3V led protection etc

T

tim_ber

Well-Known Member
Joined
12 May 2009
Messages
1,039
Visit site
I've read this thread:

http://www.ybw.com/forums/showthread.php?t=221719

I want to:
(1) use single white 3V leds from 12V (12V to 14 Volt fluctuation I guess)
(2) maybe 4 white 3.1V leds from 12V (12V to 14 Volt fluctuation I guess).


One website said to use the 4 leds in series for use (2) - that sounded odd. If they were 3V leds, the website said the 12V drop across all 4 in series would be OK. But most white LEDs are 3.1 V, so the 4 in series doesn't seem a great idea?

What do you think of that to start with?

But of course the Voltage spikes must be taken into consideration.

The thread above mentioned 2 methods of delivering a regular 12V as far as I can tell:

(a) LM2940CT-12 with a capacitor
(b) a transient suppressor (what does that do?) and a 'resettable' fuse

what is the difference in method (a) and method (b)?

Even if a regular 12V supply was established, to use a single white 3V LED at 30mA, the next thing is to stick in a 330 ohm resistor isn't it.

Is any of what I written there correct?

Is there a better method?

PS, I've always used LEDs in parallel before.

thanks in advance.
 
If you buy "12 volt" LEDs from Maplin I think you will find they simply have aresistor in series.

In fact Maplin's catalogue gives the formula R= (Vs-Vf)/If, to calculate the value of a series resistor, where Vs is the supply voltage, Vf is the forward voltage drop and If is the forward current
 
tim_ber said:
But most white LEDs are 3.1 V, so the 4 in series doesn't seem a great idea? ...
Click to expand...

It isn't. You can't feed LEDs without some form of current limit.

Using LEDs with a dropper resistor is always going to be a compromise. The maximum number of 3V LEDs you can reliable power from a nominal 12V is three, and even then there is a danger they will get burnt out at higher voltages. With fewer LEDs there is less chance of this, but you lose more power in the resistor.

There are lots of calculators for the LED series resistor, such as this. If you choose 14.4V as the voltage, and 20% below the maximum LED current, you should be OK, but the LEDs will not be as bright at 12V.

The best way to drive LEDs is with a Buck-Boost Constant-Current LED driver, but the additional circuitry required makes this rather complex; at least it was when I last looked into this a couple of years ago. If anyone knows a simple circuit, please let us know.
 
nigelmercier said:
If anyone knows a simple circuit, please let us know.
Click to expand...
There are some good ideas here:
http://www.instructables.com/id/Super-simple-high-power-LED-driver/

The above link is to a driver for one of the higher power LEDs, but I think you can some ideas for 5mm leds

For most projects I find the high power 1w or 3w LEDs are generally far superior. They have a much greater spread of light and one of these high powered leds will replace 20 or more 5mm ones. They are quite efficient at lower brightnesses if you don't require much light.
 
Most white LEDs that I have bought are rated at 3.4 volts. I use 4 in series but never have them on when the engine is running
 
nigelmercier said:
...... If anyone knows a simple circuit, please let us know.
Click to expand...

The simplest circuit is a resistor.
The next level of sophisitication is an analogue current source or sink, 3 resistors and a transistor. This has the advantage of not generating interference.
You can improve on it with a zener diode.

http://en.wikipedia.org/w/index.php?title=File:Const_cur_src_111.svg&page=1

The most efficient will be some sort of switching regulator, but this implies the possibility of creating interference.

In practice, running LEDs at a bit below max current is OK, your eye won't notice all the lights getting 5 or 10% brighter when you start the engine. Unlike incandescent bulbs they don't change colour very much.
 
nigelmercier said:
Well yes. I meant a simple Buck-Boost Constant-Current LED driver.
Click to expand...

That's a bit oxymoronic really.
But you could try the Maxim IC catalogue.
Or Linear Technology, National etc etc.
Or maybe start from the RS/Farnell/digikey dogalogues.

Most commercial circuits will be a suitable IC, and inductor and a resistor or two.
Then you just need the suppression filtering to stop it killing the VHF, Navtex etc etc.

A current source can be easily made to give 5% regulation and say 80% efficiency at minimum volts. Given that that gives you a big saving on normal bulbs, it gets questionable whether its really worth chasing the 90+% efficiency of an active converter, especially as most of the efficiency improvement will only be evident when you are charging the batteries anyway.

http://www.maximintegrated.com/datasheet/index.mvp/id/5608
 
Thanks for all advice.

I like electronics, but am very much at the beginner, bottom of the learning curve.

But I have also made some cool stuff mostly owing to the knowledge you guys have kindly imparted.

