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

Will They Work? 2 Rusty Generic 14-15-20W Fluorescent Ballasts

Will They Work? 2 Rusty Generic 14-15-20W Fluorescent Ballasts

Click to view full size image

These were in a wall wart housing for an aquarium light, however, at some point, the wall wart got filled with water. I have not tested these things yet so who knows whether they will work or not.

20191019_2025245B15D.jpg 20191013_2050065B15D.jpg 20191007_1738055B15D.jpg IMG_20191007_1636425B15D.jpg

File information

File information

Download: Download this File
Filename:20191007_1738055B15D.jpg
Album name:High Intensity / Ballasts & Fixture Parts
Keywords:Gear
File Size:500 KB
Date added:Oct 07, 2019
Dimensions:2464 x 1848 pixels
Displayed:97 times
Date Time:2019:10:07 17:38:05
DateTime Original:2019:10:07 17:38:05
Exposure Bias:-2 EV
Exposure Time:1/281 sec
FNumber:f 2.2
Flash:Compulsory Flash
Focal length:3.519 mm
ISO:50
Make:LG Electronics
Model:LM-X220
White Balance:0
URL:https://www.lighting-gallery.net/gallery/displayimage.php?pos=-165619
Favorites:Add to Favorites
Comments
High Intensity
Full Member
***
Offline

Gender: Male
View Posts
View Gallery

EYE 175w MV


GoL
View Profile Personal Message (Offline)
Oct 07, 2019 at 06:59 PM Author: High Intensity
So i did a resistance test using an older GE ballast as a baseline...

GE Ballast, 26.2Ω

Generic 1, 23.4Ω

Generic 2, 23.3Ω

With this info, i assume it's safe to move on to lamp testing.

Old lighting is, and will always be, the best lighting.

xmaslightguy
Hero Member
*****
Online

Gender: Male
View Posts
View Gallery

zzz


GoL ATL
View Profile WWW Personal Message (Online)
Oct 07, 2019 at 08:48 PM Author: xmaslightguy
Wouldn't be surprised if they still work...
Just use a near EOL lamp (and ofcourse test outside )

Its All For The Christmas Lights...
And A Good Summer Thunderstorm!

High Intensity
Full Member
***
Offline

Gender: Male
View Posts
View Gallery

EYE 175w MV


GoL
View Profile Personal Message (Offline)
Oct 08, 2019 at 12:29 AM Author: High Intensity
Final results from lamp testing...

Lamp used is a Philips F20T12.

GE Ballast (Baseline) 290mA 21w.

Generic 1, 450mA 31w.

Generic 2 300mA 21w.

I was surprised to see ballast 2 perform as well as it did, however, i wonder if ballast 1 did so poorly due to lower inductance. Since it appears the resistance has very little to do with the current (the resistance of ballast 1 went up to 24Ω after testing).

Old lighting is, and will always be, the best lighting.

Ash
Hero Member
*****
Offline

View Posts
View Gallery


View Profile Personal Message (Offline)
Oct 08, 2019 at 12:58 AM Author: Ash
If the resistance changes like this thre is a bad connection. Most likely corroded solder joint between the external wires and coil wire hidden under the tape. But the resistance change on this scale have virtually no effect on the lamp current

The difference in current could be from change of the air gap (does one of the cores "play" or have part of it disintegrated ?) or a shorted turn in the ballast with the higher current. This is some 1.2x overdrive for an F20
High Intensity
Full Member
***
Offline

Gender: Male
View Posts
View Gallery

EYE 175w MV


GoL
View Profile Personal Message (Offline)
Oct 08, 2019 at 01:05 AM Author: High Intensity
One of the cores is rusted to the point were it's starting to flake apart. I don't think a shorted turn would cause the resistance to go up. What's the normal current for an F20?

Old lighting is, and will always be, the best lighting.

Ash
Hero Member
*****
Offline

View Posts
View Gallery


View Profile Personal Message (Offline)
Oct 08, 2019 at 01:31 AM Author: Ash
Little flaking probably won't change the current much. I meant another thing : The core E's and I's are inserted so that all E's are from one side and all I's from the other (unlike most transformers where sides are alternating with each layer). The center part of the E is a little short so it makes an air gap towards the I, that sets the ballast inductance. If they rust, they can push out and the gap will grow. If the support (weld or whatever) that holds the core together disintegrated they may not be held together well anymore and the gap will change too. The gap have the most significant effect on lamp current of all ballast parameters

A shorted turn will change the resistance down. If it's 1 turn you may not see the change on the multimeter, but it will have bigger effect on the ballast reactance (effectively turning it into a transformer with short on the secondary). This will increase the lamp current and cause localized overheating around that turn in the coil

A corroded connection on the solder joint (or just twisted wires) will change the resistance up, that is what i think might have made the resistance go up. but 0.6 Ohm won't make a difference in the lamp current. (The bad connection itself may continue degrading, which will make it heat up, until it overheats and starts melting isolation or fails open)

Proper F20 current is 370mA
High Intensity
Full Member
***
Offline

Gender: Male
View Posts
View Gallery

EYE 175w MV


GoL
View Profile Personal Message (Offline)
Oct 08, 2019 at 01:54 AM Author: High Intensity
After taking a closer look at ballast 1, it became obvious that the laminations are being 'pushed apart' by the rust.

