Originally posted by: arch_8ngel
Originally posted by: rlh
This is the first I have ever heard of this but being that I'm now not 10 years old, my semi-educated guess is that certain capacitors within the NES might abruptly release their stored capacitance when the NES is powered off. This wouldn't be a big deal for RAM but SRAM is much more sensitive to voltage shifts. I have no clue how old SRAM chips worked but they could have had voltage thresholds and if the voltage was ever to high, it might have tripped an internal safety switch that simply cut power to the whole chip so that it wouldn't blow by being overloaded. It would be better to lose some save states than kill the whole chip where save states wouldn't work at all or possibly brick the whole cart.
If my guess is correct, by pressing the reset button, those caps might have been reset in a way that would have allowed a safer drain of the power.
I don't think that would necessarily explain losing saves on the AVS, since I would have hoped Bunnyboy would have solved that particular issue.
Then again... we know that GB games of that same era obviously only required the power to be turned off (no holdable-reset available).
Primary difference between the two systems being a power supply that converts AC to DC versus being directly on DC/batteries.
Good point. That said, I've noticed that many NES carts also have a single cap on them. Maybe that cap has it's capacitance properly released on a reset, which would mean the issue is on the NES carts rather than the NES (or AVS) units. I've not paid much attention to the caps on NES games and I've not noticed when they were or weren't present. If they are only present on battery save carts, then there's a strong chance that this specific cap is related to the issue.
Also, tech changes with time. There's also a chance that the SRAM chips improved within the first half of the NES. This might have only been an issue with games manufactured, say, before 1988. To be safe Nintendo might have always suggested holding the Reset down, instead of making special rule. If the chips in the Game Boy games were of a better quality, it might not have mattered, plus they would have been aware that they would have only had one switch for the GB, so they would have manufactured it such that a power-off couldn't cause this type of issue.
I'm just throwing out extra theories. I'm still not good at my electrical troubleshooting and fixing skills, so it's all an "arm chair" educated guess, but I bet if someone looked into this who had the know-how, they could figure out the reason for this if they've not done so already. I do know that SRAM chips use to be (if not still were for many years after) quite finnicky. I imagine Nintendo had to make sure that they were properly treated during a power on and a power off.