It could well be down to the delay that you have set between switch over, 40ms is two cycles at 50Hz but if your delay isn't accurate and takes into account the phase difference between the star and delta configurations your change over might be closing onto a reasonable voltage difference driving the increase in current and the kick.  I dont have any data on changeover times for conventional starters but my guess is that they could be shorter.  If you have flexibility you could try incremental adjustments (say +/-1ms)in the delay to see which direction reduces the spike and then optimise the delay in that direction to minimise the spike. Ofc grid frequency varies throughout the day and that will limit what you can ultimately achieve.  +/- 0.5Hz is allowed but normally its much tighter.  That would suggest fine tuning in sub 1ms increments would not be worthwhile.  Let us know how you get on.

One other possibility would be differences in switching times between phases.  You dont say what sort of switching you are using but any differences would produce spikes and kicks.

Hope this helps

Is the motor up to full speed before change over? Usually you can hear it. I'd lengthen the change over time to something like 5-6 seconds to ensure that the motor is up to full speed in Star before changing over to Delta. Also, are you starting up off-load? I.E. with no hydraulic circuits activated before change over? There should be a by-pass which dumps straight back to the oil tank when none of the valve spools are operated.

Motor is up to full speed in star before changeover. If we start offload the issue is reduced but not resolved.

The issue does seem to be from the new control panel, as I've hooked it up to one of the old HPU motors and the problem still exists. If the old panel is put back in circuit the problem goes away.

In answer to Satter, the old panels have a basic star delta timer fixed at 40ms with no issue but, to be thorough I've tries 20/40 and 60ms all of which have no meaningful effect

Do double check that the starter is wired correctly !

And not for example reversing direction when changing from star to delta. This may sound basic but I have found cases of this.

It does not have to reverse to be awkward, the standard wiring configuration is correct for hand over with a 60 degree fallback, (*) but it is possible to create a situation where the phases are advanced by 180 degrees, and the motor then has to instantly twitch ahead to a new shaft position, while rotating at the same speed and same direction.  If there is some dead time between the start breaking and the delta making it becomes less serious, as then the core has demagnetized and the delta case is just doing a DOL onto a motor that is already spinning but without a pre-magnetised core a rather  just more sophisticated version of pulling the belt and then flicking the switch.

Mike

* by way of explanation

the 3 phase to neutral voltages reach their respective peaks 60degrees offset from the peaks in the phase to phase voltages, but depending which phase to phase voltage you put across which coil there may be any multiple of 120 degrees to add to the shock the core feels.

So  when the stator coils are energized  in time with phases 1-2-3 in star, once switched you want them to be  energized  (1-2) (2-3) (3-1) The rotating core behaves like a permanent magnet going round a touch slower than shaft speed, so the currents in the core, and which parts of the spinning iron are the north and south poles are changing steadily at few Hz (the slip frequency). This does not like to be asked to jump violently.
Mike

Yeah, I've checked multiple times (going slightly crazy tbh) and the starter wiring is fine. I have had to change the motor direction at the motor terminals rather than just switch at the input, as the panel is for site use and we can't guarantee correct phase rotation. We have a phase sequence relay fitted, hence why we can't just do the simple fix. Here's the original and reversed wiring connections (in case I've made a daft mistake)

Original wiring
from Line Contactor  L1-U1  L2-V1  L3-W1    From Star/Delta Contactor  L1-W2  L2-U2  L3-V2

New wiring to reverse motor

from Line Contactor  L1-V1  L2-U1  L3-W1    From Star/Delta Contactor  L1-U2  L2-W2  L3-V2

The behavior seems odd.

When it comes to incorrect phase rotation on transition from start to run, the mcb or mccb should knock off in short order, especially if starting under load.

Changed out the Star Delta timer (fixed at 40ms changeover delay) for a Schneider unit with adjustable changeover delay (20 to 140ms in 20ms increments). Problem went away at 100ms; every other setting the problem existed to a degree.
Cheers for putting this potential issue in my mind. Oddly, I found that the very few posts I found with people having similar issues all involved hydraulic pumps

If you mean the gap between letting go of star and connecting delta, that will be the time for the core magnetisation to die off a bit and to slip round to be more or less in line with the delta derived stator field when it comes on. Pumps are a funny to predict, as they can have a very odd start-load torque profile depending how much fluid has to be got moving with various bypasses or load valves engaging as well.. Glad you have a fix, not too surprised it is a bit 'tweaky' .

Mike

In case you didn't see my solution (buried down a sub converation above), I changed out the standard star delta timer for an adjustable one and changed the star-delta delay from 40ms (as per the original starter) to 100ms and the starter is working like a dream. Thanks very much to you for your input