The instructables was a great link, thank you - just my level.

Some of what you guys said whizzed over my head.

I've used LMxxx thingy bobs such as the LM2940CT-12 before to knock 12V down to 5V and I know there is another type of LMxxxx thingy bob that can be used to output a voltage other than 5V if a resistor is placed in the correct place and on the correct leg.

BUT What is a transient suppressor? Like the one in the tread
http://www.ybw.com/forums/showthread.php?t=221719

that I mentioned in my OP that looks like a diode but isn't? What does it do? I bought some of the diode looking suppressors a long time back and have only just got around to start using them. I also have a load of the LMxxx thingy bobs. Need to get some heat sinks now.

Google didn't help: "A transient suppressor for 12V systems can be quite useful, especially in automotive/marine environments which have a high degree of short transients and can exhibit load dumps with peak voltages in excess of 100V. This is of course a very nasty environment for electronics, so making a transient suppressor for sensitive/cheap electronics is a good idea.
The LT4356 is a really nice "

That thread
http://www.ybw.com/forums/showthread.php?t=221719
gave 2 solutions, the transient suppressor and the LMxxxT thingy bob.

What is the difference between these 2 solutions?

Thanks
 
Last edited:
A transient suppressor is a device often closely related to a Zener diode, which conducts if the voltage across it exceeds a certain value.
For a 12v system, you might put a 20V transorb or similar across the rails close to sensitive equipment, this will remove any big spikes from the power rails.

The LM series covers a range of voltage regulators originally made by National.
The LM317 is classic linear regulator which takes in a higher voltage than the output and basically dissipates the excess as heat.
There are also switching regulators which use inductors to reduce the voltage more efficiently.

But for LED's you want a controlled current not voltage, because the voltage across them varies slowly with the power they are using, and varies with temperature. Controlling the current is stable and will work over a range of individual parts and across the temperature range. Some of these parts will draw more current from a fixed voltage as they warm up. So they get hotter still. Hence connecting in parallel is a bad idea as the hotter of the two will take more current!
 
'just use a resistor' or put 4 leds in series so that the 12v nominal system voltage drops across the summ of them, is the whole reason LED's get a bum rap.

In a vessel system with a high power alternator circuit and potentially a bunch of solenoids and motors ( autopilot, gas cutoff, bow thruster, anchor solenoid, engine starter etc ) there could be spikes of as much as 600v occasionally.

LED's are semiconductor devices ( like transistors or IC's ) just glueing them into circuits willy nilly is bound to get very variable results.

Trnsient suppression and current controlled drivers are the only reliable way to go.
 
LEDs

If an LED is running with a simple resistor current limit then I would suggest that it is very immune to short term voltage spikes. Indeed it is common to drive LEDs grossly over current for short time periods. The average of course must remain within LED limitations. I suspect the spike protection is needed for protection of the current regulator semi conductors. Higher 14+ voltages of system when on charge must be protected against.
As has been said and must be reiterated you can not run 4 x 3v leads in series off 12v without current limiting. Unlike an incandescant lamp or resistor an LED draws no current until voltage rises to its required voltage ie 2 to 3 volts. At that voltage current will flow with no limit usually releasing the smoke from the LED. A relatively small resistor in series will reduce the voltage at the LEDs as the current rises so limiting current to that required of the LEDs. This can be fine for a stable 12v supply. Better a 12.5v supply and .5 volt dropped across the resistor.
However it is not practical to use a series voltage regulator on a 12v battery supply system to hopefully give 12v stable as the series regulator will drop several volts. So best you can do is a 9v regulator and be satisfied with 2 or 3 LEDs in series.
Using the given formulars for calculating resistor values and resultant currents you can see that one LED with a series resistor will have just a small current increase from 11v to 14v supply. However you lose a lot of efficiency. 2 LED in series with a resistor gives you 2 LED so double the light at no extra current but more variation of LED current from 11v to 14v supply. 3 LED even more efficiency but more concerns about current variation. You have a dim light at 11v and or risk smoking the LEDs at 14v.
My little boat runs on solar charging so no lights on when battery is being charged. I use Nicad battery so voltage is fairly stable at 12v so I can run 3 LED in series happily. I am happy with resistive current control.
I think however that you might as well buy LED lamp with inbuilt regulator if you are concerned with high voltage supply although they can be built.
good luck olewill
 
You must log in or register to reply here.

Other threads that may be of interest

Y
The Seafarer 700 and NASA clipper sounders saga
Replies
10
Views
436
Dellquay13
Dellquay13

Share this page

Top