Return to the thumbnail page Display/hide file information See previous file See next file

LED conversion

LED conversion

Click to view full size image

Just some more fun. This GLS lamp has been converted to LED running on 12V. I needed to keep the GLS shape and run it on a low voltage and not too bright. This worked ok with warm-white LEDs, it is quite an acceptable conversion.

giessler2.jpg sodium_flash.JPG led_GLS.jpg candela.jpg

File information

File information

Download: Download this File
Filename:led_GLS.jpg
Album name:tuopeek / misc
Keywords:Miscellaneous
File Size:70 KB
Date added:Apr 30, 2013
Dimensions:800 x 861 pixels
Displayed:240 times
Date Time:2013:04:26 19:02:12
DateTime Original:2013:04:26 06:15:29
Exposure Bias:0 EV
Exposure Time:1/52 sec
FNumber:f 4.2
Flash:No Flash
Focal length:9.1 mm
ISO:200
Make:FUJIFILM
Model:FinePix S3400
Software:Adobe Photoshop Elements 2.0
White Balance:0
URL:https://www.lighting-gallery.net/gallery/displayimage.php?pos=-79026
Favorites:Add to Favorites
Comments
Medved
Hero Member
*****
Offline

Gender: Male
Posts: 4050
View Gallery

View Profile Personal Message (Offline)
Apr 30, 2013 at 01:01 PM Author: Medved
What ballast do you use there? 4 chips seems to me too much to leave any voltage room for the voltage dropping ballast, for boosting converter the voltage is way too close to the input,...

No more selfballasted c***

rjluna2
Sr. Member
****
Offline

Gender: Male
Posts: 264
View Gallery

Robert


GoL
View Profile Personal Message (Offline)
Apr 30, 2013 at 01:18 PM Author: rjluna2
Cool

Pretty, please no more Chinese failure.

Ash
Hero Member
*****
Offline

Posts: 3560
View Gallery


View Profile Personal Message (Offline)
Apr 30, 2013 at 01:21 PM Author: Ash
Cool indeed

The range between 12-13v would cover exactly the "dim" to "bright and burning up" range for the LEDs

So i'd guess 1 of 3 options :

Voltage match closely to what is needed without any resistors, but on a slim tolerance and it would easily go unstable with the slightest change in voltage or temperature

There is higher voltage and resistor (possibly parasitic resistor), and 12v is measured across LEDs

The step up converter is used, maybe it is running on eg 5v input
tuopeek
Newbie
*
Offline

Posts: 5
View Gallery

GLS on Tesla Coil


tuopeek 111967450636623837217 tuopeek1 77334065@N05/
View Profile WWW Personal Message (Offline)
Apr 30, 2013 at 01:54 PM Author: tuopeek
Thanks, Yup, runs directly on 12V but you are right it is on the edge of safety for the LEDs It was an experiment that has worked quite well so I may put a DC-DC converter in as a regulator.
Ash
Hero Member
*****
Offline

Posts: 3560
View Gallery


View Profile Personal Message (Offline)
Apr 30, 2013 at 02:49 PM Author: Ash
LEDs need constant current. The "linear" (ie, not high-frequency-based) approach to that is run them on DC voltage either through a ballasting resistor, or through a current source. A current source can be made by using a LM317 for example, but also with lower loss components like mosfet etc

+vdc -----\/\/\----->---->---->---->---- earth

+vdc -----(-->)----->---->---->---->---- earth

The losses from a linar circuit are exactly the voltage across the resistor or the current source, multiplied by the current. To reduce losses, you down this voltage as much as possible. For example, use a DC-DC to get voltage of precisely 12.5v and drop just 0.5v on a resistor, instead of using unstabilized voltage of around 15v and drop 3v on a resistor. The more stable voltage you have, the less voltage you have to drop on the resistor or current source to still be able to keep the current within tolerance
tuopeek
Newbie
*
Offline

Posts: 5
View Gallery

GLS on Tesla Coil


tuopeek 111967450636623837217 tuopeek1 77334065@N05/
View Profile WWW Personal Message (Offline)
Apr 30, 2013 at 03:22 PM Author: tuopeek
Hi Ash, when I get round to it I will be keeping the volt drop as small as possible no point wasting energy in heat. This bulb is one of 5 in a fitting running in an off grid hut. Some protection will be required with the variations of solar charging.. Just now I watch the battery voltage and run the when it low enough to be safe then it only drops slowly from there
Ash
Hero Member
*****
Offline

Posts: 3560
View Gallery


View Profile Personal Message (Offline)
Apr 30, 2013 at 03:32 PM Author: Ash
Sounds like a fun project

As for powering them, you can use one common DC-DC with precise output voltage and then use pretty small resistor in each. Other option would be DC-DC to higher voltage (around 65v) and run all in series, but im unsure if you'd be able to get as precise voltage there (not precise voltage requires higher resistor value so higher losses)
Medved
Hero Member
*****
Offline

Gender: Male
Posts: 4050
View Gallery

View Profile Personal Message (Offline)
Apr 30, 2013 at 11:06 PM Author: Medved
If there is a DCDC, the most energy saving to use the DCDC regulation mechanism to stabilize the current.

