Fly-back transformer turns depends on switcher IC?

Thread Starter

picstudent

Joined Feb 3, 2009
91
I am trying to design and make a flyback SMPS for about 26V 3 Amps.
I tried Power integrations designer with TOP switch at 66Khz switching frequency.
Since I have Viper100A with me, I tried to design in old STMicro's Viper design software.
Both cases I use same frequency and same transformer core and same input and output.

But the resulting transformer number of turns is very much different. Top switch gives very few turns but Viper gives more than tripple turns.
Anything I missed in the process or is there any special reason behind this?
 

Papabravo

Joined Feb 24, 2006
17,303
Why do you think that SMPS designs are unique? If two different pieces of software give different results then either the inputs are different or the algorithms are different. It would not be correct to draw any conclusions about the results without more information.
 

Ian0

Joined Aug 7, 2020
3,767
A usual flyback voltage is about 125V, so I would expect a turns ratio of about 5:1
Inductance will depend on frequency, number of turns will depend on peak flux density and reluctance, which depends on gap length.
 

Thread Starter

picstudent

Joined Feb 3, 2009
91
A usual flyback voltage is about 125V, so I would expect a turns ratio of about 5:1
Inductance will depend on frequency, number of turns will depend on peak flux density and reluctance, which depends on gap length.
In both deigners I gave input as 195V to 250V AC, both outputs 26V 3.5A. Both cases frequency I selected 66Khz and slected same core. Viper100 designer gives primary around 76 turns and Top switch only 20 tuns. Something I am overlooking in this case. Anyway thanks for the response.
 

Ian0

Joined Aug 7, 2020
3,767
In both deigners I gave input as 195V to 250V AC, both outputs 26V 3.5A. Both cases frequency I selected 66Khz and slected same core. Viper100 designer gives primary around 76 turns and Top switch only 20 tuns. Something I am overlooking in this case. Anyway thanks for the response.
What size core are you using? I'll put it in my spreadsheet and see what it says. I found Power Integrations' software to be pretty good. I only used Vipers once, and found them to be very fond of blowing up. Maybe they have improved since then.

The flyback voltage is an arbitrary choice - you'll see anything between 100V and 150V - which makes quite a difference to turns ratio and duty cycle.
 

Thread Starter

picstudent

Joined Feb 3, 2009
91
Input range 190V-250V AC
Output 26V 3.5Amps
Frequency 66Khz
Core is EE42/21/15 (Datasheet Image attached)
If you don't mind please clarify the term "Fly-back Voltage" in this case .

Core.jpg

Thank you very much in advance for the support.
 

Ian0

Joined Aug 7, 2020
3,767
Input range 190V-250V AC
Output 26V 3.5Amps
Frequency 66Khz
Core is EE42/21/15 (Datasheet Image attached)
If you don't mind please clarify the term "Fly-back Voltage" in this case .



Thank you very much in advance for the support.
DFAA3A89-1673-4F9A-934F-5A5A2D2E1268.jpegThis is the MOSFET drain-source voltage, on the primary side.
 

Ian0

Joined Aug 7, 2020
3,767
For a flyback voltage of 125V I get the lowest total loss at a flux density of 0.1T (55 turns/9 turns/1.9mm gap). However, the flux density could be increased as high as 0.2T which would need 28 turns/5 turns/0.5mm gap).
For a flyback voltage of 100V, it is 48 turns/9 turns/1.9mm; and for a flyback voltage of 150V it is 62 turns/ 8 turns/1.9 mm at 0.1T and 31 turns/4 turns/0.5mm for 150V and 24 turns/5 turns/0.5mm at 0.2T
A lower flyback voltage requires a higher PIV diode on the output, but the output is on for longer so the secondary ripple is lower, but the MOSFET is on for a less time, taking more current and making more primary side ripple (and vice versa).
 
Last edited:
Top