|
|
13-11-2013, 12:31
|
#1
|
Registered User
Join Date: Sep 2008
Location: Camden, ME
Boat: A Thistle and a Hallberg-Rassy 36
Posts: 848
|
LED Lighting Using a Dropping Resistor
Most of us know that the ideal way to drive an LED is with a constant current regulator. This provides a stable current to the LED that will allow it to maintain a desired brightness over a range of voltages presented to the regulator. Unfortunately, a constant current regulator is also a switching regulator, which can be and often is a source of radio frequency interference.
It is possible to use a linear voltage regulator to provide drive to an LED without creating RFI, but it is less efficient and since it does not limit current, LEDs can self destruct via thermal runaway as they heat up.
A third and rather low tech means of driving LEDs is through a dropping resistor. This method is often dismissed as being inadequate given the large range of voltages that can be seen on a cruising boat, but I maintain that LEDs driven this way actually work rather well, and better than incandescent or halogen bulbs under the same voltage fluctuations. Over a range of current, an LED actually becomes more efficient as voltage/current falls, and luminosity falls much less quickly than it does with an incandescent bulb.
In order to test this, I put together a test circuit with three white 5mm LEDs in series with a 220 ohm dropping resistor. I applied a range of voltages to this circuit with a BK1743 variable voltage power supply, measured circuit current with a Fluke 117 DMM, voltage drop over the resistor with a Fluke Scopemeter 123, and light output from the LEDs with an Extech Lux meter. This type of LED is often used in light fixtures, and is rated at a maximum of 30 mA. It is better to drive them well under this to improve luminous and ultimate lifetime, however. Higher power LEDs that are driven at 300 to over 1000 mA are often used too. The relative results of this experiment are still applicable to these higher power LEDs, but the value and wattage rating of the dropping resistor will necessarily be different in those applications.
If you would like to see the spreadsheet, I have it on Google Drive:
Dropping Resistor Current Limiting for LEDs
For a quick summary:
At 16.0 volts, total current is 28 mA, relative light output is 609, and power wasted in the dropping resistor is 38%
At 14.0 volts, total current is 21 mA, relative light output is 490, and power wasted in the dropping resistor is 32%
At 12.5 volts, total current is 15 mA, relative light output is 382, and power wasted in the dropping resistor is 26%
AT 10.0 volts, total current is 7 mA, relative light output is 186, and power wasted in the dropping resistor is 15%.
This is how I see the take-home: At 12.5 volts, you have a circuit efficiency of 74%, compared to a good constant current driver which might give 85%, maybe 90% at best. You are safely under-driving the LEDs, they will last a long time (if you also put an avalanche diode across to protect against spikes).
At 16.0 volts, you are still not over-driving the LEDs.
At 10.0 volts you are still getting useable light.
This is just to make the point that dropping resistors can be a viable approach to lighting when simplicity, cost, and/or radio quiet are important. Careful design can yield a robust solution that provides efficiency close to that of a good switching regulator without the RFI headaches.
Chip
|
|
|
13-11-2013, 13:34
|
#2
|
Moderator Emeritus
Join Date: Sep 2006
Posts: 4,663
|
Re: LED Lighting Using a Dropping Resistor
For what it's worth, a linear (not switching) constant-current regulator is pretty common. Bebi used these in their lamps. It is possible to fashion one with two transistors and and two resistors, although the temperature stability of that design is somewhat poor. You can also make one using an LM-317 voltage regulator and a single resistor, with good performance.
I sometimes use resistors to drive low-power LEDS, and as you say, the loss and regulation can be acceptable as long as you don't stack too many LEDs in series. It's the "thrown away" voltage across the resistor that delivers the regulation.
__________________
Paul Elliott, S/V VALIS - Pacific Seacraft 44 #16 - Friday Harbor, WA
www.sailvalis.com
|
|
|
13-11-2013, 13:42
|
#3
|
Registered User
Join Date: Jan 2012
Posts: 143
|
Re: LED Lighting Using a Dropping Resistor
Constant current regulators don't have to be switching type. Check the LM317 regulator data sheet. They show a simple, one-resistor constant-current circuit. As you pointed out, it's less efficient than a switching type, but a lot easier to implement.
|
|
|
13-11-2013, 14:17
|
#4
|
Registered User
Join Date: Sep 2008
Location: Camden, ME
Boat: A Thistle and a Hallberg-Rassy 36
Posts: 848
|
Re: LED Lighting Using a Dropping Resistor
I'm not seeing an application on the datasheet for the LM-317 that turns it into a constant current source, I would be interested to see the schematic. I can easily imagine using an LM-317 to give a constant voltage, then a following resistor in series to provide current limiting, but then you would lose a bit more efficiency to get something close to constant current.
How would you use an LM-317 to give a true constant current source?
Edit: Google is my friend.... I see Linear Technology has some application notes for this.
