What voltage per LED for this lamp unit?

[NeilP]

Today I was given a outdoor halogen lamp style lamp, but it had either been converted to an LED board or bought like I had it

Like the one in the image below but with only 48 LED's in it.

I wonder if it had been converted, as it was a mess inside, I certainly hope Sealey did not produce them like I found it

RThe back of the board was marked up as 12 volt, but half of the tracks on the back had been cut and it had been re wired so there was 12 LED 's in series instead of the original design which put just 4 in series.

Inside was s small 1 inch by 1 inch circuit board, with a tiny rectifier in a 4 pin package about 5mm x 5mm. The whole lot wrapped in black insulating tape!!

Under no load I measured 66 volt on the output, so that makes 5.5 volts per LED as it was wired.

Now the original wiring put 12 volts across 4 LED's
That gave 12 strings of 4 in parallel.
When I put 12 volt across it all the LED's lit up, and the board took 0.2 amps

I never ran it up at the 66 volts from rectifier board, so did not see it working at that voltage, and seeing the shoddy way it was built, I did not fancy it..just in case it was too much.


So anyone know what I could likely run these LED's at per LED? is the 5 volt safe or is that likely to cook them in short time...Or maybe measuring the voltage of this little rectifier board unloaded was not the correct thing to do, under load it probably gave a voltage drop

 

[cor]

Hi Neil,
See C1? That 1uF capacitor is the main limiting component here, everything downstream is just passing (rectifying, consuming) the *current* that this capacitor allows through.
This means that you cannot really measure output voltage, especially not when unloaded, but rather the output current.
The varying voltage of the 240 AC input will continuously charge/discharge C1 up to the difference between the output voltage and the 240V input.
So, at 12V DC output you see almost the entire 240V AC across C1 and the current to charge it can be calculated, this is the current that the circuit behind the bridge rectifier will receive as DC.
I am guessing that with the 1uF capacitor the current will be approx 70mA (700V swing times 1 micro-Farad divided by 0.01 second period) so the brightness will be very low at this current and 12 parallel strings, because each LED will receive 6mA and the typical spec of LEDs is 20mA continuous.
Since typical white LEDs drop about 3.5V but should never get more than 4V, you can't just put 4 series LiPoly packs on them, they need a *current* controlled supply. The easiest way to limit current is with a resistor, most of these LED lights will use a DC/DC converter, but this cannot be used with the circuit that you already drew, because that citcuit is already a current source.
To get the light output at optimum, you need to measure the current through the LEDs and add a resistor that gives the best results. For example, if your LiPo packs are freshly charged (4.2V) then you can expect that you need to drop about 2.5V-3V to give each LED the max 20mA (240mA total). Do not go beyond 20mA per LED if you want the light to live long, LEDs will die quickly when overdriven with even 30mA. The 2.5V drop at 0.24A means 10 Ohms 1W resistor.
The way that the board was modified (4 strings of 12 LEDs in series) makes sense for the 1uF value, then each LED should get just under 20mA from the approx 70mA that C1 lets through.
I suspect that originally for the 12V variant, the capacitor was 3x bigger, for example 2.7uF.
NOTE that the whole board and also the LEDs are "live" when powered from 240V as there is no isolation in the circuit!

 

[NeilP]

I did wonder about measuring the voltage, across the rectifier unloaded.

To get the light output at optimum, you need to measure the current through the LEDs and add a resistor that gives the best results.

At what voltage should I measure the current? just as it is at the moment with 12volt across them or as I increase the voltage?
So put a pot in series..10k? and a higher voltage..and adjust the pot to get 20mA

 

[cor]

Yes, you can measure at the current 12V how much current is drawn. 0.2A means 200/12 = 17mA per LED, close to the max. You will not see much increase from going to 20mA.
If this light is not giving enough light then try to add a different light to the mix.
After all, even though LEDs are efficient, this light is only 48 LEDs x 20mA x 3.5V or a little over 3 Watts. Probably as bright as a 25W incandescent bulb.
One of the most useful devices for inspecting small objects (like SMT PC boards) is a loupe-lamp. This is a large magnifying glass on an articulating arm like a desk lamp, having a circular TL tube around the loupe to throw light from all directions on the object you are looking at.

 

[NeilP]

If this light is not giving enough light then try to add a different light to the mix.

Yes, I could do, but I was using this because I had it and it was free. It will do as it beats working by the light given off by the PC screen

 

[cor]

It will do as it beats working by the light given off by the PC screen

Wheel your bike inside and switch the headlight on.

 

[NeilP]

The GF has put a stop to that..too much mud and oil on the hall carpet