Solar charger problem.

al98 said:

Mmm ok i sort of understand. Is there anyother way or type of circuit which is reliable?

I am using 2 x 1/3aaa Ni-mh batteries.

Click to expand...

hi al,
Which part of your circuit do you think is unreliable.?
What do you want to change.?

Attached is a graph, double the voltage on the graph when using two cells.

What colour is the LED.?

Im a bit consfused but it seems like the threshold cannot be kept constant over the change of battery voltage. From the graph the batteries should have a final voltage of around 3.2v, should i use a 3.3v zener across the supply?

The LED is Orange 2V. Does the output of comp 2 affect anything?

Do you know of any other circuits that would be better? I cant seem to find any on the net.

Thanks.

al98 said:

Im a bit consfused but it seems like the threshold cannot be kept constant over the change of battery voltage. From the graph the batteries should have a final voltage of around 3.2v, should i use a 3.3v zener across the supply?

The LED is Orange 2V. Does the output of comp 2 affect anything?

Do you know of any other circuits that would be better? I cant seem to find any on the net.

Thanks.

Click to expand...

hi,
Look at this dwg.
The +feedback resistor is now 2.2M, this will reduce the hysteresis.
Add the two diodes from the Solar cell[ they give approx 1.3V drop]
I dont follow what you mean by Does the output of comp 2 affect anything?

Tell me what you want the circuit to do and I could suggest other circuits.

For example.
how long do want the LED to stay ON.?
whats the specification of the solar cell and the LED [how many mA's]

Sorry ill start again i wasnt clear about it from the beginning.

I would like to create a solar garden light ciruit with 2x1.2v batteries and a 4.5v panel that turns on a led ats night and i was thinking of using a caomparator to do this as i tried using transistors previoulsy but couldnt get the circuit to wait a litle longer until it got darker.

With the circuit i have shown i have used hysetresis because i thiught the LED might flicker but it dosnt.

So ditching the original circuit with hysteresis and just using a simple circuit shown in the file i would just like the led to fully turn (not gradually) on at night at the same light level every time regardless of the voltage of the batteries then turn off when light appears.

Solar Panel 4.5V 60mA batteries 1/3aaa or 1/4aaa (120mA capacity)
LED 2V 10mA.

Does this help?

Sorry forgot attchement.

al98 said:

Sorry ill start again i wasnt clear about it from the beginning.

I would like to create a solar garden light ciruit with 2x1.2v batteries and a 4.5v panel that turns on a led ats night and i was thinking of using a caomparator to do this as i tried using transistors previoulsy but couldnt get the circuit to wait a litle longer until it got darker.

With the circuit i have shown i have used hysetresis because i thiught the LED might flicker but it dosnt.

So ditching the original circuit with hysteresis and just using a simple circuit shown in the file i would just like the led to fully turn (not gradually) on at night at the same light level every time regardless of the voltage of the batteries then turn off when light appears.

Solar Panel 4.5V 60mA batteries 1/3aaa or 1/4aaa (120mA capacity)
LED 2V 10mA.

Does this help?

Click to expand...

hi,
Seen the circuit.
You say you dont want any hysteresis and the LED to come on 'not gradually'.
What do you think is going to happen when the Solar cell voltage drops to and equals the 0.12V threshold...?

How are you testing the last circuit I posted.?

Now no hysteresis and for the LED to turn on not gradually (just on or off).

When the solar voltage equals the threshold it will gradully turn the led off or on. I know increasing the gain by connecting two comparators the flicker is less noticable but still exists over a shorter span.

With the last circuit i fully discharged the batteries. Charged them for 5 min checked the ammount of 'darkness' needed to turn the LED on then repeted this every 10min noticably seeing the ammount of darkness needed reduced each time the battery potential increased (this i cacnt fix).

al98 said:

Now no hysteresis and for the LED to turn on not gradually (just on or off).

When the solar voltage equals the threshold it will gradully turn the led off or on. I know increasing the gain by connecting two comparators the flicker is less noticable but still exists over a shorter span.

