Driving 20+ relays from 12V? Boost Converter?

[nigelmercier]

I've made an LED lamp that I run from a nominal 12V (range 11.8V to 14.2V); there are over 20 LEDs in it. Because of the possible voltage variation I incorporated a 7812 regulator.

At first I used groups of 3 LEDs and a dropper to give 20mA, but the LEDs kept blowing, so I removed one LED from each chain and used a different dropper, but the LEDs still keep blowing.

I think I want to do this properly: the 7812 is not very efficient, and driving LEDs hard from a fixed voltage is not a good idea.

Can anyone suggest a suitable boost converter that might give me 60V+ at constant current of around 20mA?

 

[Hero999]

The chances are it's wrong.

For a start, the LM7812 needs an input voltage of at least 14V to regulate properly so unless you're no using a 15V supply then it won't work properly, the voltage will range from about 10V to 12V.

Please post a schematic.

 

[nigelmercier]

OK, I lied! It isn't a 7812, it's a 12V LDO regulator and most of the time the input is around 13.8V from a lead-acid battery. (I thought I was making it easy).


Schematic for proposed change:

[12V] > [60V/20mA] --[>|-- --[>|-- --[>|-- --[>|-- --[>|-- ..... --[>|-- --[>|-- --[>|-- --[>|-- [0V]

 

[Hero999]

The LEDs should NOT have blown when connected as a string of three with a forward current of 20mA.

Did each string have its own series resistor?

Is 20mA the absolute maximum rating or is it the recommended operating current?

If it's the absolute maximum rating, then you need to reduce the current. I would suggest going back to the 12V LDO regulator and reducing the current to 15mA per string and each string using its own resistor.

 

[nigelmercier]

The LEDs are rated at 30mA, 20mA is the recommended current. I have checked the current in each leg, they were all under 20mA. I don't like using the regulator and dropper resistors method, it wastes a lot of energy.

So, back to the subject of my post: can anyone suggest a suitable boost converter that might give me 60V+ at constant current of around 20mA?

 

[Gary B]

What color are the LEDs? I am interested in the forward voltage drop and the size of your current limiting resister. If the things are red, the forward drop is close to the PIV. Don’t take this wrong but, is it possible that you have the polarity wrong?

If everything else is as you say, then that, thermal damage from soldering and high frequency voltage spikes from somewhere are the only reasons I can think of that would cause the type of failure you describe. Short strings with its own resister from a 12v. to 14v. supply should work fine. If your power source has a battery charger attached, bypass the input with both a large electrolytic and a small ceramic cap just to make sure that any voltage spikes never get to the diodes.

I agree about the waste of power in a resister if it is not selected properly. Therefore, the voltage drop across the resister should be as low as possible. Example, if the forward drop of the LED is 2 volts, then each string should be 5 LEDs. If you use a 600 ohm resister, that would provide a forward current of 20ma @ 12 volts 23ma @ 14 volts and about 1/3 watt of power loss per string. Increasing the voltage and the number of LEDs in the string will reduce the total power loss in the resister if regulation can be maintained so it is still dropping only 2 to 4 volts. That means as you double the supply voltage you have to halve the regulation tolerance to save anything. I wonder, is it actually worth it in the long run?

 

[Hero999]

My previous contained a typo which I've corrected: the LEDs should not have blown with a current of 20mA.

The fact that some of the LEDs have blown suggests you have a quality problem with the LEDs, which I suggest you solve before using a constant current driver.

There are plenty of ICs, have a look on the National Semiconductor or Linear Technology websites.
**broken link removed**
Linear Technology - Step-Up (Boost) LED Drivers

 

[nigelmercier]

Yes, it is possible that voltage spikes cause the problem, although there is a fairly large C, and a couple of small ones as well.

Please can we move away from problems with the existing circuit, I want to start anew.

 

[kinarfi]

How simple ------ complex do you want to go, complex being a handful of parts, I made an oscillator as shown, don't know if it will give the power wanted, you may need to change components, but the lm317 connected as shown will limit the current to 20 ma.

 

[nigelmercier]

How simple ------ complex do you want to go, complex being a handful of parts, I made an oscillator as shown, don't know if it will give the power wanted, you may need to change components, but the lm317 connected as shown will limit the current to 20 ma.

Thank you very much for that, certainly food for thought. I was rather thinking of a single boost converter IC with a few added components though.

 

[nigelmercier]

My previous contained a typo which I've corrected: the LEDs should not have blown with a current of 20mA.

I figured

Thanks for the links, will take a look.