LT4320 with center tapped Transformer

Hi everyone,

I´m wondering if the LT4320 is able to handle with symmetric power rails. I designed the following schematic and build the circuit but as soon I connect the 0V (center tap) rail from the Transformer to the circuit GND, I get high current flow through the center tap.

I measured the circuit with the Oszilloskope an found the same as in the Datasheed. So the basic circuit is as expected.

Any suggestions and recommendations are welcome.

  • I would like to add some more Information:

    Without the center tap to GND, the Voltage between GND and OUTP is +24V and between GND and OUTN is -19V. There is a difference of -2,5V on each rail. That let me assume that the LT4320 is not working symmetrical. 

    If the center tap is connected to GND (as it should) the current is too high (fuse pops out) so I can´t measure for further investigation.

    If I substitute the LT4320 with a regular bridge-rectifier it works as expected.

    Where is the mistake?

  • "solved, center tapped Transformer with LT4320"

    The LT4320 works with a center tapped transformer if the secondaries are in phase. Normally the two winding of a CT are out of phase.

    I would recommend some explanation in the Datasheet of the LT4320.

    If you are  interested in more you can read here.

  • 0
    •  Analog Employees 
    on Sep 12, 2019 5:16 PM in reply to dremeier

    Hello Dremeier,

    I am glad you found a solution to your issue. Connecting both windings of your transformer in phase would create a single AC input, which the LT4320 will happily rectify. There is another option that may work with your center tap transformer.

    It is difficult to see what's going on with your schematic, but you could use the LT4320 with both windings out of phase. In this configuration, the center tap would be ground to your load while the LT4320 would have a separate local ground that would the be the "lowest" potential. These nodes are completely separated, with perhaps 10uF of biasing capacitance between them.

    IN1 would connect to pin 3 on your transformer, and IN2 would connect to pin 6. The circuit would then use only two MOSFETs to rectify the positive output, these MOSFETs would connect from transformer pins 3 and 6 to OUTP. Then two diodes are connected from OUTN to each input pin to assure OUTN stays low. These diodes will not conduct the load current. The bottom gate pins are unused in this configuration. Finally, the load is connected between OUTP and the center tap.

    Best Regards,

    Eric

  • Hello Eric,

    thank you for your advise. I guess your second option is in principal like this:

    This would result in an single output.

    What I need is a bipolar output with positiv and negativ Voltage in respect to GND or CT like this:

    In this scenario it is mandatory that the two windings of the transformer are in phase. And this is not described in the datasheet. Please can you update the datasheet of the LT4320 to advise others. It took me two week and a new transformer to solve the issue. 

    Kind Regards

    Andre

  • I liked your advise on the lt4320 with CT transformer...just built it and want to understand if I got it right...you wrote:

    "...but you could use the LT4320 with both windings out of phase. In this configuration, the center tap would be ground to your load while the LT4320 would have a separate local ground that would the be the "lowest" potential. These nodes are completely separated, with perhaps 10uF of biasing capacitance between them."

    ...So i connected two diodes with cathode to the input ac pins and anodes to the outn pin of the chip. Not sure if I need the 10uF now between outn and CT ? Normally the 10uF is between outp and outn, no ?...by the way  In audio amplifiers it is to my surprise this cap (quality and build construction, lytic vs. film etc) is very audible ?

    Any advise is gratefully welcomed

    All the best

    Frank