Thanks,
Chip
|
|
|
13-11-2013, 14:27
|
#5
|
Registered User
Join Date: Sep 2008
Location: Camden, ME
Boat: A Thistle and a Hallberg-Rassy 36
Posts: 848
|
Re: LED Lighting Using a Dropping Resistor
Looks like the Linear Technologies LT3092 will provide constant current from .5 mA to 200 mA with about a 1.2 Volt dropout. I wonder why this isn't used more often in lieu of switching regulators?
|
|
|
13-11-2013, 14:29
|
#6
|
Registered User
Join Date: Apr 2013
Location: So Cal
Boat: Catalina 387
Posts: 967
|
Re: LED Lighting Using a Dropping Resistor
Constant current source with LM317
Pretty easy to do... But you will end up needing around 3v over the LM317 and as previously mentioned, it''s not maximally efficient.
|
|
|
13-11-2013, 16:20
|
#7
|
Nearly an old salt
Join Date: Jun 2009
Location: Lefkas Marina ,Greece
Boat: Bavaria 36
Posts: 22,801
|
Quote:
Originally Posted by SoonerSailor
Most of us know that the ideal way to drive an LED is with a constant current regulator. This provides a stable current to the LED that will allow it to maintain a desired brightness over a range of voltages presented to the regulator. Unfortunately, a constant current regulator is also a switching regulator, which can be and often is a source of radio frequency interference.
It is possible to use a linear voltage regulator to provide drive to an LED without creating RFI, but it is less efficient and since it does not limit current, LEDs can self destruct via thermal runaway as they heat up.
A third and rather low tech means of driving LEDs is through a dropping resistor. This method is often dismissed as being inadequate given the large range of voltages that can be seen on a cruising boat, but I maintain that LEDs driven this way actually work rather well, and better than incandescent or halogen bulbs under the same voltage fluctuations. Over a range of current, an LED actually becomes more efficient as voltage/current falls, and luminosity falls much less quickly than it does with an incandescent bulb.
In order to test this, I put together a test circuit with three white 5mm LEDs in series with a 220 ohm dropping resistor. I applied a range of voltages to this circuit with a BK1743 variable voltage power supply, measured circuit current with a Fluke 117 DMM, voltage drop over the resistor with a Fluke Scopemeter 123, and light output from the LEDs with an Extech Lux meter. This type of LED is often used in light fixtures, and is rated at a maximum of 30 mA. It is better to drive them well under this to improve luminous and ultimate lifetime, however. Higher power LEDs that are driven at 300 to over 1000 mA are often used too. The relative results of this experiment are still applicable to these higher power LEDs, but the value and wattage rating of the dropping resistor will necessarily be different in those applications.
If you would like to see the spreadsheet, I have it on Google Drive:
Dropping Resistor Current Limiting for LEDs
For a quick summary:
At 16.0 volts, total current is 28 mA, relative light output is 609, and power wasted in the dropping resistor is 38%
At 14.0 volts, total current is 21 mA, relative light output is 490, and power wasted in the dropping resistor is 32%
At 12.5 volts, total current is 15 mA, relative light output is 382, and power wasted in the dropping resistor is 26%
AT 10.0 volts, total current is 7 mA, relative light output is 186, and power wasted in the dropping resistor is 15%.
This is how I see the take-home: At 12.5 volts, you have a circuit efficiency of 74%, compared to a good constant current driver which might give 85%, maybe 90% at best. You are safely under-driving the LEDs, they will last a long time (if you also put an avalanche diode across to protect against spikes).
At 16.0 volts, you are still not over-driving the LEDs.
At 10.0 volts you are still getting useable light.
This is just to make the point that dropping resistors can be a viable approach to lighting when simplicity, cost, and/or radio quiet are important. Careful design can yield a robust solution that provides efficiency close to that of a good switching regulator without the RFI headaches.
Chip
|
Constant current linear regulators are easy made. And represent the best way to drive a led. Dropping the voltage over a resistor is fine if the input voltage is stabilised otherwise its a very poor method that leads to limited led life, not to mention wide changes in luminosity as the battery voltage changes. . Furthermore it's as energy inefficient as a linear regulator.
Rf quiet switching is easy to achieve , it's just that it needs more filters hence more components and so then output filters are often dropped.
Note all linear regulators with fold back control can be rigged as constant current sources.
__________________
Interested in smart boat technology, networking and all things tech
|
|
|
13-11-2013, 16:22
|
#8
|
Nearly an old salt
Join Date: Jun 2009
Location: Lefkas Marina ,Greece
Boat: Bavaria 36
Posts: 22,801
|
Quote:
Originally Posted by jeepbluetj
|
Try the lt3085. Ldo regulator. Only needs about 300mv across the reg to work.
Dave
__________________
Interested in smart boat technology, networking and all things tech
|
|
|
13-11-2013, 19:31
|
#9
|
Registered User
Join Date: Mar 2006
Location: Winnipeg
Boat: None at this time
Posts: 8,462
|
Re: LED Lighting Using a Dropping Resistor
The LM317 certainly isn't very practical for a 12 v nominal system. You would need an LM317 voltage regulator to take the input voltage of the current regulator down to a more practical level.
|
|
|
13-11-2013, 20:21
|
#10
|
Moderator Emeritus
Join Date: Sep 2006
Posts: 4,663
|
Re: LED Lighting Using a Dropping Resistor
Quote:
Originally Posted by DeepFrz
The LM317 certainly isn't very practical for a 12 v nominal system. You would need an LM317 voltage regulator to take the input voltage of the current regulator down to a more practical level.
|
I submit that in a 12V system the LM317 is better used as a current regulator than as a voltage regulator. The dropout voltage is no worse when used as a current regulator. I see no need to throw in a voltage-regulator as a first stage -- the current regulator can do it all.
By the way, I only mentioned the LM317 because it is (or was) probably the most-used general purpose linear voltage regulator out there. In the applications we are considering there are definitely regulators better-suited for the task, but the LM317 is familiar to many people.
And if anyone is designing a regulator circuit, please consider the issues of transient voltages, reverse current flow from any output bypass capacitors during shutdown, accidental reverse-polarity connection, bypass capacitors to prevent oscillation, heat dissipation, and any of the dozen other ways that the real world can zap your circuit.
__________________
Paul Elliott, S/V VALIS - Pacific Seacraft 44 #16 - Friday Harbor, WA
www.sailvalis.com
|
|
|
13-11-2013, 20:41
|
#11
|
Registered User
Join Date: Jan 2011
Boat: Catalina
Posts: 25
|
Re: LED Lighting Using a Dropping Resistor
How about using an incandescent bulb as a dropping resistor? ~
|
|
|
13-11-2013, 20:58
|
#12
|
Registered User
Join Date: Jun 2009
Location: Elyse is in New Zealand
Boat: Amel Super Maramu 2000
Posts: 589
|
Re: LED Lighting Using a Dropping Resistor
We have a 24 volt house system and I have used small flat pack LM7812 regs mounted inside the lamp housing, to reduce the 24 to 12v for certain lamp fittings that are only available in 12v models. Works a treat.
|
|
|
13-11-2013, 21:12
|
#13
|
Registered User
Join Date: Jul 2013
Location: San Diego, CA
Posts: 6,619
|
Re: LED Lighting Using a Dropping Resistor
Why not just use LEDs to drop the voltage? Back before I could buy any LED configuration I needed from China on Ebay for $3 ea. (including shipping); I used to solder 4 LEDs in series for 12v systems. This would give 3v ea for 12v, up to 3.6v ea at 14.4v. I just never saw any reason to waste current in a resistor when it would be better used in more LEDs.
Your battery system shouldn't be any higher than 14.5v or so, or lower than 12.0v anyways, so there's no need to go as high as 16v or as low as 10v. It worked fine then, is there any reason why it wouldn't work now?
|
|
|
13-11-2013, 22:01
|
#14
|
Registered User
Join Date: Mar 2006
Location: Winnipeg
Boat: None at this time
Posts: 8,462
|
Re: LED Lighting Using a Dropping Resistor
Quote:
Paul Elliot said: I submit that in a 12V system the LM317 is better used as a current regulator than as a voltage regulator. The dropout voltage is no worse when used as a current regulator. I see no need to throw in a voltage-regulator as a first stage -- the current regulator can do it all.
|
I guess you didn't read the link showing the power drop across the LM317 when used as a current regulator. The bigger the voltage drop the higher the power dissipated. I suppose if you wire 3 or 4 LEDs in series after the output of the LM317 you could get away with it, but if you are using it to power 1 or 2 LEDs then you have a lot of power to dissipate.
|
|
|
13-11-2013, 22:04
|
#15
|
Moderator Emeritus
Join Date: Sep 2006
Posts: 4,663
|
Re: LED Lighting Using a Dropping Resistor
Quote:
Originally Posted by socaldmax
Why not just use LEDs to drop the voltage? [...]
|
Because the current through an LED rises exponentially once you reach the forward bias threshold, and this threshold varies with temperature. You *need* some form of current limiting unless you are powering the LEDs from a tightly-regulated and temperature-compensated power supply. The voltage from your boat battery varies way too much for this to work using normal LEDs.
Now if the LEDs you were stacking had built-in series resistors or something like that, then yes, you might be able to put them in series without further regulation.
__________________
Paul Elliott, S/V VALIS - Pacific Seacraft 44 #16 - Friday Harbor, WA
www.sailvalis.com
|
|
|
|
|
Thread Tools |
Search this Thread |
|
|
Display Modes |
Rate This Thread |
Linear Mode
|
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
|
Advertise Here
Recent Discussions |
|
|
|
|
|
|
|
|
|
|
|
|
Vendor Spotlight |
|
|
|
|