UCC28950 / resistor of reset the current sense transformer CT

Hello All,

I'm looking forward to the reply.

Best Regards,
Ryuji Asaka

 

Hello All,

I'm looking forward to the reply.

Best Regards,
Ryuji Asaka

I'm looking forward to your reply.

Best Regards,
Ryuji Asaka

Hello John san,

Thank you very much for your reply and sorry for my repeatedly post.
I understood that the Rre is for limiting the voltage.
And Rre adds the error.

Best Regards,
Ryuji Asaka

I do not think that Rre is only used for the very uncommon case of Rs becoming open. It resets the secondary magnetizing inductance current. When current goes to 0, the magnetizing current (Vxt/L) cannot flow through the diode and this resistor dampens it. Soemtimes the resistor is omitted and the energy goes to the parssitic capacitance of the winding, or to a zener diode, but something has to be done to avoid the current to build up. Note that if the capacitance of the winding is to big, there is no time to reset the transformer.

Actually,

Volt x sec balance of the current transformer means that

V(on)*T(on) = V(off)*T(off)

So V(off) = V(on)*T(on)/(1-T(on))
V(off) = V(on)*D/(1-D)

Because the max value of V(on) is set to  2.06v by the cycle to cycle current limit and because D has a max value of 0.97

This will give a max off voltage of  66.6v without any resistor used in the Rre position.

What you say is true as the way to know the voltage needed to reset the transformer within the needed time.

But you do not say were the 66.6 V come from. The energy has to go somewhere: or to a voltage source (i.e. zener) or to a resistor or to a capacitor.

A practical example (a true one) will show best what I mean:
A current sensing transformer, 1/100, Ls = 7 mH, resonant frequency = 500 kHz (so interwinding capacitance = 14,5 pF). Converter at 100 kHz, Dmax = 0.95 so ton = 4.75 us, toff = 250 ns.
Sec. voltage = 1.6 V + diode loss = 2.2 V. Magbetizing current Im = 2.2 V x 4.75us / 7 mH = 1.5 mA.
If there is a 42V zener diode (in series with a diode) then, as you said, toff = 2,2 V x 4,75 us / 42V = 249 ns and the transformer will be reset during toff.

But if there is no zener, no resistor and in an ideal case no parasitic capacitance, the 1.5 mA will continue to flow and theoretically the voltage would rise to infinity, but then the output diode will go into avalanche and the energy will be dissipated into this diode. But the avalanche voltage of the diode has to be higher than 42 V in this example.

In a real world there is a parasitic capacitance. During the toff this current will charge the interwinding capacitance. In the example, 1/4 period is only 500 ns and there is no time to pass all the energy to the capacitance.
Adding an insulating layer between the windings changed the resonant frequency to 1.2 MHz, capacitance = 2,5 pF, T/4 = 208 ns and then yes, it is possible to have a transformer reset without resistor or relaying in the avalanche of the diode. The voltage will reach 1,5 mA x SQRT(7mH/2,5 pF) = 79 V, and the output diode has to be choosen according to this.

I forgot to say that if there is no resistor and the ressonant voltage is lower than the diode voltage avalanche, at certain D it may happen than during the resonance of Ls with the parassitic capacitance, all energy goes to the capacitance.

In the next half cycle the ouput diode clamps de voltage, and no more oscillations occur. So it seems that a damping resistor is not necessary, provided the ressonant frequency is high enough.

Hello All,

Where can I find formula to calculate resistor reset CT? seem  datashhet and application note are not discribe on how to calculate this resistor.

 

Thanks in advance.

NT