Most motor commutators shouldn't be subjected to large currents during DOL starting, so some means of limiting the current is required. If I remember correctly Bull Electric recommend less than about 3x full load current on a medium-sized machine. The old school method uses resistors: essentially you will have a number of resistors in series with the armature which are sequentially shorted out as the motor picks up speed. The timing between steps depends on how fast the motor accelerates. More steps gives smoother acceleration and better control of peak starting current.
How many steps do you have - four judging from the contactors? And what limit do you want on the starting current? If you have motor data then it's not too hard to calculate. Do you have the armature resistance and nominal current and voltage for the motor? You'll also need acceleration data for the load if you want to set the times accurately.
If you're re-wiring this starter be careful to ensure that the field is tapped off the full voltage supply and not from a point downstream of the armature resistors. Bad things happen if you tap off at the wrong point. Trust me.
Thanks for this. Ive got the original motor data sheet, so armature and field resistances. There are 3 stages, the fourth contactor is the last (but first to close) that acts with the overload to protect the motor should that be required. Are the calcs straightforward then?
I've attached a rough & ready method which seems to give useable results. I don't claim any credit for the 'calculation' - it was taken from a book much, much older than I am. I've typed it up for you because I don't have a scanner at home. The example is for a two-stage starter, but it looks simple enough to extend it to include a third stage. The supply was 120V - you will need to substitute your own voltage, current, and resistance.
At rest you have all starting resistors in series and the back EMF is zero, so you can apply Ohms Law to determine your overall starting resistance if you know the maximum acceptable current and the supply voltage.
For a two stage starter the author suggests the starting resistance is split into two stages of 75% and 25%, shorting out the 75% stage first. For a three stage starter he suggests 65%, 25% and 10%, shorting out the 65% stage first. The crude assumption made is that during each acceleration period the armature back EMF rises to a value equal to the voltage applied to the armature terminals at the start of the stage, which is not true. Purists will rightly say that this method has some glaring errors in terms of motor theory, but the approximations don't seem to result in a wildly wrong solution in practice - I found it was slightly conservative on a DC pump starter which I had to rebuild.
You could almost certainly model this in Mathcad and get a far better result: if anyone has time to try I would be interested to know how accurate (or not) the method in this old book turns out to be.
Using the method I described, calculate the required total resistance 'R' to limit the initial current to your desired value, then split R into three sections of 0.65*R, 0.25*R and 0.1*R. The resistance in series with the armature will be R during stage 1, 0.35*R during stage 2, 0.1*R during stage 3, and zero when the sequence completes. Calculate the peak current at each stage.