mikrocontroller.net

Forum: Analoge Elektronik und Schaltungstechnik Flyback transformer- Design


Autor: Daniel P. (daniel-p)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
Could any body please help me with the following problem (You could 
write in German :-)):

Flyback transformer in discontineous mode:

"Vin"=11-12V Dc
"Ipeak"(primary)=520mA, ton(turn-on time)=5us
"Lp"(primary inductance) is calculated by: Lp=(Vin*ton)/Ipeak

A EF20 core is used with "Ae"=31mm^2 and "AL"=1200nH.
Wire with "d"=0.38mm is used.

Air-gap calculated by: 
"Lg"=[0.4*3.14*Lp*Ipeak^2*10^8]/(Ae*Bmax^2)=0.02mm
"Bmax"=2000Gaus selected.
Primary number of turns  by: "Np"=(Bmax*Lg)(0.4*3.14*Ipeak)=8 turns
------
Surprisingly the primary current taken from the power supply during the 
turn-on time is too small, about 100mA, expected 540mA!!
Questions:Why is the current so small?
          Why is the inductance not equal to 100uH,it's about 80uH.
          (In each case the current is too small for Lp=80uH or 100uH)

I'll be gratefull if any body can help me with this problem!
thank you very much in advance

Autor: hgd (Gast)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
are you measuring the primary current with load on the secondary?

Autor: daniel-p (Gast)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
Thanks for quick reply:-)

I'm using this flyback transfprmer to charge a capacitor(my load is 
capacitive).
I saw that my transformer doesn't take the expected current, therefore I 
removed the secondary winding to see whether the primary works as a 
normal inductor, what actualy schould be.I recognized the current taken 
from the power supply is always the same:-(

Autor: Daniel-p (Gast)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
Hi!
I have a flyback configuration as a part of a more complicated low power 
design.
in this design I must disconnect the secondary winding from output 
capacitor using a switch or a device funktioning as a switch.

Since this switch must be placed between the secondary winding(which has 
floating voltage levels)and output capacitor(which is being charged and 
has no constant voltage), both sides of the switch have floating 
potentials, therefore there is no reference potential to activate the 
switch.


       Requirements:
                    1: an absolute disconnection must exist(no leakage 
current)
                    2: Maximum current flowing through this switch is 
about 1A
                    3: reverse voltage on the switch(off status) about 
900V (difference of the voltage on the
                           output capacitor and the secondary winding)
                    4: switch power dissipation must be very low
                    5: switch must be very fast, some micro seconds
                    6: rise and fall time for the switch turn on and off 
must be in nano second

Important: -MOS transistors are not adequate because current leakage 
through free wheeling diode
            makes the absolute disconnection impossible!
           -Relays are too slow and bulky


I do appreciate your idea!
I'll be gratefull if any body can help me with this problem!
thank you very much in advance



----     ----------/ ----
    |    |      switch   |
   ...  ...              |
   . .  . .             ---
   . .  . .             ---
   . .  . .              |
    |    |               |
----     ----------------

Autor: eProfi (Gast)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
Scenario 1:
What I do not understand is, do you have a diode between secondary and 
capacitor? Or should the switch act as a diode (active rectification)?

Scenario 2:
What I think you want to have is charging the cap and then turn of the 
flyback, hoping that the cap would hold the charge as long as possible.
Then I suggest a semiconductor and a (reed-) relay in series.

Plase tell us more about your circuit.

Autor: eProfi (Gast)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
Some more thoughts (I just sav that the first postings are more than 2 
weeks old):


"AL"=1200nH    8 turns
L = 8  8  1,2µH = 76,8µH

Why did you expect it should have 100µH?
The inductance is not constant, it depends of the field strength / 
current.

Did you succeed in getting an higher primary current?


> I saw that my transformer doesn't take the expected current,
> therefore I removed the secondary winding to see whether the
> primary works as a normal inductor, what actualy schould be.

You know that you cannor simply remove the secondary of a flyback?

The energy stored in the inductance (air gap) must be "discharged" / 
zeroed , otherwise the core would saturate.


A literature tip:
TI / unitrode   magnetics design handbook

http://focus.ti.com/docs/training/catalog/events/e...

http://focus.ti.com/lit/ml/slub007/slub007.pdf
(many pdfs!)


The german word for flyback is Sperrwandler (to use the search 
function).


