Converter In-Rush Thermistors etc, UPDATE -4
UPDATE on 17 Nov, Update 2 on 22 Nov, Update 3 on 16 Jan 14, Update 4 on 19 Oct 14. --------------- In a recent thread there was mention of how you can wreck your gizmo (in my case a converter) ...
Forum Discussion
ken white wrote:BFL13 wrote:
So my 5R020s are ok, my 2R025s are not ok, but if I use two 2R025s in series to get 4R that would be ok? (or should that be parallel? I never did get that straight)
I have to go for now, but series only since NTC devices will not share the energy transfer equally when in parallel.
x2. Series ONLY!!! NEVER parallel for NTC thermistors.
And consider this. As long as the thermistor can handle the inrush current (joules spec) and the power at max steady state current, a higher than "optimal" thermistor initial resistance within reason, really will not hurt anything and has upsides. inrush current will be lower (good), time for the caps to fully charge will be longer, but only by a few cycles, not noticeable. No real downsides.
Lower than "optimal" by contrast will result in higher than desired inrush current, higher stress on everything downstream and no real performance advantages.
And if I was picking a thermistor, once I determined the thermistor value I wanted, I'd pick one that can handle about 2x the Joules on inrush and 25% or so more steady state current. Don't want to go overboard on steady state current margin, since you do want the thermistor to get hot enough to have low resistance at high current, but don't want to operate it a max power either.
And for an initial target as I SWAG'd earlier, I'd limit inrush to 2x maybe 3x worst case over max current at load. So around 30 to 45A, especially since we have limited knowledge of the rest of the circuit and PCB.
So I'd go with Ken's "5" recomendation.
The tutorial is a good overview, covers basics, but is not an in depth lesson on use of NTC thermistors.
As I have mentioned previously, when I worked in T&M design, power supply designers spent more time than you might think determining the best choice of NTC thermistor values, power handling capability and handling of "short cycle" conditions to end up with a solution that gave long life, high reliability, for the thermistor and power supply. It usually involved a fair amount of computer simuation of the input circuit over a variety of power on conditions followed by some rather rigorous worst case testing.
Something we don't have access to us here, so better off being conservative.
