ETD29 power loss capacity

[analityk]

Let's assume free air convection, room temperature, temp rise 40 Kelvins or something more realistic - core temperature equal 100 Celsius degrees (minimum core losses).
How many power this core can exchange? I assume 1W, this core is rather small and have small heat transfer capabilities. I want take out 120W from this transformer  (200kHz, N97) so I think it isn't anything unrealistic.

[Wolfram]

Practical dissipation depends a lot on local airflow and how well the transformer can sink heat through its pins to the local environment.

Whether this is practical depends a lot on the topology as well. Doing 120 W in an ETD29 would be pretty realistic in an LLC converter with good airflow around the transformer, but very very difficult in a flyback converter running in a box where the ambient air heats up to 60 degrees and nearby components heat up the area around the transformer for example.

[analityk]

We will use LLC or hard switched half bridge, testing is in progress. But anyway I see now that I should prepare myself to long an boring testing of heat transfers.

[Faringdon]

You should get core loss chart from core datasheet......it shows the power/m^3 for your particular delta B and at your frequency.

Generally you shoudl not go above 300mT in delta B....so for you its +/-150mT...
..but you can exceed if you have fans or only short interval times at max load.

Also remember not to saturate, so dont go too far above 320mT if you can help it.

Small core is good as the volume is low, so the overall dissipation is low, as their is less material to heat up, and because it is small, all parts are close to the core surface, so to speak.

Some people say the hotter it gets...the more convection current tend to flow, which cools it back down again!

[analityk]

I think you missed clues. Part that you described is fine, my transformer work up on fixed frequency 200 kHz with max B =115 mT (usually something around 95mT). That's mean core have to dissipate something around 2 W in cold condition and 1W if core temp reach 100 Celsius degrees. (because core losses decrease in higher temperatures, up to 100 Deg C).
But how many power this small core can dissipate? 2W? Maybe 5W? How to calculate it?

Anyway my transformer temp was 30 K over the room temperature (and natural air flowing) in highest stress conditions so it is ok but it must be verified in lab.

[Faringdon]

You are asking how to calculate what is the max power that can be dissipated by core loss and winding loss before the transformer “blows up” so to speak(and this in your particular thermal environment).
Sorry There is no exact way to calculate this.
Best you can do is put a thermocouple to the core and measure temp rise in your setup, with max ambient temp.

Alternatively, and better, put on full power and in thermal chamber at your max ambient and run it all day, monitoring vout and pout. If it survives then you’re good to go

[dietert1]

Anyway, as a safety measure the board could have a temperature sensor. Similar to Li-Ion battery packs. Maybe you find a way to handle overload/overtemperature conditions. That sensor can be used to verify the margins of your design.

Regards, Dieter

[analityk]

My approach was easy, any way i was thinking that is easy in past. We know thermal conductivity of core material, dimensions etc so may be someone was build some plot with power lost versus temperature differences beetwen core and air in some defined circumstances.

Any way i got some good resutls with this core and overal effciency over 90%. With full power mode transformer temp. rise was 32-35 degree of Celsius.