0
\$\begingroup\$

I'm doing some pretty basic "experimenting" with some transformers to step up voltage for an embedded project. I've got various 1:N transformers from coilcraft: http://www.coilcraft.com/lpr6235.cfm

I'm driving a FET with a 5V to 0V square wave (not -2.5V to 2.5V) from a function generator, and I'm powering the primary side with a power supply. Unlike the flyback circuit example in the data sheet, I've got a high power resistor in the circuit to limit the current to not exceed the saturation current of the transformer.

I was expecting that with a 1:N transformer, my output voltage would be N times my peak voltage, provided that I reach the saturation current. So, with a 1:10 transformer and a 5V square wave, I expected a peak to peak voltage of 50V, but I'm not getting anywhere near that.

edit: Added information

schematic

simulate this circuit – Schematic created using CircuitLab

The FET I'm using is actually FQP30N06L

Since I'm using a function generator, I vary the frequency until I hit peak voltage on my rms multimeter. I'm using an N-fet to switch a 5V supply, and I am expecting 100V, since the transformer ratio is 1:20, but my multimeter only shows about 19 V.

Here's the output of the scope: The yellow output is measured at Test_Point. The Blue output is measured at V_supply. The gate signal is not shown, but it's very clean by comparison. Also worth noting: I switch out the R1 resistor for various values to try to hit saturation current. When the resistor is 2 Ohms, and the current is about .56 A, I get a higher voltage than when I use a lower resistor and run at saturation current (.7 A). Either way, the output voltage is about 1/5 what I was expecting.

enter image description here

\$\endgroup\$
7
  • 4
    \$\begingroup\$ Although I can almost "see" what you're trying to do, it would be better if you included a schematic just to avoid any confusion. There's a circuit editor: edit your question and press the circuit-symbol to draw a schematic. Including a picture is also OK. \$\endgroup\$ Jun 13, 2017 at 13:26
  • \$\begingroup\$ One of the reasons is that you don't have 5Vpp across the transformer... attach a scope and observe that. \$\endgroup\$
    – PlasmaHH
    Jun 13, 2017 at 13:37
  • \$\begingroup\$ At what frequency are you driving the transformers? You say you are using a high power resistor but power does not matter, resistance does. What is the value of the resistor? \$\endgroup\$
    – Glenn W9IQ
    Jun 13, 2017 at 14:00
  • 1
    \$\begingroup\$ do you get about half of what you expect? \$\endgroup\$
    – dandavis
    Jun 13, 2017 at 14:53
  • \$\begingroup\$ Your load also makes a difference, what is the load on the secondary? \$\endgroup\$
    – Voltage Spike
    Jun 13, 2017 at 21:30

2 Answers 2

1
\$\begingroup\$

It sounds to me like you may have an average DC voltage across the transformer, which means you will saturate the core, or drop most of the voltage across your power resistor in series.

This would be the case if you were simply switching the primary with one end tied to your supply and the other to a FET to ground. If you are doing something different than that please add a schematic to show what it is and we can supply a better answer.

Try driving it with a bridge, to allow the flux to reset or coupling to the primary with a capacitor.

\$\endgroup\$
2
  • \$\begingroup\$ That's pretty much it, with the exception of the resistor in-line on the primary side to limit current. \$\endgroup\$ Jun 14, 2017 at 14:07
  • \$\begingroup\$ I've updated the question with the schematic and oscilloscope output, as well as other details. \$\endgroup\$ Jun 14, 2017 at 21:34
1
\$\begingroup\$

Apart from anything else, a 1N4001 @ 400KHz??! Check the reverse recovery time, you will want a much faster diode then that!

You will also want some mechanism to reset the core flux when you turn off that mosfet or you will probably wind up killing it.

\$\endgroup\$
1

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.