Hi,

Thank you for the interest in PI products. 

PIExpert platform is oriented towards conservative designs considering most tolerances and guarantees complete performance. 

PIXLS though is oriented towards experimenting with the schematic and parameters more freely so, the experienced user can orient its design in a particular direction. 

Best Regards

Thanks, though PIXIS wont let me do the said flyback with EFD15/8/5, do you know why?

I can do it myself...but boss wont use power.com chips if the pi expert or pixis wont do it.

Hi,  

The reason why PIXls cannot provide a solution within the restrictions is because it is very unlikely such a design to fly in a mass production.  

I would suggest seeing our local representative for detailed discussion of the requirements. Our FAE's have a lot of experience with finding solutions for tough design requirements. 

Best Regards, 

PIXLS will let you do a 17W, 24V design using the EFD15 core, but it requires an 80W TOPSwitch (TOP269) to do the job, and the design runs deeply in discontinuous mode (KP = 4.5). This is a classic example of a design where a too small of a core was forced into an application too big for it. The current stress is very high at both the input and the output of the supply due to the highly discontinuous operation. A somewhat larger core and a much smaller TOPSwitch would be a better pairing by far. I'm appending a *.pdf file for the PIXLS spreadsheet so the tradeoffs can be seen in forcing a small core to do a job too big for it.

Here is a design with a much more sedate pairing of TOPSwitch and core, again done using PIXLS. Using a somewhat larger core allows using a smaller TOPSwitch much better matched to the job at hand. I'm not sure that enough copper can be fitted into the RM7 to meet thermals, but use of triple insulated wire for the secondary with flying leads will allow the design to meet safety creepage.

Thanks,  though  the following is a 24V, 16W Flyback design with EFD15/8/5 intended for TOP264.

The peak primary current is 1.2A. Peak secondary current is 4.97A. Bpk is 292mT.

I checked the core loss's for 3F3 and they are not bad. It uses the AL=63 core set.

The turns ratio is 58/14 = Np/Ns.

It is as attached. Surely there is nothing electrically or magnetically stressful about this?

There is blown air blowing all over this flyback ,and the enclosure has vent holes to the outside air.

Do you find anything about this objectional?....

Not quite - this is what I got when I checked the design in PIXLS. You can run the numbers.

Thanks, but you have used an ETD15 core....it is required for us to use EFD15.

You have also wound turns that give your peak B over 370mT, so yes of course what you attach is unworkable.

The excel i attach shows it can be done with TOP264 and EFD15/8/5  gives Bpk 292mT.

With all respect, Your PDF is comparing apples with pears..its a completely different situation.

But thanks for your efforts in this.

That was a typo on my part - the parameters entered in the spreadsheet were those for the EFD15. I can show you what your proposed EFD15 design with  58/14 turns will do with a TOP264 - I doubt you will like the results.. The PIXLS tool incorporates the current limit tolerances into the transformer design, making sure the design will deliver the required  power at the lowest current limit, but not saturate at the high end of the current limit tolerance. there were no warning messages generated with the TOP265 design shown - if you don't like the peak flux, you can throw one more turn on the secondary and open up the gap a bit. Appended is the file PIXLS gave me for your 58/14 turns design using the TOP264 and an EFD15 core, dialed down to the lowest current limit it would take before being unable to deliver the required power at the low end of its current limit tolerance.  

Thanks, yes, what you have chosen gives too high B, but that is not what we have chosen here.



Choosing the AL=63 core set of EFD15/8/5 and with 58 turn prmary gives Bpk = 0.292

i(sat) = B(sat).A.N/L = 0.3 * A.58/ 212uH = 1.23A

TOP264max tol currtn lim = 1.392......for this B = L.i/N.A = 0.339

For 16W output and efficiency of 0.82 the peak current is 1.2A  (B = 0.292 for this)

fsw=132khz, its in DCM over most of 100-265vac.

There is nothing wrong with this, and i am surprised tht PIXLS won't allow this to be put in. Do you know why?

My boss will not  choose power integrations if the PIXLS or PI Expert do not take it.

(i did shave the power output down to 16W i confess here)

The problem coudl be the dson conduction loss, though we have blown air...but just in cse, do you have "soic8 style" heatsinks to go on top of TOP264 SMD version.

The transformer will be directly under the TOP264 so we cant thermal via down to a heatsink.

I suggest that you test your spreadsheet by winding up your design and trying it with a TOP264, closely monitoring the drain current.