Solar charger problem.

al98 · Sep 16, 2008

Hi,

I am trying to make my own solar powered garden light using a 4.5v 60mA solar panel incorporating a comparator with hystereis. The circuit works well, as the light level falls to a certain threshold the LED turns on. The solar panel charges the battery as well as acts as a sensor to turn the LED on.

My problem is as the battery slowly decreases the turn on time of the comparator drifts. I was thinking of using a zener diode to establish a constant reference voltage to the input of the comparator but since the desired turn on time of the solar panel is so low at around 0.15v a zener cant be used to establish a threshold.

I don’t know how clear this is but can anyone suggest anything?
I have included a attached drawing.

Thanks.

ericgibbs · Sep 16, 2008

al98 said:
Hi,

I am trying to make my own solar powered garden light using a 4.5v 60mA solar panel incorporating a comparator with hystereis. The circuit works well, as the light level falls to a certain threshold the LED turns on. The solar panel charges the battery as well as acts as a sensor to turn the LED on.

My problem is as the battery slowly decreases the turn on time of the comparator drifts. I was thinking of using a zener diode to establish a constant reference voltage to the input of the comparator but since the desired turn on time of the solar panel is so low at around 0.15v a zener cant be used to establish a threshold.

I don’t know how clear this is but can anyone suggest anything?
I have included a attached drawing.

Thanks.

Hi, Its possible to split R1 into two resistors.
At the junction of this split you could use a 1N4148 forward biassed diode to give about 0.65V.
Select R1A and R1B to give about 2mA thru the diode and then adjust R2 to that you get the Vref 0.12V

Do you follow OK.?

EDIT: see this dwg.

al98 · Sep 16, 2008

Do you mean like what i have attached?

Thanks.

ericgibbs · Sep 16, 2008

al98 said:
Do you mean like what i have attached?

Thanks.

hi,
No,
Look at my dwg changes made to your original dwg.

EDIT: your diode is the wrong way around.

ericgibbs · Sep 16, 2008

hi,
Added some values, check the calculations.

al98 · Sep 16, 2008

Ok great, il give it a try and get back.

Thanks.

ericgibbs · Sep 16, 2008

al98 said:
Ok great, il give it a try and get back.

Thanks.

hi al,
If you have a 1K pot, make that R1b [530R] adjustable, it will give you more control. OK.

audioguru · Sep 16, 2008

An LM10 has only 8 pins. It is an an accurate low voltage (200mV) adjustable voltage reference plus an opamp. It works when its supply is as low as 1.1V and its operating current is low. Its output current is up to 20mA.
It is perfect for your project.

al98 · Sep 17, 2008

I tried the circuit. But unfortunately the LED does not fully saturate instead it gradually turns on im not sure but im guessing it has to do with the current through the diode not being proportional when the comparator switches.

The LM10 would be great but I need a very low cost circuit to use along side the 4.5V panel. I probably could use a typical solar garden light circuit with a single 1.2v battery and use a few series diodes from the panel to the battery to inefficiently step the voltage down a couple volts. I treid building a couple of these circuits from the net and found the circuit to consume around 45mA with the LED consuming around 3mA!

Do you guys have any more suggestion maybe different circuit ideas?

Thank you for your help.

ericgibbs · Sep 17, 2008

hi al,
Can you post your full circuit with component values.? [including the comp]
The circuit looks OK, its most likely a value that needs tweaking.

So that I can see why the comp is not switching fast.

al98 · Sep 17, 2008

No problems please see attached.

al98 · Sep 17, 2008

Sorry i forgot the comparator its a LM393

ericgibbs · Sep 17, 2008

al98 said:
Sorry i forgot the comparator its a LM393

hi,
I'll make one up on a project board, the LM393 is a good comp.
Let you know.

Torben · Sep 17, 2008

You might find something on Dick Cappel's site: http://www.cappels.org/dproj/ledpage/leddrv.htm#Solar_Powered_Garden_Light

I built one of those and it worked great in simple tests. Never put one out in the garden for a year, though.

I know I've seen more complex circuits based on the same general idea around but I can't remember where. I'll see if I can dig up the links.

Torben

al98 · Sep 17, 2008

OK that would be great Eric.

Thanks for your help Torben i would like to use a 4.5v panel and i was thinking i could use a simple comparator circuit without boost.

Torben · Sep 17, 2008

al98 said:
OK that would be great Eric.

Thanks for your help Torben i would like to use a 4.5v panel and i was thinking i could use a simple comparator circuit without boost.

Mmm. . .you're right--I missed one of your earlier posts. Sorry 'bout that. I mean, you could charge the 1.2V cell from the 4.5V panel but you might need to add more circuitry to make the charging work out.

Torben

ericgibbs · Sep 17, 2008

al98 said:
OK that would be great Eric.

Thanks for your help Torben i would like to use a 4.5v panel and i was thinking i could use a simple comparator circuit without boost.

hi al,
Problem solved.
Missing resistor from pin #1 to +V added a 3K9.
Remember the outputs of the LM393 are open collector.

The circuit works fine, used a 1K pot to simulate the solar panel.
Had to use +5V for test, no 2.4Vbty on the bench.
I used a 1K pot inplace of the 560R, set the Vref at 0.12V

You are operating the LM393 very close its lower limit at 2.4V.

EDIT: connect a 10K from the solar panel to the LM393 pin #2 [not a direct link]

Added extra components to the circuit.

al98 · Sep 17, 2008

Thank you again guys. It does work thank you very much. But it does still drift due to the battery volatge a bit not a lot though, i think thats because the output voltage of the comparator 1 to the input of comparotor 2 across the 3.9K resistor does change slightly and impacts the Von/Voff times off the hysteresis.

Vthreshold = (10K(Voutput - Vreferance)/1.01Meg) - Vreferance

Is there anyway of keeping Voutput constant so the threshold keeps constant? i have tried with diode configurations on the output but still no luck.

Also i forgot to say as the batteries charge they end up having a potential of 4.0V (for 2x1.2v cells) dont no how safe this is.

ericgibbs · Sep 17, 2008

al98 said:
Thank you again guys. It does work thank you very much. But it does still drift due to the battery volatge a bit not a lot though, i think thats because the output voltage of the comparator 1 to the input of comparotor 2 across the 3.9K resistor does change slightly and impacts the Von/Voff times off the hysteresis.

Vthreshold = (10K(Voutput - Vreferance)/1.01Meg) - Vreferance

Is there anyway of keeping Voutput constant so the threshold keeps constant? i have tried with diode configurations on the output but still no luck.

Also i forgot to say as the batteries charge they end up having a potential of 4.0V (for 2x1.2v cells) dont no how safe this is.

hi,
The output voltage of comp1 will effect the switching point, its designed that way.
Its positive feedback, thats required to give the comp1 a fast switch action.
Also the change in threshold is to ensure that comp1 output dosnt 'dither' at the switch over point.
Its normal for that type of circuit.
[Its the 1M0 thats the +ve feedback path]

You havnt said which type of battery you are using.?
But 4V does sound too high.

al98 · Sep 17, 2008

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.

Solar charger problem.

New Member

Attachments

Well-Known Member

Attachments

New Member

Attachments

Well-Known Member

Well-Known Member

Attachments

New Member

Well-Known Member

Well-Known Member

New Member

Well-Known Member

New Member

Attachments

New Member

Well-Known Member

Well-Known Member

New Member

Well-Known Member

Well-Known Member

Attachments

New Member

Well-Known Member

New Member

Similar threads

Privacy & Transparency

Privacy & Transparency