View Full Version : LDR for solar light

01-12-2009, 06:57 AM
For christmas I got given some expensive Solar lights (two LED models $40 each which we got for $20 as they were old old stock), they worked well until on new years eve we got hit with a day where the air temperature hit 41C and since then they have worked erratically...

I have pulled one apart and it seems the light dependant resistor is buggered...I snipped the wire leading to it and the lights sprung into life....

So I need a new resistor but am not sure what to get..Any ideas???

01-12-2009, 08:42 AM
Measure the good one in the dark and in the light. Let it sit for at least a minute in each condition before measuring. It's a cadmium sulphide LDR in nearly all those units. If you can't find a replacement part of the right spec then buy a set of the cheapest and steal one. It will probably work.

J Tiers
01-12-2009, 08:50 AM
Since yours is semi-bad, it may not be a good test, but.....

LDRs have two resistances.... their "dark" resistance, which is normally megohms, effectively open, and their "light" resistance, which may be anything from ohms to some intermediate value like a few tens of k-ohms.

Both have "adapted" values and "initial" values.... for instance what the resistance is when it has been dark for 24 hours, and what it is if you first turn off the light. The second case is going dark FROM the "light adapted" condition. The values are different.

In your specific case, there is a "switch-over" resistance, with an associated illumination level. it's actually sometimes tricky to get the right value, depending on the design.

Since you are turning ON a light when dark, it is the INITIAL DARK resistance you want, AND the voltage withstand.

Dark presumably megohms, and voltage at least 3x your mains voltage.

But, what I would do, not necessarily recommending it, because you have your hands in the AC...... First, take the bad one outside, and measure resistance lit up. bring in and measure when dark.

Then I would try some resistance values in place of it in the circuit, until I found the value that "just " turned it on. Then I would make sure that a replacement LDR was above that value when illuminated with the amount of light you want to have the light go "on" in.

The resistances should be similar for LDR measured and circuit requirements.

DIGI-KEY has a selection, not very large, but some.

01-12-2009, 01:05 PM
That sounds like an odd design. Why would a solar light need a separate LDR?

The solar cell itself should all the input the device needs in order to know when to run from battery power.

01-12-2009, 03:14 PM
Good one, Iwalker. I think the answer has to do with planned obsolesence. Those LDRs are one of few components made that aren't as 'solid' as most any other semiconductor. They are subject to having their operating parameters drift. I suspect that the one in question is not 'bad' per se, but has gone off spec too much to be useful anymore.

The designers could have easily taken the signal off the solar cell to tell when to turn the light on- on the other hand it's possible that it would have taken a few more other components, so maybe its cheaper anyway to use the LDR.

I wonder if the charging circuit uses a diode in series with the solar cell stack to prevent reverse discharge of the batteries through the cells- if not there may not be enough of a voltage drop when light is dim to trigger the leds to come on-

01-12-2009, 04:29 PM
Thanks everyone, I will look into trying to measure the resistance..

01-12-2009, 05:01 PM
That sounds like an odd design. Why would a solar light need a separate LDR?

The solar cell itself should all the input the device needs in order to know when to run from battery power.

The LDR could be there to ensure the light comes on when needed, not necessarily when it's dark enough to prevent the solar cell from generating. Also perhaps to adjust for changing ambient light conditions. It is also a good way to reduce flickering by giving the system some hysteresis.

01-19-2009, 07:02 AM
I finally got into town where the electronics supplier had a massive range of 1 type of LDR...I purchased one and have tested it out...It makes the light start up in darkish conditions but it turns on when conditions are not as dark as the original....Would putting a suitable resistor in series make it turn on when it is darker???

J Tiers
01-19-2009, 08:39 AM
larger selection:


select "in stock" box and 3 or 4 350VDC parts will show up, ranging from 1 megohm to 300K initial dark resistance.

An added resistor would likely do the reverse, since it seems the LDR you have is too HIGH in resistance..... dark it INCREASES resistance.

You can measure the one you have, by showing it to light, then putting it in the dark conditions that you know it turns on in.... measure resistance. You know that resistance will turn it on, so you want one that is LOWER in resistance by "some amount" (which we don't know).

I would, if it were me, pull out a resistance box and see what value turns the light on... then I'd have a reference for selecting parts.

Most in the Digi-key list are rated for resistance at 10 lux, which I confess does not produce a picture in my mind as to how light or dark that is. Evan will know immediately, I'd have to look it up.

You will have to "W.A.G." any deviations for lighter or darker than that.

01-19-2009, 09:56 AM
10 lux is ten candles in the same small room as you. :D

More specifically, one lux is the illumination provided by one candela (a candle) at a distance of about 1 meter.

J Tiers
01-19-2009, 10:24 PM
Sounds like THAT "unit" is begging to be "metrificated"...........

Anyhow, that's pretty bright..... odds are that spec level won't be much help, considering the result of the one he found.

It makes the light start up in darkish conditions but it turns on when conditions are not as dark as the original