Non Isolated 20W or more buck converter

8 posts / 0 new

i need to identify the best solution to realize a non-isolated buck converter with this parameters:
Vin: 85-265 (Universal input)
Vout: approx 20V
Pout: 20W nominal or more

Which is the best solution for this apllication?

Thanks very much

Keep in mind that for a non-isolated buck converter, the switching device will need a current capability of at least the output DC current plus the peak inductor ripple current. If the switching device fails, the output load circuit will be subjected to the input B+. The Link-TN is designed with a non-isolated buck converter application in mind, but its peak current capability is only 360 mA.

Additionally, the output inductor for a non-isolated buck will need to be rated for the peak B+ voltage at high line, limiting the choice of off-the-shelf inductors for the application, especially given one of this power capability.

In years past, there was a design example using a TOPSwitch-II in a buck converter design, but using the newer TOPSwitch families in a non-isolated buck application has not been proven out.

If you choose a non-isolated flyback instead, you will have a wider choice of devices for your primary controller + switch, as the peak current can be lower for the same output power, allowing one to use choices from either the TinySwitch or LinkSwitch product families. I ran a spreadsheet using PIXLS, and it looks like this application could be served with a TNY288 and an EF20 size transformer. A similar flyback could be designed using a device from the LinkSwitch family.

Given these choices, please let us know how you would like to proceed.

Thanks for the reply.
Here more details of our apllication.
This power stage would be the main conversion stage from our industrial project. The input is a 120Vdc (100V to 130V range) and we need an output of 18V-20V with approx 20W ouput power. The input is DC and not AC, but we can remove the diode bridge. We need to evaluate a non isolated solution. Max ambient temperature is 50 to 60°c. Which tools i need to use to simulate/evaluate a non-iso flyback topology? PI Expert and PIxls seems not to have a non-iso solution. Thanks

Here is a sample circuit using the TinySwitch 4 series. I used an EE25 core instead of an EF20 to reduce the transformer turns.Also, bobbins are more readily available for the EE25.You can use PI expert or PIXLS to calculate the turns and inductance value for the transformer, as well as the parts around the TinySwitch.

Since your temperature range goes up to 60C, you may need a larger TinySwitch-4 using the reduced current feature in order to keep operating temperature within limits. A larger output diode may also be necessary.


Thanks for the schematics. If i remove the diode bridge, does this circuit will work with a DC input between 100Vdc to 130Vdc?

DC input is no problem - I'm looking at the effects of using 100VDC input. This can be simulated in the spreadsheets by using a min input voltage of 70VAC rather than 85 VAC, and greatly increasing the value of the input bulk capacitor in the spreadsheet (1000-2000 uF). I will also look at the possibility of using a TOPSwitch-JX in this design, as it offers the possibility of using an external heat sink - a nice option, considering the 60C temperature requirement. The circuit using TOPSwitch will be a little bit different than for the TinySwitch option, as I have to change the feedback network, but the overall architecture is the same.

The lower DC input will change the transformer design slightly and perhaps require a larger device.

If the upper input DC voltage is truly 130V, this changes things a bit, as one can use a lower voltage output Schottky diode with a lower drop, for reduced dissipation/higher efficiency. Please confirm...

Still waiting to see if input range is really 100 to 130VDC. This definitely changes the requirements for the output rectifier.

Hi, we are still in requirements definition. Don't know at the moment the exactly range. Thanks. Regards