could give to me. I am a hobbyist with no training in electronics. I have got all my knowledge by reading forums like this. I have not tested the circuit, since it still is a dark and cloudy winter here.
 

Attachments

  • solar delay.jpg
    solar delay.jpg
    68 KB · Views: 235
Is this what you are trying to accomplish:?

1. You want to charge the battery whenever the sun shines?

2. When the button is pushed, the fan runs for a few min, and then shuts off automatically, hopefully leaving some energy in the battery?

3. If the button is pushed again, the fan runs again; could happen several times until the battery is exhausted.

4. If the sun comes back, then charge the battery without overcharging it?
 
Yes Mike, those are the things I want to accomplish. In addition, I want to run the fan any time when the battery is fully charged and there is extra sun power available, but not run it too long to ruin the battery charge

An additional useful feature would be shutting down the fan when the battery charge has been fallen to a critical value. I think this could be accomplished using another TL431 connected to the gate of an N-FET. I think here the resistor at the cathode of TL431 must be bigger in order to reduce the idle current.
 
Last edited:
Lets see if this is the spec?

Suppose we have two voltage detectors:

X=V(bat)>11.5V (used to prevent battery discharge below 11.5V)

Y=V(bat)>14.2V (used to prevent the battery from being overcharged)

We have a time delay T, initiated by momentarily pushing a button, duration = 60s?

We have a fan F. We want the fan to run if the (battery voltage is > 11.5V) and (T is true) or whenever the (battery voltage > 14.2V).

F=X*T + Y

Note that I am using the fan as the dummy load into which the solar panel current is shunted after the battery reaches 14.2V

Would this work?
 
Last edited:
In the equation X*T and Y haqe different dimensions. Should it be F=X*T1 + Y*T2, where T1 is the button time and T2 is the overvoltage time? Let us assume that T1 is 2 hours and T2 4 hours.
The idle current could be minimized by using another unit of the inverter, which activates the low-voltage regulator only when the fan is running. I will modify the schematics later.
 
With respect to T2, are you thinking that when the battery reaches full charge, you trigger a "timer" that runs for a fixed period, just like T1. If fact, since the period of T1 = period of T2, why not just have a single timer T. Now the condition for the fan F to be running is:

F = X*T (never run the Fan if V(bat) <= 11.5V)

The new, separate expression for what triggers the timer T, is

Tin = P + Y (PushButton is pushed or V(bat)>14.2V)

This begs the question. If the Fan is running, and the sun is shining to produce best panel output, is the solar panel current higher than the Fan current? If so, we still need Y to turn on a switch that shunts current from the battery. If the Fan current exceeds the max solar panel current, then the Fan load is sufficient to prevent the battery voltage from exceeding 14.2V?
 
I realized that my T1 and T2 values are useles, because I gave the total daily times of the fan running.
In my present circuit the fan is running as long time as the switch is closed (lid is open), and after opening the switch (closing the lid), still for a short delay. In the case of overvoltage, it runs first for the delay time, and starts again, if the overvoltage still persists. But, I can estimate that one session lasts about 12 minutes and I can change the circuit so that there is a fixed fan running time when the switch is closed or overvoltage found. This way T1 = T2 = 12 min.

The panel maximuim current is 280 mA. It propably is an overoptimistic value. I think the true maximum current I can get is 20 % smaller, about 225 mA. This value, however still exceeds the fan current, which I measured to be 120 mA at 12 V. At the maximum power of the panel the difference, 105 mA, must be shunted when the battery is fully loaded. The lead gel battery I will buy has the capacity of 1.3 Ah., which means about 10 h use (50 sessions) without charging.

The fan current at different voltages is:
12V 120 mA, 9 V 102 mA, 7.5 V 66 mA, 6 V 54 ma, 4.5 V 45 ma, 3 V 3.6 mA(not running)
 
Last edited:
Or just let T1 = L be the lid open time directly, and T2 is not needed. We are back to:

F=X*L + Y, which in English is: run the fan if [the lid is up and V(bat) > 11.5V] or [V(bat>14.2V]

Just pure logic; no timing required...
 
This mimics my previous engineering experience. 1o days to get the specs, 2 days to do the engineering, and 20 days to document it...;)

I'll work on a design at night. It is 11am local here...
 
Ok, here is my hack at the NoSmell circuit.

Because this should have a low stand-by current, I changed the approach a little bit. I am now using a **broken link removed** "micropower 10V Voltage Reference" and a dual **broken link removed** micro-power voltage comparator(s). This makes for a design that draws less than 500uA during idle. Refer to the schematic.

U1 detects that the battery voltage is less than 11.6V (about as low as a SLA so be discharged), and is used to prevent the fan coming on.

U2 detects that the battery voltage is greater than 13.8V (the highest allowed float voltage for a 12V SLA), and starts the fan. I put hysteresis on this voltage detector, so the fan runs until the battery is discharged about 200mV. This turns the overcharge regulation into a multivibrator, where the fan turns on, runs for a while, then turns off while the battery voltage increases, then runs again.

Since the load caused by the fan is less than the output of the solar panel, I would be inclined to get a second fan, rather than just waste the excess power in R8. I am using an NFET as the fan switch because it is voltage controlled, so low power.

