Soft-start

Thread Starter

Iodem_Asakura

Joined Sep 14, 2004
140
I was reading about the soft-start technique used to drive transformers, but i have some doubts. Have anyone some good information about it?
 

beenthere

Joined Apr 20, 2004
15,819
Hi,

Not sure this is good information, as I have no idea of the size and function of the transformer in question.

However, by observation over time, I have noticed that it is common for a transformer core to retain a bit of magnetism. If it happens that the residual magnetic field is in a direction that fights the field that wants to build under the first application of power, then the initial surge current will be significantly larger tahn expected. Even a small power supply transformer (for a linear supply) can draw over double it's expected current, on occasion.

The soft start frature may have this little gotcha in mind. It will certainly allow the transformer's reactance to be fully in synch with the driving current before full powe(and draw) is reached. It's one thing to pop a 3/4 amp fuse, and quite another to have to interrupt 25 KVA because of a start-up surge.
 

bipin

Joined Mar 21, 2004
80
In normal context there require no softstart for a transformer as it is an inductive device and the primary itself will alow a slopy current rise even given with a step voltage as input. Its slop depends on the L/R time constant.
Inrush current and surge current are not a big constraint unless its a heavy trafo and (where coincidental alignment of residual magnetism in the opposite direction and causing a little higher starting current which lasts for one cycle of the input ) you dont desire any abnormal current behavior even in the first cycle.

Soft start is very crucial as far as motor drives are considered. You need to power up the motor slowly untill Eb(back emf) is generated.
 

Thread Starter

Iodem_Asakura

Joined Sep 14, 2004
140
Thanks for the information, i know what you said me, but have some doubts. I think it was better if i put my doubt all along (sorry :p ). So...
I only need the soft-start at beggining of the entire opperation, after the transfo has been at rest?

What happen when i have to switch the transfo with a wave like this? Here i need to switch the transfo with a square ac wave with a dead time between positive cycle and negative cycle. Do i need to use the soft-start every positive and negative cycle? Would it be the ideal thing? It is frecuently used like that?
.........____
........|.......|
........|.......|
____|.......|_____........._______
.......................... |........|
.......................... |........|
.......................... |____|

Very thanks for your help
 

bipin

Joined Mar 21, 2004
80
Hi Now the things are clearer,
You only have to make sure that your on time of the wave is not exceeding to such a level that the transformer core starts saturating.
If its a 50Hz(60Hz) well designed transformer then you can easily give 10ms(8ms) width to your square wave.

but I feel you are going get caught by the assymetry of the two (+ve,-ve) half cycles of your wave. If there is a little dissimilarity with these two half cycle and then this will cause a small average DC current through the trafo and based on the input RMS voltage of the square wave and the total Ohmic resistance of the transformer input circuit path, your transformer will start developing stairecase saturation. mmmmm..... Thats what I feel.. because once I had it, when I did some thing similar.

Good luck :)
 

bipin

Joined Mar 21, 2004
80
:) hi that looks as a useful link.

Asakura,
off hand i dont have the formulas which you would reuqire. one thing I can tell you is the maximum on-time is dependent on your L/R value of the primary, the peak voltage of your square wave, and the maximum saturation current that your input of the transformer can with stand with out going in to saturation.

Practically you can find out the maximum on time using an Oscilloscope.

What you have to do is apply a square wave (the wave that you are intending to inject) to the inductor through a low value resistor(1-Ohm or 10-Ohm) and observe the wave form across the resistor it should be a triangular wave. if there exists any overshoot distortion in the triangular wave that means your inductance is saturating, if not then you can go ahead with your project for that particular wave form.

if the transformer works for one positive and negatice pulse with out saturation and both the pulses are symetrical then, I dont think you will have any problem in continuously operating it with the same square wave.
 

Thread Starter

Iodem_Asakura

Joined Sep 14, 2004
140
Thanks, that's a really good information.
But i still have a doubt. What's the problem if the positive and negative pulse are not simetricals, if the width and amplitude of each pulse don't saturate the transfo? :huh:
 
Top