Yes, you’re correct, it’s based on the assumption of kV60 for the 70H, then assumption that kV is inversely proportional to stator length to 60*(70/90) which gives kv=47 for a 90H with same winding as the 70H.
Then an assumption that both motors have the same copper fill which means that the possible number of turns is inversely proportional to the phase wire area. Any good motor is wound to the maximum possible copper fill so this assumption is quite safe.
The assumption can very simplified be expressed that:
1turn x 100mm2 and 100turns x 1mm2 will make the same copper fill but a 100x difference in kV.
Why i don’t think they did anything else than make the stator and magnets longer is that for each stator width there’s an optimum magnet width. For each magnet strength (airgap field) and slot count there’s an optimum lamination design (this assumption is going out on a limb a bit) and i assume they use the same laminations for both motors also since the lamination stamping is tooled already —>
Same magnets spec, just longer.
Same laminations, just longer stack
If QS was aiming for the same kV with the 90H motor as the 70H it’s reasonable since then it will reach the same rpm on same voltage with the new motor.
I wonder if they use the new fardriver controllers in combination with the 90H to use the full power potential? This would be interesting: