Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Designing a simple Isolated DC/DC converter

Status
Not open for further replies.

udhay_cit

Full Member level 6
Joined
May 16, 2008
Messages
346
Helped
37
Reputation
74
Reaction score
37
Trophy points
1,318
Activity points
3,895
I have an isolated 1W, 5V-5V DC/DC converter from Murata (NXE1S0505MC) and I'm impressed with the converter design. As I can see the input primary circuit is looks like a "Royer Oscillator" and the toroidal core is embedded inside the multilayer PCB. The internal X-Ray view is in the link. https://www.youtube.com/watch?v=CT3yCUjMXlc

I'm interested to make one working circuit of royer oscillator based isolated DC/DC converter for 1W to 5W power rating just for ground isolation purpose. The primary driving circuit has very few components with two MOSFETs, 3 capacitors (including the decoupling) and only one resistor. From the X-ray video, I don't see more primary turns (My count is 8+8 push pull) and more turns are possible because of multilayer PCB.

I try to make my prototype circuit with few toroidal cores available with me and it doesn't work. Any one has experience to make DC/DC converter like this.

Udhay
 

Attachments

  • Murata.jpg
    Murata.jpg
    56.7 KB · Views: 175

Royer Oscillator is also known as a self saturating oscillator.
The transformer is a push-pull type. When one side saturates, it turns off and the other side turns on.
The oscillation frequency depends on the L of the coil and the C of the capacitor, and the stray LC and parasitic LC.
You need a core material that will support successive saturation at a reasonable frequency.

You do not want matched transistors, as it is the minor differences that get the thing started. Same type yes, but not matched.

There are several good sources on this on the internet.

As a side issue, the X-Ray video is very pretty, but I have concerns about Murat Erie's intellectual property and the propriety of X-raying it.
 

I'm familiar with those DC-DC's and quite like them.

Did you fully reverse engineer the schematic? Seems like you should start with that and try to simulate it.

You could also remove the components and move them to your own board. Perhaps move the primary off while still using the tranformer and secondary. And so-on step by step so you can isolate the issue.

Self oscillating circuits are simple but might be tricky.
 

The primary driving circuit has very few components with two MOSFETs, 3 capacitors (including the decoupling) and only one resistor.
4 resistors on the photo...
 

The transformer is a push-pull type.
Not necessarily.
Any one has experience to make DC/DC converter like this.
Linear technology (ADI) uses the Royer oscillator for their WPT DC DC converters (LTC4120 charger, check it on YouTube) (more on app note 138).
 

4 resistors on the photo...

The photo attached earlier is differ from the actual converter. The previous image is not even a product photo and it just a 3D image. I have attached the picture of my DC/Dc converter which has only one resistor.

- - - Updated - - -

Not necessarily.

Linear technology (ADI) uses the Royer oscillator for their WPT DC DC converters (LTC4120 charger, check it on YouTube) (more on app note 138).

Yes. You can see the attached schematic of the LTC4120 application diagram for wireless power charging. The circuit arrangement nearly looks like a astable multi vibrator. It requires more components compared to the Murata converter. Its very hard to include all the components inside a 1W DC/DC converter with a very tiny form factor. One another thing is, its working circuit in a commercial product.

- - - Updated - - -

The application notes AN118 from LT is almost matching with the Murata circuit including the resistor value 820Ω. From my previous post I mentioned about two MOSFETs is actually a BJT NPN transistor FMMT619. I can not able to verify the AN118 circuit with the murata converter because the components are soldered and the coils are showing 0Ω in multimeter.
If requires I will de-solder one converter to verify the circuit and to measure the inductance of the coils. But I hope the circuit is same

Udhay
 

Attachments

  • NXE1S0505MC.jpg
    NXE1S0505MC.jpg
    67.9 KB · Views: 273
  • NXE1S0505MC-1.jpg
    NXE1S0505MC-1.jpg
    37.7 KB · Views: 231
  • h947G.png
    h947G.png
    26.6 KB · Views: 362
  • an118fb.pdf
    2.4 MB · Views: 142
  • AN118.PNG
    AN118.PNG
    72.1 KB · Views: 264

The circuit arrangement nearly looks like a astable multi vibrator.

View attachment 151722

Absolutely right. It has inductors instead of capacitors. To make a 5v to 5v power supply, a buck-boost can do the job. However it inverts supply polarity. And it is not isolated.

astable twin buckboost +5v to -5v 2W.png

Adjust bias resistors to determine what voltage goes to the load. More effort is needed to add voltage regulation.
 

Finally, its working:thumbsup: Only 2 transistor and single resistor.

As I can understand from the operation, the switching frequency is depends on the toroidal core permeability and the number of turns. The switching frequency varies based on how fast the core is saturating. I managed to get the oscillation at 25KHz with my core. NiZn or any other high permeability core will increase the switching frequency. The Murata converter is oscillating at (Measured) 126KHz.

This circuit may help me to design upto 10W of isolated converter for my application. The efficiency is less then 70%. I have not used any capacitor in parallel with the coil. May be the suitable capacitor value can help the circuit to resonate.

Udhay
 

Attachments

  • Working circuit.PNG
    Working circuit.PNG
    18.8 KB · Views: 237

I think you really need to add a resistor in the base in the base of each transistor and in series with the 5V center tap.
There are 2 saturations of the core each cycle. Each saturation causes path from the 5V center to ground with no current limit.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top