With the last circuit i fully discharged the batteries. Charged them for 5 min checked the ammount of 'darkness' needed to turn the LED on then repeted this every 10min noticably seeing the ammount of darkness needed reduced each time the battery potential increased (this i cacnt fix).

Click to expand...

So we agree that some level of hysteresis is required for the circuit to give a positive switching action to the LED.
If you increase the 1M0 to say 2.2M or 4.7M the hysteresis [dead band] will decrease.

When you fully discharged the batteries what voltage did they go down too.?

Did you measure the 0.12V reference voltage while doing the darkness test at different battery voltages.?

Let me know these details.?

What is your location.?

When you fully discharged the batteries what voltage did they go down too.?

Did you measure the 0.12V reference voltage while doing the darkness test at different battery voltages.?

Let me know these details.?

What is your location.?

Click to expand...

The batteries go down below 2v.

Yes the 0.12V referance voltage remains constant. What changes is the output voltage of the comparator 1.

Australia.

al98 said:

The batteries go down below 2v.

Yes the 0.12V referance voltage remains constant. What changes is the output voltage of the comparator 1.

Australia.

Click to expand...

hi,
Should have plenty of Sun in Oz.

I have been testing the circuit, the Vref changes less than 10mV from 3Vbty down to 2Vbty, so as you have found thats OK.

Can you tell me how the comp1 output changes, it should switch either low about +10mV or high about greater than +2V. ?

Have you removed the 1M0 ..??

Haha its nice and getting warmer!

Can you tell me how the comp1 output changes, it should switch either low about +10mV or high about greater than +2V. ?

From the diagram (if its correct) when c1 switches the - input of c2 becomes Vcc and therfore Vout changes according to Vcc. Everytime i put a voltmeter between V2 and ground i get Vcc when c1 switches on. This is what i think is causing the threshold to change??

Have you removed the 1M0 ..??

No the 1meg is still there.

ericgibbs said:

hi,
Look at this dwg.
The +feedback resistor is now 2.2M, this will reduce the hysteresis.
Add the two diodes from the Solar cell[ they give approx 1.3V drop]
I dont follow what you mean by Does the output of comp 2 affect anything?

Tell me what you want the circuit to do and I could suggest other circuits.

For example.
how long do want the LED to stay ON.?
whats the specification of the solar cell and the LED [how many mA's]

Click to expand...

Sorry to nag in on an other persons post but i just wanted to know(learning) 1)how did you set the value of the FB resistor(by trial and error?)
2)why there is the need for 2 comparators since one can compare the voltage wrt ot the batteries?

Wond3rboy said:

Sorry to nag in on an other persons post but i just wanted to know(learning) 1)how did you set the value of the FB resistor(by trial and error?)
2)why there is the need for 2 comparators since one can compare the voltage wrt ot the batteries?

Click to expand...

hi,
The amount of +feedback should be between 1mV and 10mV to ensure fast switching on slowly rising input voltages.
When a LM393 is configured as a NI comparator and it has a 10K source input, adding a 10M0 +feedback resistor will give about 10mV.

However, when used in some applications the user wants a wider hystersis in order to control the switching levels.
One such application would be a daylight detector switching a charger.
The user may decide for example that a 0.2V change in switching level between ON/OFF is required, so a lower value feedback resistor would be used in order to 'widen' the voltage band.

If you want to adjust the hysteresis, a simple way is to connect a pot say 50K from the output of the LM393 to 0V,
then connect the usual feedback resistor to the pot wiper.
To get the precise switching voltage points will require adjusting the pot and checking the switching limits.

With ref the second half of the OP's circuit for the LM393, it isnt required, but its not my design.
IIUC the OP is experimenting/learning electronics and is asking questions about his design, he is not asking for a fully design circuit.
If he did, most members would tell him to post his work to date.

If I can help with your design please ask away.

EDIT: added a pdf that explains hysteresis in a more mathematical way.