I remember back a few hundred pages ago we postulated that the OG has some TCR determining coil temp from resistance and PID like software to maintain temp under load. I don't think it's as simple as a potentiometer, inhale strength would then play a much bigger role in performance.
It's sounding like the reliability is based purely on the usage model... which means its quite possible for the user to damage the TM via overheating. The PCB doesn't look like a purely discrete solution, so bad things can happen above 150'C... by the time you hit 240'C or so solder can start to reflow. There's going to be a temperature gradient, but I wonder why session mode hasn't created more problems. I'm imagining that some output driver, say a current source (it could even be a trim DAC) is driving the heater (load) I think you are right that headroom has to be given so this driver can product more current under load (inhale). I believe the original poster (
@WhyWhyWhy? ) was inhaling during his temperature measurements to simulate these load conditions.
I say all this because it means that under normal session mode (no load) the device probably sits near the rated temperature. However, the load (inhale) will create more current.. creating more heat... until the device eventually is damaged. I'm not saying WhyWhyWhy? is seeing the results because he caused damage... I'm saying he will eventually damage his device because the only protection is the usage model... but TM doesn't give guidelines on how much thermal guardband is left if any. So I believe he had a valid point. And the response is also correct... the TM isn't robust enough to regulate the temperature under any conditions... so it's based on the usage model.... if it feels hot, best to stop. It's not a robust solution but seems to be holding up over the years or the warranty takes care of it.
Since we aren't inhaling the whole time in session mode (no load condition)... the device stays in a reasonable temp range and that is why damage isn't frequent. However the potential exists for the user to damage the device by doing long inhales, and there is no thermal monitoring to protect the circuit. With configuration states, led modes, calibration this is all much easier done in the digital domain. I thermal monitoring circuit would help reduce the reliability concerns... even something simpler like a short circuit protection ckt could help limit the temperature. Thermal monitoring can get complicated, so I don't think we'll ever see tight temperature control... but a reliability-based temperature limit is reasonable. i hope this is easy enough to understand.. rushing for dinner