> Important: -MOS transistors are not adequate because current
> leakage through free wheeling diode
many IGBTs do not have an intrinsic diode.

Autor: Daniel-p (Gast)
Datum:
Angehängte Dateien:

Bewertung
0 lesenswert
nicht lesenswert
In my design I try to charge and discharge a capacitive load, 
demonstrated as output capacitor.
Primary is connected to a low power supply and output capacitor should 
be charged to a high voltage level. I switch the primary (low side) in 
order to charge the capacitor and I switch the high side(secondary) to 
discharge the capacitor.
So the diode in the secondary(used to charge the capacitor) is realized 
by the free wheeling diode of the switch in the secondary side(as 
depicted in photo) and the diode in primary(used during discharging of 
capacitor) is realized by the free wheeling diode of the switch in the 
primary side.

Circuit is depicted in the following address.

http://www.4shared.com/photo/ElAXGiDe/question_pic.html

thanx a lot!

Autor: Daniel-p (Gast)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
eProfi schrieb:
> I suggest a semiconductor and a (reed-) relay in series.


may I have more information please?

Autor: Daniel-p (Gast)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
> Why did you expect it should have 100µH?
>
> The inductance is not constant, it depends of the field strength /
>
> current.
> >
> Did you succeed in getting an higher primary current?

Yes I get highet currents, the current meter was not calibrated!!
About the inductance: you are right,N*N*AL=L which results in a smaller 
inductance than expected.
The procedure I used to calculate and make the transformer is written in 
many refrences, but actually it's also confusing.As explained in these 
refrences, using this procedure results in an expected value of 
inductance, what indeed doesn't work!
**please give me an advice how schold I calculate the number of turns in 
primary and the Airgap length?
**Is this a try and error process in which I calculate the number of 
turns and the air gap and then I check the behaviour of the transformer 
and consecuently modify the number of turns and the air gap length?
*How is the formel N*N*AL=L for transformer with an air gap?I mean 
should I make more turns for example 10 and calculate L=N*N*AL=120uH and 
then try to decrease the L by an air gap?

Autor: Daniel-p (Gast)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
eProfi schrieb:
> many IGBTs do not have an intrinsic diode.

do these IGBTs fulfil my riquirements listed above(Current,Voltage,...)?
may you please name a few? I could not recognize from the data sheet 
whether they are with or without an intrinsic diod.

thanx in advance

Autor: eProfi (Gast)
Datum:

Bewertung
0 lesenswert
nicht lesenswert
> whether they are with or without an intrinsic diod.

It should be visible in the diagram and in the text.


There are only a few manufacturers: IXYS and IR , to name a few.


You select the core with the correct air-gap, look to the datasheet for 
the Al-value and calculate the L.
Siemens has a broad palette of "sifferit" magnetics:

Here is a collection of their products of year 2000:
www.dextermag.com/uploadedFiles/Literature_Ferrite_EPCOS.pdf  (4 MB)


Read the slup124.pdf ... slup129.pdf of TexasInstruments.

Antwort schreiben

Die Angabe einer E-Mail-Adresse ist freiwillig. Wenn Sie automatisch per E-Mail über Antworten auf Ihren Beitrag informiert werden möchten, melden Sie sich bitte an.

Wichtige Regeln - erst lesen, dann posten!

  • Groß- und Kleinschreibung verwenden
  • Längeren Sourcecode nicht im Text einfügen, sondern als Dateianhang

Formatierung (mehr Informationen...)

  • [c]C-Code[/c]
  • [avrasm]AVR-Assembler-Code[/avrasm]
  • [code]Code in anderen Sprachen, ASCII-Zeichnungen[/code]
  • [math]Formel in LaTeX-Syntax[/math]
  • [[Titel]] - Link zu Artikel
  • Verweis auf anderen Beitrag einfügen: Rechtsklick auf Beitragstitel,
    "Adresse kopieren", und in den Text einfügen




Bild automatisch verkleinern, falls nötig
Bitte das JPG-Format nur für Fotos und Scans verwenden!
Zeichnungen und Screenshots im PNG- oder
GIF-Format hochladen. Siehe Bildformate.
Hinweis: der ursprüngliche Beitrag ist mehr als 6 Monate alt.
Bitte hier nur auf die ursprüngliche Frage antworten,
für neue Fragen einen neuen Beitrag erstellen.

Mit dem Abschicken bestätigst du, die Nutzungsbedingungen anzuerkennen.