I have an old (but very useful) Linksys E4200v2 that has long suffered from the will-not-boot-while-warm problem that is mentioned here. This condition continues to grow worse with age, to the point that external fans no longer help. The device now has to be placed in a freezer for 5-10 minutes every time it needs to boot.
With the failure so easily reproducible, I decided to investigate and was hoping that a minor, replaceable component might be the cause. Using a can of freeze spray and working down to tiny drops on individual components, I have some pretty strong evidence that the Hynix TSOP flash is to blame.
That's all for now. Hope the information helps someone.
have you check the capacitors relating to regulators or filtering
I would check this & voltage
while it won't boot add a an extra capacitors to the CPU supply rail
I would try adding a 10uF tantalum this would cover mono's & electrolytic on VCC
or a say 100uF electrolytic on the regulator supply side
I would use this temperately to see if it fixed the problem
and then chnage the main ones if proven to be faulty
Knowing that electrolytic capacitors cause a significant number of problems in aging electronic equipment, they were the first thing that I checked. I performed a visual inspection of vent seals and tested them in-circuit with an ESR meter - all nominal. I also focused on them with freeze spray, to no avail.
Also, capacitor failure doesn't fit the symptomatology. An out-of-spec capacitor would tend to be worse when the device is cold and better as it heats up. In addition, this problem doesn't just occur when the device is being powered on - it also manifests for software-initiated reboots long after the power circuits are stable.
Nevertheless, I did spend a good deal of time probing and chilling the individual components of the power circuits including the Texas Instruments 54231 DC-DC converter. However, I didn't find any evidence that they contribute to this problem. The operation of the main LED also suggests that power isn't an issue.
I have seen lots of circuits where a monolithic capacitors are faulty
these can show as the circuit get warn
if you do not have functional high frequency filtering the regulator brakes out into oscillation
this is confounded by he fact that when you add test equipment the capacitance of the leads alters or fixes the problem so it looks like it working
tho I do admit this usually heat up the voltage regulator another possible sign
anyway these are the little capacitors located at each device there are lots of them
but they usually all come in the same batch & with age tend to turn from capacitor to resister over time
anyway I don't think this is your problem but something people have to be aware of as a possibility