Previous in Forum: Voltage Drop Analysis Next in Forum: Energy Conservation By Star Connection
Associate

Join Date: Mar 2011
Posts: 52

# Design of Magnetic Ballast

11/18/2012 10:23 AM

Hi,

I was trying to design the magnetic ballast for 36W fluoresecent lamps design parametes 220V, 0.3 AMps, 60Hz, cosφ=0.54.

I want to know the equation to find the length of the ballast, in other words I want to know how is the length of the ballast related to the inducatance charactersitic of the ballast.

If there is any link to a technical document on this or an ebook then it will be very helpful.

Interested in this topic? By joining CR4 you can "subscribe" to

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, rate them!
Guru

Join Date: Oct 2008
Posts: 42290
#1

### Re: Design of magetic Ballast

11/18/2012 10:32 AM
Guru

Join Date: Oct 2007
Location: India
Posts: 1246
#2

### Re: Design of magnetic Ballast

11/18/2012 11:05 AM

There is no such equation, to find the length of a magnetic ballast.

Guru

Join Date: Oct 2008
Posts: 42290
#3

### Re: Design of magnetic Ballast

11/18/2012 11:17 AM

Ballast salesman to Shoeb, "how long do you want this ballast?"

Shoeb, "I'd like to keep it forever".

Associate

Join Date: Mar 2011
Posts: 52
#4

### Re: Design of magetic Ballast

11/18/2012 11:19 AM

this is what I did:

VA=220V * 0.3Amps= 66VA (since its similar to 1:1 trans it will be /2)

Core design Calculation

VA rating = 33VA;

Te (Turns/volt) = 6 (from the table, if possible its as per your table as well)

Bsat = 1 T (keeping flux density a const.) ;

φ = (1.44 * freq * Te)-1 = (1.44 * 60 * 6) = 1.93e-3 Wb ;

Net Iron Area= Ai = φm / Bm = 1.93e-3 / 1 = 1.93e-3 mm2 ;

Gross Iron Area = Ai / 0.95 = 2.03e-3mm2 (dboutful value) ;

Winding Calculation

No. of turns = VA * Te = 33*6=198 turns ( I am sure this value is incorrect) ;

We are using a copper wirer= 2.3A/mm2 over here ;

Area of conductor required = 0.3Amps / 2.3 = 0.13mm2 ;

Dia of wire = sqrt(area of Cu * 4/pi) = 0.4mm (practically it was 0.3mm dia enamelled wire) ;

Size of Lamination : "EI" type

All dimen. in (mm)

a=41.2; b=26.5; c=6.8

d=7.1; e=13.4; g=11.8

thickness=0.6

Area of window Aw=d*g

=83.78mm2

(Dimension of the stampings were obtained from the design table)

so far so good I guess, how do I determine the length of the stamping needed and is the equation that considers the effect of length on Inductance/Magnetic properties.

Kindly if there is an equation please post. Also if you can find mistakes in the design then please post them also.

Associate

Join Date: Mar 2011
Posts: 52
#5

### Re: Design of magnetic Ballast

11/18/2012 11:30 AM

The PCu=(Vb2 * Ib2 * ρcu) / ( (4π2f2 ) * (Aslot * pfcu * Tlam * Lstack * Bsat ) )

ρ= resistivity of copper;

Aslot = Area of slot

pfcu packing factor = 0.9-0.7 as per design

Tlam= thickness of Lamonation

Lstack =length of stack of Ballast (which I am trying to determine)

Bsat = Flux density 1T (always)

form the above equation We find that Lenth of stack is inversely proportional to Copper losses so there must be an optimum design point, which think we dont know so far.

If I believe you Mr. Joshi then why ballast manufacturers are making ballast which vary in sizes from 70mm to 40mm (stack length)?

Associate

Join Date: Mar 2011
Posts: 52
#6

### Re: Design of magetic Ballast

11/18/2012 11:33 AM

I am using that software its for electronic only, but my question is regarding the design aspect of magnetic/Inductance ballast

Associate

Join Date: Mar 2011
Posts: 52
#7

### Re: Design of magnetic Ballast

11/18/2012 11:34 AM

I want it to be approx 40mm to 70mm (length of stack) window factor and stamping dimension can be altered

Guru

Join Date: Oct 2007
Location: India
Posts: 1246
#8

### Re: Design of magnetic Ballast

11/18/2012 11:36 AM

Have you heard of reverse engineering?

Score 1 for Off Topic
Associate

Join Date: Mar 2011
Posts: 52
#9

### Re: Design of magnetic Ballast

11/18/2012 11:39 AM

Yes there is see my second last post

Associate

Join Date: Mar 2011
Posts: 52
#10

### Re: Design of magetic Ballast

11/18/2012 11:43 AM

No you are not allowed to do reverse engineering or any back calculations

Guru

Join Date: Aug 2010
Location: Melbourne, Australia
Posts: 740
#11

### Re: Design of Magnetic Ballast

11/18/2012 10:33 PM

Inductance and saturation or not are controlled by a number of factors:

Core Material

Number of turns of wire

Cross sectional area of the core

I suggest you search on Inductor design, or magnetic field or a variety of other similar terms

Guru

Join Date: May 2006
Location: Placerville, CA (38° 45N, 120° 47'W)
Posts: 5989
#12

### Re: Design of magetic Ballast

11/18/2012 10:44 PM

Then this must be homework! And we are not supposed to solve homework problems.

