With help from Philiegram's photos, I've made some "back of the napkin" models for the convective heat transfer for the Venty vs Mighty.
Interestingly, the Mighty is actually
more efficient than the Venty at convection. See:
In this model, a Mighty-like heat exchanger (black) is compared to the Venty (red) across volumetric flow rate from 0.1L/s to 0.5L/s. The Venty's relatively short convection channels mean that at high volumetric flow, the ratio of convection to conduction (also known as the Nussalt number) does not outscale mass flow and the energy required to heat the air fully. However, in practice, the Mighty has a weak heating element, so flower is wasted waiting for the exchanger to reach temperature. The Nussalt number for the Venty is about 4 (4:1 convective vs conductive heat transfer) whereas the Mighty is around 30. They sort-of balance out because the Mighty has far less exposed surface area for convection, however, at higher flow rates, the Mighty's Nussalt number is much larger due to turbulent currents from the spiral channel design. In general, most vapes that advertise a hybrid design are mostly conduction, objectively. It's very difficult to design a session vape that doesn't overheat, is inexpensive, and actually features majority convection. The TM2 is an exception, but from what I understand you can't keep it at temperature for long - it's really a challenge to keep the heat in the heat exchanger and not burn the shit out of the user.
On top of this, more than 70W is lost due to convection, even more considering conduction. This means the exchanger temperature rapidly drops as stored energy is not replenished. This is somewhat the case with the Venty, despite the advertised 130W output. This is in part due to the fact that most resistive heaters have a positive temperature coefficient which means they dissipate less energy as they heat up. But more importantly power scales with voltage squared and most LiPo batteries only put out 4.2V at maximum state of charge.
Looking at the
Venty's battery discharge curve, at half charge and 12.5A, each cell contributes ~3.5V. Working backwards from 130W @ 4.2V, we have a heater resistance of 0.55 Ohms (very close to the actual measurement of 0.6 Ohms). This means at at half charge the output is 90W and less as the batteries drain. The temp co for Nichrome is pretty small, about a 3% impact at 200C.
This is also why pass-through isn't supported. With a resistance of 0.55 Ohms and a minimum USB-C voltage of 5V there would be almost 10A draw which is way more than USB-C supports. It's technically possible to power pass-through with a buck converter but that adds complexity and the design isn't trivial.
The construction of the heat exchanger is likely aluminum, but this introduces a problem. The heating element (I think Nichrome) is wrapped around the exchanger, and there will be the full voltage drop of the batteries across the coil. If there is no insulation then the coil just shorts. I think they are hardcoat anodizing the exchanger for this purpose (also, a thin oxide layer builds on the Nichrome). The heating wire is crimped onto an adapter that is then soldered to the rigid flex PCB.
The same PCB houses some readout circuitry for the temperature sensor (thermistor) which is fed through a stainless tube close to the top of the heat exchanger.
The PCB also has a differential pressure sensor which measures the difference between ambient pressure and the pressure inside the PEEK shroud for the heat exchanger. It's quite expensive, about $20 at 1000qty. There is some delay between starting a draw and feedback due to temperature drop, but it seems like added complexity for little benefit as there is already a feedback loop controlling temperature.
The heater control is pretty standard - a series of P-channel MOSFETs switching on the high side. There are two FETs in the power path because you need to control power to the heater and also block charge and discharge currents. For this purpose, there's also an additional FET in series with the charger as a FET can only control power in one direction. A polyfuse protects from over-current and there's a shunt resistor for charge current feedback. Because the heater is effectively an air core inductor, they also need some flyback protection in the form of a diode.
The airpath is isolated through a silicone gasket, although that's not really a priority when it comes to designing a safe vape. Printed circuit board assemblies, when properly cleaned, do not appreciably offgas at room temperature. In fact, silicone outgasses significantly more at temperature (not all silicone is the same, too). Silicone is even shown to decompose into formaldehyde among other refractories and this decomposition is very temperature dependent. There aren't many (inexpensive) alternatives, so then it's prudent that the silicone temperature is limited in design.
S&B sticks to the mighty in the choice of materials: PEEK as a high temperature thermoplastic, a polycarbonate/PET blend (commodity plastic) for the outer casing, an aluminum heat exchanger (not 100% confirmed but likely), and silicone gaskets. I'd say that's better than 99% of vapes on the market. The electronics seem relatively well engineered, although there is zero technical reason why a removable battery wouldn't be possible. DIY battery refurbishment is also definitely possible - there are inexpensive spot welders available and it would even be possible to mod in replaceable batteries. For such an expensive device it's frustrating that it's essentially designed to fail once the batteries reach their lifespan. To me, the device isn't very cost optimized either.