And for the voltage headroom: 0.5V sounds quite low for 4 LED's. You have to count for LED to LED variations (~30..50mV per chip in series), thermal dependency (-4mV/K per chip, so with 4 chips and operating temperature at 75degC on the junction it mean 400mV difference from 25degC) and an accuracy of the input voltage (there mainly it's long term and temperature stability; 2% is a good one, so +/-0.24V). So on the end, the voltage difference with 4 LED's may vary over about 0.75V range, so for an active series regulator and 12V typically measured on the LED's you would need at least 12.6V typical (voltage room 0.25..0.75V), for a resistor, when you want to keep the current in +/-20%, you would need at least 14V nominal.

Normally for 12VDC, 3chips are connected in series with a resistor as a ballast. This give the best ratio between minimum brightness and an average power input: Lowering the voltage mean higher efficiency of the ballast, but yielding wider tolerances. And that mean for given specified minimum, you have to add more margin for the "typical" power (so the brightness tolerance of +/-30% mean the same as 30% lower efficiency). Higher voltage drop cause lower ballast efficiency, but you don't need as much design margin to meet the specified minimum output. And the "sweet spot" then lies around the 12V.

And as with LED's the majority lies on the temperature dependence, it would be correlated even over larger installation, so you can not count on the averaging effect as you could do with "traditional", so you have to really count with the minimum corner.

Of course, the ability of the active linear regulator to keep the current constant helps a lot, as you may keep the voltage room way narrower and still have no current variation, so you improve both on ballast efficiency, as well as the required headroom for the tolerances (the later being even more significant).

DCDC vs linear: The DCDC seems to be more efficient, but a linear regulator is way simpler. That mean the control part of the DCDC would consume way higher power. And that mean, than for the low power LED's, the DCDC efficiency would become way lower than the simple linear series regulator.

No more selfballasted c***

LegacyLighting
Jr. Member
**
Offline

Posts: 99
View Gallery

Blake


View Profile WWW Personal Message (Offline)
Dec 16, 2013 at 01:17 AM Author: LegacyLighting
Very clever - just a question too. How do you separate the envelope from the base without destroying the envelope?
AngryHorse
Sr. Member
****
Offline

Gender: Male
Posts: 350
View Gallery


View Profile Personal Message (Offline)
Dec 16, 2013 at 01:26 AM Author: AngryHorse
You can `crack` the lamp at its base if your careful Blake, with little taps all the way around

A cleaver and interesting idea!

"Beauty fades, dumb is forever".......Judge Judy Cheesy

LegacyLighting
Jr. Member
**
Offline

Posts: 99
View Gallery

Blake


View Profile WWW Personal Message (Offline)
Dec 16, 2013 at 01:57 AM Author: LegacyLighting
I must give it a try on an old bulb.
tuopeek
Newbie
*
Offline

Posts: 5
View Gallery

GLS on Tesla Coil


tuopeek 111967450636623837217 tuopeek1 77334065@N05/
View Profile WWW Personal Message (Offline)
Dec 16, 2013 at 03:28 AM Author: tuopeek
Thanks for the comments. Separating the base can be a bit hit or miss but I find an electrically heated hot wire around the glass often produces quite a clean break.
BG101
Full Member
***
Offline

Gender: Male
Posts: 165
View Gallery

EYE H80 Mercury Vapour


Brian TheTellyman
View Profile Personal Message (Offline)
Dec 16, 2013 at 05:32 AM Author: BG101
When I've opened lamps in the past I've simply broken the stem away after snapping the pinch seal to let the air in then opened up the hole piece by piece, leaving the outer part and bottom seal intact. Obviously this leaves less room to get an LED assembly in and you lose the original stem but the lamps I had were pearl or opaque anyway, but you can make an invisible join when you re-fit the cap I'm sure a bit of ingenuity with some stiff wire can open out a suitably formed LED string once inside the bulb.

A second bulb could be used as the stem donor


BG

Say NO to DICTATORSHIP in the form of bulb/tube/ballast bans !!

© 2005-2019 Lighting-Gallery.net | Powered by: Coppermine Photo Gallery