Anyone who asked you to design a magnetic ballast in 2012 is either way out of date, or is a teacher who thinks the experience is worthwhile. And indeed it may well be, but the latter means homework!

__________________
Teaching is a great experience, but there is no better teacher than experience.
Guru

Join Date: Jan 2008
Posts: 1753
#13

### Re: Design of Magnetic Ballast

11/18/2012 11:42 PM

I was blinded by a shower of equations of unknown origin and even less known pedigree. If this would be animal husbandry, I would not know, if I am dealing with a fish, a bird or a mammal.

In plain, forget it.

Build an inductance as you like it, then test. Repeat, and repeat again.

Here we are engineers. There is little liking to purely theoretical exercise, not coupled to any reality we can recognise.

Associate

Join Date: Mar 2011
Posts: 52
#14

### Re: Design of magetic Ballast

11/19/2012 12:11 AM

this is not a homework problem, just hoping to manufacture inductance ballast .Got some technical papers and here I was stuck with the length of the ballast.

Guru

Join Date: May 2006
Location: Placerville, CA (38° 45N, 120° 47'W)
Posts: 5989
#15

### Re: Design of magetic Ballast

11/19/2012 12:36 AM

Then why "No you are not allowed to do reverse engineering or any back calculations"?

And why waste time designing an inductive ballast, when electronic ones are cheaper and more efficient, unless you have some new breakthrough magnetic technology.

__________________
Teaching is a great experience, but there is no better teacher than experience.
Associate

Join Date: Feb 2010
Posts: 35
#16

### Re: Design of magetic Ballast

11/19/2012 8:46 AM

Who says you're not allowed to do reverse engineering?

Guru

Join Date: Apr 2008
Location: Lexington, KY
Posts: 1640
#17

### Re: Design of magnetic Ballast

11/19/2012 8:50 AM

VA=220V * 0.3Amps= 66VA (since its similar to 1:1 trans it will be /2)

Core design Calculation

VA rating = 33VA;

I believe you are going to find that the VA rating will change for various lamp lengths.

I think the information you may be missing is what output voltage is needed to run and start the various lamp type. That will determine the number of windings needed and thereby the overall length of the ballast. You must have enough room for those windings in order to achieve the desired voltage output at your nominal current.

ANSI has all of the design parameters you really need to know.

For an instant start lamp the open circuit voltage is a bit higher than for a pre-heat or rapid start application. Basically you need to provide about 10 volts per inch of lamp arc in order to have stable operation. Gas type in the lamp will vary that requirement.

What kind of ballast are you making? Are you providing any coil heating windings?

__________________
A great troubleshooting tip...."When you eliminate the impossible, whatever remains, however improbable, must be the truth." Sir Arthur Conan Doyle
Power-User

Join Date: May 2007
Location: Flanders (Belgium)
Posts: 344
#18

### Re: Design of magnetic Ballast

11/19/2012 8:53 AM

PCu=(Vb2 * Ib2 * ρcu) / ( (4π2f2 ) * (Aslot * pfcu * Tlam * Lstack * Bsat ) )

Pcu = V*V*A*A*ohm*m*s*s/m*m*m*m*V*s*m-2

Pcu = V*A*A*V*A-1*s/m

Pcu = V*A * V*s/m

Pcu = power x magnetical flux/m

Power cannot be equal to power times flux per meter

Or did I make a mistake?

2
Guru

Join Date: Jan 2008
Posts: 1753
#19

### Re: Design of Magnetic Ballast

11/19/2012 12:30 PM

Phew!

It appears, that the OP desires to become a manufacturer, without building any prototypes, including reverse engineering, and working to find out, what they do and how do they do it.

The OP sticks to linear equation formulas of quite foggy origin, on a device nonlinear 5 ways from sunday.

Magnetics are inherently nonlinear, simply drive them toward saturation and see.

A ballast works in one mode before, during and after ignition. Take the simplest case of a fully operating tube. Describe to us step by step, what happens as the assembly goes thru a single cycle of 50 (or 60) Hz. That is a minimum prerequisite to design anything.

OR, buy a few ones on the market, and copy the best looking one. And be happy with the results. No understanding needed.

Associate

Join Date: Mar 2011
Posts: 52
#20

### Re: Design of magnetic Ballast

11/20/2012 10:52 AM

I am a hobbyist, I am designing a Inductance Ballast for 36W FL lamps, if successful I will design for Floodlights and Industrial Highbays that use Incandescant lamps. Right now its all a simple 36W FL lamp.

Guru

Join Date: Oct 2007
Location: India
Posts: 1246
#21

### Re: Design of magnetic Ballast

11/20/2012 10:57 AM

You are welcome. Invent the wheel.

Associate

Join Date: Mar 2011
Posts: 52
#22

### Re: Design of Magnetic Ballast

11/20/2012 11:04 AM

Thanks for all you suggestions. The back calculation was very helpful. I was changing the length of the length to find different values of copper losses but then I have to adjust the flux, area of window etc to arrive at the most rational values. Thanks again for your suggestions.

Guru

Join Date: May 2009
Location: Richland, WA, USA
Posts: 20964
#23

### Re: Design of Magnetic Ballast

11/20/2012 11:23 PM

How about designing some buggy whips too?

__________________
In vino veritas; in cervisia carmen; in aqua E. coli.
Interested in this topic? By joining CR4 you can "subscribe" to