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

Osram dulux electronic 11W

Osram dulux electronic 11W

Click to view full size image

This is one of the first Osram electronic CFLs (date code AZ3). When I’ve powered it I’ve noted that it was really bright for a 11W lamp, so I’ve measured the power consumption and I’ve found that it drawn 22W!
There was also something rattling inside the ballast box, so I’ve opened it and I’ve found that half of the ferrite of the current limiting inductor (not so limiting without the ferrite!) was unglued and lying in the box!

Another curiosity in the lamp circuit is the preheating system, it uses a PTC resistor (like a lot of cfl), but the PTC is not placed in parallel with the resonance capacitor (Cs) between the lamp electrodes.
In the schematic on the photo you can see that the PTC is placed directly after the current limiter inductor L1, in series with two 47V zener diodes and then connected with two diodes on the the ground and positive power supply.
When the lamp is first powered up, the voltage is limited by the current path through the PTC under the ignition point of the tube. In the meantime the current flowing in the resonance capacitor warms the electrodes. When the PTC is warm enough and “switch off” the voltage is no more limited and rise until the tube is ignited. Once the lamp is running the voltage is kept low by the discharge on the tube, no more current is flowing through the PTC which can cool down for the next power cycle (similar behaviour as the SIDAC-PTC circuit described here).


Mini-Lynx.jpg osramdulux.jpg youtube_MAArf1G1QVA.jpg STARTE2MOSSCH.JPG

Light Information

Light Information

Manufacturer:Osram
Model Reference:Dulux electronic
Lamp
Lamp Type:CFL
Base:E27
Service Life:8000h
Fixture
Ballast Type:electronic
Electrical
Wattage:11W
Voltage:220-235V
Optical
Lumen Output:600
Color Temperature:2700k
Physical/Production
Factory Location:Germany
Fabrication Date:Datecode: AZ3
Application/Use:collection

File information

File information

Download: Download this File
Filename:osramdulux.jpg
Album name:Kappa7 / Electronic ballast circuits
Keywords:Lamps
File Size:433 KB
Date added:Feb 11, 2014
Dimensions:1600 x 1200 pixels
Displayed:448 times
Date Time:2014:02:11 20:25:44
DateTime Original:2014:02:11 18:31:47
Exposure Bias:0 EV
Exposure Time:1/50 sec
FNumber:f 2.65
Flash:No Flash
Focal length:3.97 mm
ISO:50
Make:SAMSUNG
Model:GT-I9100
Software:ACD Systems Digital Imaging
White Balance:0
URL:https://www.lighting-gallery.net/gallery/displayimage.php?pos=-90595
Favorites:Add to Favorites
Comments
Medved
Hero Member
*****
Online

Gender: Male
View Posts
View Gallery

View Profile Personal Message (Online)
Feb 11, 2014 at 02:14 PM Author: Medved
It is indeed quite unusual circuit. With this, I will be afraid, the Zeners will overheat before the PTC switch OFF.
By the way taking the idea, I will use there rather a SIDAC: It will limit the voltage as well, at the same time it will create extra current through the filament (each cycle discharge the 3.3nF capacitor from the ~200..250V back to ~150V), so the warmup will be faster (the extra current will pass the PTC as well, so the I^2*t integral dictating the amount of heat the filament need will stay the same). With the SIDAC, there will be no principal losses except the PTC, but within the PTC it is desired...

No more selfballasted c***

Kappa7
Jr. Member
**
Offline

Gender: Male
View Posts
View Gallery

View Profile Personal Message (Offline)
Feb 12, 2014 at 04:47 AM Author: Kappa7
Yes I've thought also that the losses in the zeners are the main problem in this circuit even if each zener take only one half of the power as they conduct alternatively. By the way they are rated 1.3W.
Yes the sidac would do a better job here. Indeed I've found in a more recently osram dulux el sensor (the ones with the dusk sensor) which have the classic PTC preheating with a series sidac. Probably they have added the sidac (it was not even expected on the pcb, they mounted it by cutting a pcb trace) because the dusk sensor was disturbed by the light of the lamp(the sensor hysteresis and the minimum on time was not enough) and during the sundown the lamp was cycling for some time like this: switch on for a few minutes, switch off for few seconds and then switch on again etc... so the preheating after a short power off was necessary to not reduce the life of this (expensive) lamp.
Medved
Hero Member
*****
Online

Gender: Male
View Posts
View Gallery

View Profile Personal Message (Online)
Feb 12, 2014 at 06:19 AM Author: Medved
Usuallythe dusk sensors are made sensitive only in quite narrow wavelength range in the IR, where the lamp does not radiate. Of course, when someone form the production floor switch to "cheaper" components without cross-checking with the design team, it frequently happen exactly what you describe...

By the way I've seen quite neat solution with LED's: The LED driver was intentionally made so, the LED flickered at the mains 100Hz (= no filter in the rectifier). And it was in the "dark" moment, when the signal from the light sensor was sampled and when there was no light, the LED was turned ON for the rest of the halfwave. It was implemented just by a phototransistor shunting the startup circuit of the flyback driver, so when the phototransistor sees light, it prevent the starting resistor from charging the low voltage controller suply, so the converter from starting, but once it was already started, the power from the auxiliary winding provided the needed supply till the end of the halfwave. And there the circuit shut down. And after the zero cross, the circuit started only when there was dark.
As a result, even when the LED shine directly on the photo sensor, the light went ON only when there was really dark and OFF, when there was enough light from other sources (sun,...)

No more selfballasted c***

Kappa7
Jr. Member
**
Offline

Gender: Male
View Posts
View Gallery

View Profile Personal Message (Offline)
Feb 12, 2014 at 11:59 AM Author: Kappa7
Indeed I remember from the trials that I've done (it was several years ago) the dusk sensor (a photodiode and a red filter on the ballast box) was only a little sensible to the fluorescent light and quite sensible to incandescent light.

Nice solution for the LED lamp, the flicker in this case has also a helpful use!
Lampka
Newbie
*
Offline

View Posts
View Gallery

Collector SL lamps and other lamps and lanterns.


View Profile Personal Message (Offline)
Sep 07, 2015 at 03:57 PM Author: Lampka
This version of Osram Dulux EL looks like Narva version of CFLs. I think, it had been made in Narva's factory. Code AZ3 means September 1991.

Collector SL lamps and other lamps and lanterns.

© 2005-2020 Lighting-Gallery.net | Powered by: Coppermine Photo Gallery