So i put all three ballasts (the two generics and the GE i've been using as a baseline) on a cheap parts tester and the results gave me more questions than answers.

GE Ballast, 351mh(?)

Generic 1, 890mh

Generic 2, 1.15h

So as expected, ballast 2 has a higher inductance than ballast 1, but the GE ballast has an even lower inductance. Now there is a chance this is balanced out by the higher resistance (GE is 26Ω, Generic 1 is 24Ω), but if that were the case, then would that mean i could "fix" the current issue by putting a 1 or 2 ohm resistor is series with the ballast? (Likely no if i were to guess.) Though like i said, this was a cheap tester, so the numbers could be way off.

And i did try reconnecting the ballasts multiple times, but the results stayed the same.

Old lighting is, and will always be, the best lighting.

High Intensity
Full Member
***
Offline

Gender: Male
View Posts
View Gallery

EYE 175w MV


GoL
View Profile Personal Message (Offline)
Oct 08, 2019 at 02:59 AM Author: High Intensity
Welp, i 'accidentally' popped part of the core apart when trying to clean it. This thing seems to be friction fitted together.

Old lighting is, and will always be, the best lighting.

Ash
Hero Member
*****
Offline

View Posts
View Gallery


View Profile Personal Message (Offline)
Oct 08, 2019 at 06:00 AM Author: Ash
The ballast reactance is by far more significant than the wire resistance :

The voltage drop on the wire resistance and the lamp discharge voltage are approximately in the same phase. The lamp voltage drop for F20 is 57V, while the drop on 24 Ohm coil (at 0.37A) is 9V. If we recalculate the lamp into resistance, that will be 57V / 0.37A = 154 Ohm. (Since th voltage stays about the same for a wide range of currents, this resistance value will go down if the current goes up, this is what is called "negative resistance" but it really means "negative resistance change coefficient")

Then, the ballast reactance (~270 Ohm) and the sum "resistance" of the lamp discharge + coil wire (~154 Ohm + 24 Ohm = 178 Ohm) dont simply add together with the ballast reactance, but they add at 90deg phase (ie. impedance = sqrt( reactance^2 + resistance^2 ) ). At 90deg phase to the dominant part of the impedance (the 270 Ohm), even the entire 24 Ohm coil resistance have very little effect on the current (allthough it is accounted for in the ballast design)

The 270 Ohm is calculated as :

ballast reactance voltage drop : sqrt( 120V^2 - (57V+9V)^2 ) = 100V

ballast reactance : 100V / 0.37A = 270 Ohm

Possible guess why the so much different inductance values : If the gap in the generic ones is not uniform or not across the entire core cross section. Imagine that there is a small section of the core that touches directly without gap, and the rest do have the gap

When you measure it with a small current from LC meter, the entire flux goes through the ungapped part and the inductance is high

When you connect it in a lamp circuit, the flux is sufficient to saturate the ungapped part (so it itself starts behaving as an air gap), then the flux sees a gap and the inductance goes down

Theoretical calculation of the inductance :

L = reactance / (2 * pi * F) = 270 Ohm / (2 * 3.14 * 60Hz) = 0.71H

But this does not explain the reading on the GE ballast. My guess here is that the LC meter measures at high frequency (10..100 kHz) and the steel lamination core behaves more like a resistor than inductor at such frequencies due to eddy currents. (that is why dust or ferrites or other more special materials are used for high frequency cores)
Men of God
Full Member
***
Offline

Gender: Male
View Posts
View Gallery


jiachao.wei.71 chao_813975447 UChyTpXvlQ8ZCfBPP_lJjubg chao990613
View Profile Personal Message (Offline)
Oct 08, 2019 at 08:46 PM Author: Men of God
这阻值好像太小了,貌似有匝间短路,我随便说几个正常点的,都是日本制造的,如下:
藤井 100V 50Hz 15W 0.3A 35Ω
三菱 100V 60Hz 15W 0.3A 28Ω
东芝 100V 60Hz 15W 0.3A 31Ω

我不会英文,所以我用中文,请你们用翻译网站翻译我打的字!

I can't in English, so I in Chinese, please use the translation website to translate my words!

High Intensity
Full Member
***
Offline

Gender: Male
View Posts
View Gallery

EYE 175w MV


GoL
View Profile Personal Message (Offline)
Oct 13, 2019 at 03:34 AM Author: High Intensity
So i cleaned and rebuilt generic ballast 1 and now the reactance is way too high (about 5H according to the cheap LC meter).

It puts out barely enough current to run an F4T5.

I'll try re-rebuilding the core tomorrow.

Old lighting is, and will always be, the best lighting.

Bottled lightning
Newbie
*
Offline

View Posts
View Gallery

View Profile Personal Message (Offline)
Oct 13, 2019 at 04:57 AM Author: Bottled lightning
The gap between the E and I parts of the core determines the inductance.
High Intensity
Full Member
***
Offline

Gender: Male
View Posts
View Gallery

EYE 175w MV


GoL
View Profile Personal Message (Offline)
Oct 13, 2019 at 09:53 PM Author: High Intensity
Some modifications later and now GB1 runs F20's at 350mA.

Now i should try rebuilding GB2.

Old lighting is, and will always be, the best lighting.

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