Transformer soft start

[alternety]

I am building a line operated DC power supply for a 48V DC pump designed for battery operation. The power consumption will range from about 250W to 500W depending on water level. A secondary objective is to keep power consumption low to minimize battery backup/inverter loading, and to minimize turn-on surges.

For best efficiency I have selected a 1 KVA toriod power transformer by Plitron. This size puts the power consumption range in the high efficiency portion of their performance curve.

They warn of exceptionally high inrush current because of the charicteristics of the core and recommned doing something about, it but provide no concrete advice.

I am looking for a soft start approach for the primary side of the circuit. I have seen two approaches; a thermistor and a resistive load disconnected by a relay (or thyristor - relatively high power loss) after a few cycles. Neither of these would protect against a very short power cycle and both involve power losses. Thermistors/thyristors while the pump is running and the relay all the time. I have seen some clever approaches from the secondary side but that does not fix the transformer charging current.

Any suggestions for an approach?

 

[cbarn24050]

Hi, use a triac to apply the power, fire it 90deg after the crossing point.

 

[Chuckles159]

Put a power rated NTC resistive device or failing that just a power resistor in series with the primary. They make special devices for this function. I use them all the time. Add a relay with contacts across the series device so that the device is bypassed when the relay contacts close. A simple timing circuit can be put on the secondary side to energize the relay after a few seconds.
Don't use a triac as they generate EMI and you will need to add line filtering and conditioning. Besides, a relay is just as cheap and more reliable.
The approach I suggest is used on most high powered professional power amplifiers, like the ones I design, bacause the charging of the filter capacitors cause a huge inrush current draw otherwise.

 

[Paulcet]

I once had an application where the transformer inrush current was a problem. We used 'current limiting' fuses in place of the regular time-delay fuses. Maybe that would help?