Hmm, some sexism going on here ? The OP did say sparky having
his first child
But I'll second what Ericmark says, compressor can be a right PITA to start. Many years ago (and long before I really knew what I was doing
) I helped a mate when he bought a new compressor - 2off 3kW 2pole motors driving separate compressors that started with a time lag. Oh what fun we had
Domestic supply to Wylex fuseboard in house, 30A wire fuse for garage supply, either 10 or 6mm T&E (I think it was 10) to garage (maybe 10m run), and another Wylex fuseboard. Ran 2x2.5mm T&Es for the compressor off a 30A fuse (yeah, some will be ahead of me here). If hot, and the receiver empty, it would fire up - one unit first, then a few seconds later (pneumatic delay on contactor in control box) the second would fire up. But from cold and with pressure in the receiver, it would more usually go : first unit fires up, second unit starts humming and creeps round a bit before the fuse blows
The difference being that when the second unit fires up, the supply is already carrying the running current for the first unit. Of course, not just one fuse, but both 30A fuses in series - so a trip to the house (in the dark) to replace that one as well. The house end quickly got upgraded to "35A"
which fixed that problem.
Machine Mart sent out someone to look at it, he fitted soft start valves* and after that it was much better - although it could hiss a LOT until there was a bit of pressure in the receiver and the valves shut.
It also improved after being used for a while, once the compressors were "run in" a bit. It was for grit blasting, and we could keep it running continuously for as long as it took to empty the grit pot.
The fundamental problem is that right when you need full voltage to get the torque for the motor to start up while also spinning up the inertia connected to it, and against back pressure - the high current causes a dip in voltage, and that slows the acceleration which prolongs the high-current startup period and gives fuses more excuse to blow. In some cases (as above) it may make the difference between starting and not starting at all if the starting torque isn't high enough to overcome the mechanical load attached.
* Soft Start Valve - like a poppet valve fitted in the line between compressor and the non-return valve to receiver. When the compressor is starting, it lets air out of the line and reduces the back pressure for a few seconds before it pops out and closes.
OK, so back to the OPs cable question. Without doing any calcs, a combination of things stand out - 40m run, 30A motor (I assume that's full load running current) driving a difficult load for starting, 4mm² cable. That to me sounds like a recipe for excessive volt drop during starting and starting problems.
Ah, now I've read the other thread, it's not 30A, it's 7.5hp which by my fuzzy calculations is only 7.5A/ph FLC - rather a bit of difference there. But still starting DOL which will mean a startup current that could easily be between 50 and 100A - IIRC a quick rule of thumb says something like up to 10 running current during startup. If the compressor doesn't come with unloading during startup then you probably want to add it - it will help considerably with both reliable starting and minimising the time the neighbours' lights dim each time it starts
Do you have agreement from the DNO to start such a large load ? Probably not, and in extreme if neighbours complain about voltage dips you can be told the stop doing it - or face being disconnected.
And as to the supply questions. As mentioned, you'll need to split the supply (set of "Henly" blocks), fit an isolator and overload protection (i.e. MCB in a case, or maybe a switch-fuse which offers better discrimination), and only then does the cable connect to that. At the other end, you'd split out 1 phase for all the single phase stuff. You may want to reconsider 30A & 4mm² cable - you might find it limiting as to the single phase stuff you could run. You'll need your electrician to do the design, it's not just current carrying capacity of the cable (which depends on cable construction and installation method) - there's volt drop, prospective fault current and let through energy (c.f. adiabatic equation regarding heating of the cable before the protection trips), fault impedance (to ensure timely tripping), discrimination between protections (see above about my mates problems with two 30A fuses in series), earthing (you may need 5 core from house to workshop so as to have a neutral as the wire armour probably won't be adequate as the CPC), ...
There's an iterative element to this. Making educated estimates as to cable size, then doing the sums and see if it works - adjusting the cable size to suit and going back to the beginning till you have a solution that meets all the requirements.