Baking PC cards. That's amazing. I'm reading up on the melting temperatures of solder, and wondering how much overlap there is with actual component burnout (approx 70-80C if I recall). Where do the engineers get the max temp for overclocking from: burn out the component, or just start melting the solder around 280 - 340C...? Seems to be a HUGE gap. If oven baking works, why can't we OC our cards to 250C, just south of the lowest solder melting point?
Most IC’s are designed for lead-free reflow and are certified to tolerate temperatures as high as 260°C (lead-free solder becomes liquidous at 217°C). A hot air gun is likely to exceed this temperature.
Cheap ass reflow ovens (ovens used for baking electronic components on boards) use IR heating elements and they tend to burn components as black parts get hot faster. But in normal air driven or ceramic driven heating methods, the copper parts covered in solder, absorb heat and reach the melting point temperature much faster than the internals of the ICs due to the increased thermal conductivity of the material.
A good reflow oven is able to raise the temperature fast at high ranges. That allows to reach the solder melting point fast and not allow for internals to get damaged. Bad ones tend to take longer as the temperature rises allowing the temperature of the internals to reach critical points and get burnt.
The component burnout does not occur at 70-80'C. That is when the *SCIENCE* start.
Be advised the following content contains science:
A CPU is basically a collection of billions of on/off switches. The speed of a CPU is related to the structure of the switch. You could picture the rapid on/off switch as a binary pulse like this:
but in reality it is more like this:
since there is a tolerance level and anything above a certain level of voltage is considered on and below is considered off and there is some time required for the transition.
As we increase the frequency we need to decrease the time required to change states. This is why we increase voltage, to allow for higher tolerance within the limits by increasing the actual Vcc. However voltage increase, increases temperature and increased temperature increases resistance. And increased resistance makes it harder to switch states and increase frequency. And so on and so forth and I'm bored with typing science...