Now refer to the simulation. The time scale is arbitrary. Initially, the battery V(pos) is discharged. The panel output current is I(I1). As the sun begins to shine, the battery voltage increases as the battery charges. At ~14Ks, the battery voltage trips U2 (Y goes high), which turns on M1. The fan current is I(R9) . As the sun shines brightly until 30Ks, the fan cycles on/off, keeping the battery voltage between 13.8V and 13.6V. The hysteresis is controlled by R5, and you might have to play with it a bit to get the fan to run for a minute or so at a time.

I am using a voltage-controlled LTSpice switch to represent your cover-up switch. The signal V(up) is >0V t0 close the switch, so each time the red trace is ~1V, the cover is up. That event is asynchronous with the fan running due to excess charging. You can see that occur at 29Ks, where the switch causes the extra fan cycle. After the sun goes down, I keep cycling the switch, each time removing some charge from the battery. Finally, at ~78Ks, the battery voltage is down low enough that X (U1) goes low, so that prevents M1 from turning on thereafter, so future switch cycles do not run the fan.

What do you think?

NoSmell.gif
 
Hi namesake (Mikko is the Finnish version of Mikael, so propably we are namesakes).
The simulation shows that the circuit fulfils all the tasks I originally planned. I will build the circuit in the near future and test it, when bright spring days are here. Today the sun is still low and the sun radiation is absorbed or scattered by the atmosphere and the highest current I can get from the solar panel is about 60 mA.

I am just astonished how high-quality help I got from you on my weird smell-fan project. I started the project, because it was fun for me, not because it shoud be very important. Maybe that is true with you, too: you received the problem for fun, not becaus its importance. But, if you think how many millions of people live in this world without WC ...o_O
 
Even 60mA from the panel should recharge the battery completely each day, with an occasional fan cycle. You might put a winter/summer mode, where R8 is switched out or increased resistance during the winter...
 
Hi
I am still waiting for the key components for the solar regulator. I found from my box the following voltage bridge resistors: R1 = 86600, R2 = 91000 and R3 = 483000 (values measured By a multimeter). I was also thinking how to remove the dummy load across the fan terminals. How do you think, is it reasonable to use a third comparator to connect a dummy load as soon as the voltage rises over a critical level?
At the side of my battery they say that the standby voltage, when connected to the charger, should be 13.5-13.8 volts. If comparator 2 could connect the fan at 13.5 V and a third comparator would connect a dummy load at 13.8 V, the system would be fully automatic and the resistor between fan terminals could be removed.

For instance, if R2 (91 k) is divided to 2 resistors, R2a= 81 k and R2b = 10 k ,and the voltage connected to the comparator 2, the fan would start at about 13.5 V. The voltage at R2b - R3 joint could be connected a new comparator activating a dummy load at 13.8 V.

How do you think?
 
I thought about that, too. You already have the 10.00V reference. You just need a third comparator, and its associated voltage divider. Can you do something useful with the excess power, like run a second fan?
 
I have a more powerfull fan taking 200 mA, but its diameter is bigger, which makes the installation on a smaller tube more laborous. Maybe I use it and just increase the value of the by-pass resistor
 
Hi again
I built the controller following your schematics. It worked fine, but then I connected the battery terminals wrong way and the comparator blowed up! :arghh: I decided to put a protection diode to the Vdd pin of the comparator. In my drawer I had another comparator LM2093, which is pin-compatible with the original LT1017. I used it by adding 470 K pull-up resistors on the outputs. The modified circuit works quite OK, but the stand-by current is considerably higher, about 900 uA. I shall buy a new LT1017.
In the original circuit the overcharge of the battery was prevented by using a resistor across the terminals of the fan. The problem with this solution is that about 200 mA is wasted every time when the fan is running. In the adjoined circuit I have tried to solve this problem. I combined your schematics in the present thread and the one in the thread “Improve this solar controller.” In addition, I added a transistor switch which connects the over-voltage protection only when it is needed.
https://1drv.ms/1CD5KCs
U1 and U2 work as described in the present thread. The over-voltage circuit t6hat I added can be seen on the right side of the schematics. It is adjusted (R3) to consume the solar panel current, when the battery voltage is 13.8 V or higher. This over-voltage circuit is switched on by FET Q3 only, when pin 7 of U2 is high, i.e. when the battery voltage is 13.5 V or higher and the fan is running. Now, if the fan cannot consume all the current from the solar panel, the battery voltage rises. If the voltage reaches 13.8 V, the home-made Darlington pair consumes the extra energy. The FET switch Q3 minimizes the standby current by disconnecting the over-voltage circuit (blue frame) as soon as the battery voltage is less than 13.5 V.
 
Last edited:
Glad that it is working. Wish I had a buck for every chip I've ruined slipping with a clip lead...

Using this approach, I am building a pumping system to circulate solar-heated water through my wife's greenhouse. I am using a 1A wall-wart to charge a lead-acid automotive battery. The water pump draws ~5A, which means that on average, the pump will run about 20% of the time to use up the 1A. The pump will turn on when the battery voltage reaches ~14.2V, and due to hysteresis around the TL431, the Pump will turn off at ~13.2V. I am effectively building an astable multivibrator where the battery itself is the timing network...

Ever get to Arizona?
 
Well, Arizona is a little bit exotic place, if you look it from here. The closest place I have visited has been Big Sur, California. But anyhow, I will continue to follow the discussions on this forum, and learn...
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top