Free engineering calculator

Device / Spot Cooling Calculator

Size a thermoelectric cooler for a single device: enter the device heat load, its target temperature, and the ambient - then pick HOW the hot side is cooled (fan sink, passive sink, boundary wall, convecting panel, or a water-cooled plate). The solver couples the TEC to the sink and finds the operating point - or tells you it runs away.

▶  Watch this problem SOLVED - live animated transient

Q_c (load)device · T_cTEC moduleΔTT_h = T_amb + riseQ_h = Q_c + P_inT_ambhot side: fan sink / passive sink / wall / panel / water plate

The diagram is labeled with the same symbols as the input fields below.

Watch it solved transiently

Live pulldown simulation of YOUR operating point

A real 22-node transient network solve running in your browser, seeded from the inputs above. Watch the cold front sweep through the device while the heat wave propagates into your chosen hot-side sink - and land on the calculator's steady point.

The full engine

This preview solves a handful of lumped nodes. The NovaThermal engine behind ThermalResults.com (coming soon) runs the same physics on tens of thousands of nodes - full transients with phase change, radiation, fluid loops, and Monte-Carlo design envelopes, GPU-accelerated at 400× real-solver speed - and hands you review-ready margin reports.

The equations this calculator uses

Effective module properties from the datasheet (T in kelvin): α = V_max/T_h,   R = V_max(T_h−ΔT_max)/(I_max T_h),   K = V_max I_max (T_h−ΔT_max)/(2 T_h ΔT_max)
Q_c = αIT_c − ½I²R − KΔT  ⇒  solve the quadratic for the drive current I (efficient lower root); P_in = αIΔT + I²R;   Q_h = Q_c + P_in   (iterated with the sink)
fan sink: T_h = T_amb + Q_h R_rated √(CFM_rated/CFM)  ·  passive sink: T_h = T_amb + Q_h R_na
boundary wall: T_h = T_wall + Q_h R_contact  ·  panel: Q_h = (h_conv + h_rad) A ΔT_panel,   h_conv = 1.42 or 1.32 (ΔT/L)^¼
water plate: T_h = T_in + Q_h / (ε ṁc_p) + Q_h R_contact,   ε = 1 − e^(−UA/ṁc_p)
Q_max(required) = Q_c (1 + margin) / (1 − ΔT/70)
Assumptions and limits
  • Real single-stage module physics via the standard effective-parameter method: alpha, R, K derived from your datasheet I_max / V_max / dTmax at the rating temperature, then assumed temperature-independent (the usual first-order treatment; properties drift a few percent across the operating range).
  • Defaults describe a generic 127-couple 6 A module (V_max 15.4 V, dTmax 68 K at 27 C rating) - typical of the common 40 x 40 mm class and well matched to the default load and fan sink. Enter YOUR datasheet numbers for real work.
  • The drive current is solved each iteration to pump exactly your load (efficient lower root); the solver refuses honestly when no real solution exists, when I exceeds I_max, or when the operating dT reaches the module's dTmax.
  • Fan sink: datasheet resistance scaled to your flow as R ~ CFM^-0.5 (turbulent-film scaling); passive sink: use the manufacturer's natural-convection rating, fins vertical.
  • Convecting panel: simplified flat-plate natural convection (1.42 or 1.32 (dT/L)^0.25) plus radiation at emissivity 0.85, one face active; drafts and nearby surfaces change this.
  • Water plate: effectiveness form with your UA and flow; coolant properties from the selector at room temperature. Contact/spreading resistances are user inputs - grease and flatness matter.

Engineering notes

Thermoelectric coolers are the standard answer when a device must sit BELOW ambient or needs tight active temperature control: laser diodes, image sensors, reference cells, small chambers. Their catch is that capacity falls almost linearly with the temperature difference the module must hold, and every watt of electrical drive becomes additional heat the hot side must reject.

That is why this calculator reports the heat-sink requirement, not just the module size. The rejected heat Q_hot is often 2-4x the device load. If the sink cannot hold the hot side near ambient, Th climbs, ΔT grows, capacity collapses, and the controller responds by driving the module harder - a runaway loop that ends with a device WARMER than it would be with a plain heat sink.

Use the margin input honestly: 20% covers thermal-interface tolerances and module aging in most indoor applications; sealed outdoor or high-reliability designs typically carry 30-50%. And when the duty cycle matters - pulldown time, cyclic loads, controller tuning - a transient simulation of the device, TEC operating curves, and sink as one network is the step beyond this estimate.

Frequently asked questions

Why does my TEC make the device hotter instead of colder?

Almost always an overloaded hot side: the sink cannot reject the device load plus the TEC input power, so the hot-side temperature rises, the required dT grows, and module capacity collapses. This calculator iterates that exact feedback loop and reports THERMAL RUNAWAY when your chosen hot-side cooling cannot hold the operating point.

Can a single-stage TEC hold a 70 C temperature difference?

Only at essentially zero heat load. dTmax (about 68-72 C for standard single-stage modules) is the zero-load limit; at half of dTmax you have roughly half the module's rated capacity. Beyond about 65 C, use a two-stage cascade.

How much electrical power will the cooler draw?

A practical estimate is COP of about 0.5 at small dT falling toward zero as dT approaches dTmax. This calculator uses COP = 0.5 x (1 - dT/70): a 25 W load at dT = 25 C costs roughly 78 W of drive power - all of which the heat sink must also reject.