using transformer with higher leakage inductance
Posted by: dmosberger
on
Hi,
I'm trying to build a 5V/1.5A universal power supply (110-277Vac). PI Expert came up with a design that specifies a transformer with a primary inductance of 498uH and a primary leakage inductance of 16.4uH. However, since my volume is small, I'd like to use an off-the-shelf transformer, specifically Wuerth Eletronik's 749118215 (http://www.wuerth-elektronik.de/eisos/pdf/749118215.pdf). How do I account for this transformer's larger inductance (900uH) and larger primary leakage inductance (120uH)?
Thanks.
评论
Thanks for your prompt response!
I'll check on Ac with the transformer manufacturer and then double-check on Bpk (and Kp). I did measure I_DRAIN and even under full load, it rises linearly to about 0.55mA (with a TNY278), so AFAIK, it doesn't look like the core is saturating.
I attached the schematics of the supply for your convenience.
A prototype of the board works reasonably well, except that the output voltage drops rather quickly with load and that D205 is getting rather hot (presumably due to the higher leakage inductance).
I'm not sure what's going on with the output voltage drop. I believe partly the issue is due to the output cap having a relatively large ESR. I'll experiment with other caps, but I don't think that will completely get rid of the voltage drop. What kind of drops should I expect here?
As for D205 getting hot: would an RC snubber help to reduce the losses here or do I just have to live with the losses and pick appropriate diodes?
--david
You can directly calculate the resulting flux density as follows:
Bpk = Lp * Ip / (Np * Ac)
where
Lp = primary inductance
Ip = device peak Drain current
Np = primary turns
Ac = core center leg area
If you use all mks units (Henries, m^2), the resulting Bpk result will be in Tesla.
Most ferrite core materials will begin to saturate at around Bpk = 0.35T
You must not allow the core to saturate.
The xfrmr spec you linked doesn't appear to specity the core area of the core they use, so you'll have to ask them what it is.
If a transformer's leakage inductance is higher than you want, the loss in the drain clamp snubber will be higher. This means the devices that dissipate power (resistor usually), will have to be bigger. This may be OK depending on your efficiency and thermal requirements.
The other thing you need to check is if your transformer will meet the Kp requirements of the device.
One way to check this is to run PIXls and enter your design; then manipulate the input parameters until it describes your transformer. Then check if Kp is met.
Which device will you be using?