UCC28811 flyback transformer design

Hello,
I don't know how familiar you are with transformer design. If you have little knowledge it may be best to find a transformer manufacturer you can give specifications to. Threre are a lot of things to consider when designing a transformer, probably the most important is what safety regulations you need to meet. Based on the requirements and input voltage and working voltages that will determine the creepage and clearances needed. There's also thermal such as Class A, B, F or H. Other things to consider are frequency range, input voltage range, auxilary windings and how to deal with the leakage inductances if you have bias windings.
Another thing to consider is cost. If you have a transformer manufacturer they know the least expensive way to build what you are looking for otherwise you are guessing on core geometry and how wind it, i.e. margin tape and three layers of tape versus triple insulated wire.
I can find more information for this but if you never have done a transformer design it is quite involved. If you know how to design a transformer and need specifics I can come up with that as well.
Thanks,

Dear Sir ,

                How are you? I hope you are good .

Thanks a lot for your fast response . we design a lot of transformer before but it was for small power & the data sheet of the IC's we were use describe How to design the transformer very well . this time we use high power IC to drive( 50-100) watt led panel so we need formula to calculate the parameter of transformer from it or transformer design example specific for (UCC28811). I hope you understand what we want this time .

Thanks & Regards

Eng/ Zaki Reda

If you have designed small transformers for offline PFC flybacks it would just scale up. Larger core, thicker wire though you have to watch for AC losses, etc.
If you have not designed for offline PFC flybacks it is quite a bit more involved than DC to DC.
Some things to consider:
The ouput power of the transformer varies over the line cycle.
The flux level of the transformer will be the highest at low input line at the peak of the rectified AC.
The frequency will be the highest at high line.
The flyback MOSFET voltage rating is dependent on the turns ratio and output voltage.
At higher power leakage inductance can affect the bias windings significantly.
The LM3447 datasheet has some equations for an offline flyback.
Is this universal input?
There are so many variables when designing for these applications it would be difficult to have a set of equations that could result in optimal transformer designs for different applications.

Dear Sir ,

                How are you ? I hope you are good also your work .

Thanks a lot for helping us . I read the LM3447 data sheet . we have previous knowledge about fly back transformer .as you mentioned above There are so many variable when designing for these applications. we need to know How to calculate this variables ? we need to design our led driver . we need to understand each component

H hope you under stand our issue .

Thanks & Regards

Hello,

You have more variables than equations so how I go about this is to create either a spreadsheet in excel or something in mathcad (I prefer mathcad because it carries the units catching most mistakes).  I start with input voltage range, output power and output voltage.  I pick a turns ratio from primary to secondary that allows the duty cycle to be reasonable and also make sure the MOSFET drain voltage is not too high to have avalanche issues.  For 90 to 277 VAC I may chose 150 to 200 volts for the reflected voltage to the primary winding.  This means that if I have a 50 volt output and a 4:1 ratio the reflected voltage would be 200 volts on the primary side.  This means the MOSFET would see 200 volts plus 277 VAC * 1.414, and add 10% for high line, on the drain during the MOSFET off time (not including overshoot/ringing which would make this higher).  There are decent 800 volt and 900 volt MOSFETs, much higher and the RDSon versus cost will really suffer so you don't want 400 volts reflected.  The turns ratio will also determine primary RMS currents so you don't want 50 volts reflected either since the peak primary currents will get very high.

Next is figuring out the core/bobbin.  I start by picking something that I think will be close.  Try a PQ3535 and you may be able to shrink from there (or maybe not depending on the variables you pick).  Pick an AL value, 315 nH/turn^2 which will have the largest standard gap from the Ferroxcube book (you can always go other AL values but if the gap gets too be there can be fringing losses).  The core size is going to be dependent on desired switching frequency range (lower means bigger core), output power (need more bobbin area for wire to reduce copper losses) and input voltage range.  Peak flux should be calculated at low line at the peak of the rectified AC.  At higher frequencies/larger wire diameter you want to keep the wire AC losses in mind also.

I don't go by a set of calculations to design a transformer I just go through a few interations until I come up with what I want.  Let me know if you have mathcad.

Dear Sir

               How are you ? I hope you are good also your work .

we have question about  what is the difference between using UCC28811 in single stage & using it in two stage .

Does the transformer will be affected ?

why there is application use two stage ?

Thanks & Regards

Eng / Zaki Reda

Hello,

The transformer could be affected.  For example if the LED stack is 40 volts the transformer would be optimized for a 40 volt secondary even though you are providing current to the LEDs.  In a two stage with a flyback for the first stage and a buck for the second stage you may want to set the output voltage to 50 volts (regulating voltage instead of current) so there is overhead for the buck converter.  The overhead would be for transients, the voltage ripple caused by PFC, brown out, etc.  The turns ratio may change a little for that as well as it may need to provide slightly higher power due to buck converter efficiency.

A single stage PFC solution has a very low bandwidth.  This can cause changes in light output due to input voltage changes and also would limit how fast you could change the light output due to the PFC bandwidth.  A single stage solution will also have significant current ripple.

A two stage solution the buck converter can have fast bandwidth and absorb changes in its input voltage which will happen due to the low bandwidth of the PFC loop.  It is also much easier to PWM dim a two stage solution singe there isn't direct feedback to the AC line.  The center of a two stage solution is a buffer that can absorb anything that can happens on the input or the output.  It's also good for controlling multiple LED strings.  A two stage solution can have near zero current ripple and it's possible to use a smaller bulk (energy storage) capacitor.  Fault conditions are generally easier to deal with as well (shorted string, etc).

It's also easier to use the UCC28811 as a voltage regulator but there needs to be a second conversion stage to regulate current.

Generally, single stage for low cost reasons, two stage for performance, control and minimizing light ripple.

Thanks,

Irwin

Dear Sir ,

                 How are you ? I hope you are good also your wok .

During study transformer design . you suggest using LM3447 . I found it can drive volt up to 60 volt & Current up to 2 A

but  the data sheet  recommend use it for watt from (5-30)watt , Can I use it to make 50 watt led driver .

Thanks & Regards

I don't see why that would be an issue. The MOSFET is external and the LM3447 does not know what the output is, only by translation on the primary side since no optocoupler is needed.

Thanks