Blowing 2.5D-15 NTC Thermistor in 12v psu

[Sulrich]

This is for the really smart people (everyone but me) I have now blown 2 power sullied in 5 days and I have the brightness throttled down to 60%. I put my last spare online ordered replacement thermistors so I can fix the psu's that are currently offline. This psu powers outputs 25-32 in my f16v2. The only difference with these outputs and the first 24 is that the power and data lines are about 3ft longer. I have now reduced brightness to 50%. What can I do different? Any advice or input is appreciated

Thanks
Steve

[corey872]

I expect the thermistor is used as an inrush current limiter.  When you first turn the supply on, there are several relatively large caps which need to charge up, and you want that to happen in an orderly process.  So the cold thermistor has a high resistance, limiting current.  As current starts to flow, the thermistor warms up a bit, resistance drops and it is able to flow more current.  Finally at operating temp, it should flow the full rated current for the given power supply.

It seems two big possibilities might be:  1)  Drawing too much power through the supply output - which your mention of running 50% brightness and the power supply's own internal circuitry should prevent.  2)  The power supply located in an abnormally cold location causing the thermistor to remain in a cold / high resistance state.  The internal protection circuitry would see a 'normal' level of current, so would not trip.  But the thermistor would remain in a high-resistance state, with a higher voltage potential across it, possibly causing it to die.


...alternately, you could just shorten the data lines three feet! 

[Sulrich]

I leave my psu's on 24hrs a day and they are outside. But this one matrix board keeps popping the thermistor..

Luckily it looks like an easy fix and 5 of them are $4 on amazon with 2 day ship.

Just trying to limit the amout i have to fix them.

[corey872]

The "on 24hrs a day" should further eliminate scenario #1 I mentioned above ...you would only be dealing with 'inrush' for the initial turn-on.  Though the "outside" - especially if cold, would increase the possibility of scenario #2 being the culprit.

I suppose there is always the possibility you had some bad/marginal parts and replacing them in-kind might solve the problem.  You might also look at ways to keep the power supplies near 'room temp' if that is an option. 

If they are known to operate cold and long, meaning inrush is not a huge concern, you could look at other thermistor options.  A larger "D" number would be a bigger diameter (provided there is enough space on the board) and would be able to handle higher current...though the actual current may not be an issue in this case. 

A lower prefix number...2.0 1.5, 1.0, 0.5, etc would mean a lower resistance.  This is typically the rated resistance at 25ºC, and rises as the temp gets colder.   A lower number could offset some of the resistance from cold weather operation, at the risk of allowing higher inrush if the psu is turned on in a warm environment.