I am once more about to convert a 220VAC appliance to 12V; in this case an electronic 150W HID lamp ballast. Since the device also accepts DC this is an easy task: use a push-pull stage to generate 200-350VDC from the 12VDC input, and feed that to the ballasts' input.
But, I want to fit the 12V->200..350V converter into the void areas of the ballast so the occupied volume of the ballast won't increase.
Now, I do have some space to fit the electronics, except for the transformer. But what I can do is to use a couple of EFD20 cores, parallel the primaries, and put the secondaries in series.
Thus, instead of one 1:22 transformer I could use three 1:7.5 transformers. With the primaries parallelled and secondaries in series this acts like a 1:22.5 transformer.
I have built it as a simple unregulated push-pull using a couple of Wurth WE-Flex transformers, and it works. Drain voltage on the FET's looks nice too, even without snubbering and maximum input voltage there is little oscillation and spikes. Efficiency is above 90% with 110W output. Transformers become 23C warmer than the environment, the MOSFET's (FDP65N06) hardly need heatsinking.
I only see advantages in doing so (provided that the cores are free 😉 ). Three cores instead of one provides better heat dissipation and lower leakage inductance.
What is your opinion on this?
But, I want to fit the 12V->200..350V converter into the void areas of the ballast so the occupied volume of the ballast won't increase.
Now, I do have some space to fit the electronics, except for the transformer. But what I can do is to use a couple of EFD20 cores, parallel the primaries, and put the secondaries in series.
Thus, instead of one 1:22 transformer I could use three 1:7.5 transformers. With the primaries parallelled and secondaries in series this acts like a 1:22.5 transformer.
I have built it as a simple unregulated push-pull using a couple of Wurth WE-Flex transformers, and it works. Drain voltage on the FET's looks nice too, even without snubbering and maximum input voltage there is little oscillation and spikes. Efficiency is above 90% with 110W output. Transformers become 23C warmer than the environment, the MOSFET's (FDP65N06) hardly need heatsinking.
I only see advantages in doing so (provided that the cores are free 😉 ). Three cores instead of one provides better heat dissipation and lower leakage inductance.
What is your opinion on this?
- Status
- Not open for further replies.