Test virtually a NTC Inrush Current limiter

TEST VIRTUALLY A VISHAY AMETHERM INRUSH CURRENT LIMITING NTC prior to go in the lab. in another work presented on Hackster.io, I have proposed a template allowing to build a QSPICE (tm) model af any thermistor available on the market. If there are NTC thermistors for which virtual testing is very important prior to go the the test bench in your electrical lab, it is indeed the type of NTC Inrush Current Limiter thermistor for SMPS or other power applications. I have analyzed the data sheet of a VISHAY Ametherm MS3510018 10 Ohms 18A, extracted the information from the data sheet and after the test in QSPICE from QORVO, I have confronted the simulation results with the rest of the available data from the same data sheet. Everything fits reasonably. You are then empowered to test it on the bench without blowing unnecessarily any fuse. Let's go step by step.

Let's extract the data from the specification :Ametherm MS35 10018 ---10 ohm / 18 Amp Inrush Current Limiter Data Sheet. We need the R25=10ohms+/-25%, the dissipation coefficient 78 mW/°C and tau 240 s and we note that the component is made with the curve I Resistance Temperature Curve | Ametherm

Filling in these values into the template (see Create a QSPICE model for NTC Thermistor - Hackster.io), we immediately get the QSPICE netlist of the part which we can copy/paste in QSPICE

Of course Qspice allows to record the I(V) curve at different temperatures and tolerances but let's compare the residual resistances at 25%, 50%, 75% and 100 % of the max current 18A (thus 4.5A:9A, 13.5A and 18 A) at Microsoft Word - MS35 10018rewrite.doc (Farnell website)

.meas r25 find v(vntc)/i(v2) at =4.5:

0.446445 ohms at 4.5 A (data sheet 0.4 ohms)

.meas r50 find v(vntc)/i(v2) at=9:

0.16284 ohms ay 9 A (data sheet 0.15 ohms)

.meas r75 find v(vntc)/i(v2) at =13.5:

0.0887377 ohms at 13.5 A (data sheet 0.1 ohms)

.meas r100 find v(vntc)/i(v2) at=18:

0.0573952 ohms at 18 A (data sheet : 0.05 ohms)

We fall straight into the +/- 25 %, that is the tolerance at 25°C.

We can conclude that the agreement between the spice model and the real measurement is remarkable; despite the simplicity of the thermal model. We are now empowered to build our final application and simulate with the confidence that the final practical measurements will meet the previsions.

PS: 1) the initialization of the model (normally 25° is done here in a special way at {TEMP} instead, since the dynamic external temperature is here much less important than the internal temperature increasing due to self heating.

2) there will never be too many spice models on the underrepresented market of thermistors. Our electronic world will never be SPICE-ified enough.