Offline SMPS charging up 30mF capacitor
Hello
I am designing a TOPSwith-jx , isolated offline flyback for 28W output power.
The 36V output has 30mF capacitor on it.
-this is needed to provide some power to the output in the event of a mains failure.
The overall spec is.
Flyback
Fsw = 132KHz
Pout = 28W
Vin = 85-265VAC
Vout = 36V
Feedback by opto
However, on page 5 of the datasheet , it seems that the TOPSwitch-jx would think it had an open loop, as the power supply is switched on and this ibig capacitor charges up?
The point is, that during start-up, it would take 1 second to charge up the 30mF cap to 36V…………….and since this is a long time, would the topswitch not go into fault mode, due to it “thinking” it was in “open loop”?
Do you know of any fixes for this situation?
-Also, do you know of any powerint.com PWM controller ICs which have features built in to handle this situation?
Comments
-but won't putting a cap there make the transient response severly slow?
-also, there is a resistor in series with the opto diode and this will severly limit the current through the opto diode when the Vout is low and rising, so there will be very little current in the opto diode at low Vout, as its rising up.
Also, the opto diode needs around 0.7V minimum across it before it will conduct at all, and with the big output cap, it may be more than 125ms before this happens .
so i cannot see how placing such a capacitor will help?
Hi,
You say the voltage is slowly brought up....the thing is, we don't want it slowly brought up....we need the 30mF cap to charge to 36V in less than 1 second.
do you have suggestions because from reading the TOPSwitch datasheet, the TTOPSwitch will think its got an open loop fault and go into hiccup mode, which is no good for us.
1 second counts as "very slow" from the point of view of an SMPS.
The "soft finish" cap will not make transient response slow because its ESR (or 100 ohm resistor I mentioned), makes its impedance high compared to the TL431 circuit's output impedance. IOW the TL431 will overwhelm the capacitor once the circuit is in regulation. The soft-finish capacitor technique is in many, many PI designs and has been extensively tested. I personally made gain/phase and transient response tests with and without the soft finish cap and it had no effect on them.
If your second statement presents a problem, then use a 2nd diode from the secondary winding (and a small cap) to feed the opto LED. Place the opto LED series resistor above, instead of below the LED. Place a capacitor from the output to the opto anode in order to provide higher frequency feedback (500~5 kkHz or so). Without this the PSU may go unstable. The capacitor should be sized so the RC time constant is around 200~500 Hz.
hi,
thankyou for your replies...
...from your last thread, do you mean like this..?.........
All you need to do is to place a very slow (large) "soft finish" capacitor right across the TL431. This way the loop is closed while the output voltage is still low, and the soft-finish capacitor will slowly bring the output voltage up in a controlled manner, and the current the power supply is delivering into your large output capacitance is below the power limit of the flyback, so that it doesn't go into auto-restart during startup.
This "soft finish" cap needs to have an ESR of perhaps 100 ohms, if it doesn't, then just add a series resistor.