---
title: "Peltier stacks"
question: "Does stacking two Peltier modules make it colder?"
url: "https://nearfieldnotes.com/videos/peltier-stack/"
site: "Near Field Notes"
video: "https://nearfieldnotes.com/media/peltier-stack/full.mp4"
duration: "2:53"
date: "2026-11-15"
note: "https://nearfieldnotes.com/notes/peltier-stack/"
---

# Peltier stacks (video)

Does stacking two Peltier modules make it colder?

2:53, with sound. [Read Note 41](https://nearfieldnotes.com/notes/peltier-stack/) for the full answer, with figures, conditions and sources.

## Transcript

One Peltier module on a fan heatsink at 12 V: −29 °C.

Stack a second one on top, both at 12 V: still −29. And it takes 112 W where one took 50.

Here's why, the two ways to make a stack work, and the one holding a camera on Hubble at −128 °C.

Last two videos, this module was a cooler and a generator. Now, two of them.

The top one draws 50 W. All of that heat, plus whatever it's cooling, lands on the bottom one.

So the bottom one is busy. The joint between them sits at 40 °C, warmer than the room.

The top module works down from there. Its full 69 degrees only gets back to −29.

Wire the two in series and it's the same story. Two identical stages on one current are one module with taller legs.

The top stage has to be the weak one.

Run the top at 5 V and the bottom at 12: −49 °C, on 61 W.

People who build Peltier cloud chambers use the same recipe.

Factory stacks do it with size. Each stage up has fewer pairs of legs, and one supply feeds them all in series.

This family has 127 pairs on the bottom, then 71, 31, 17 and 8. The bottom stage is our module.

With nothing on top, the datasheets say 70, 88, 104, 116 and 128 °C below the hot side.

I built all five from the one-module model. It lands within 2 degrees of each.

Here's the four-stage one with half a watt on top, at −82 °C.

The top stage lifts that half watt, and burns 2.7 W doing it. So the next one lifts 3.2.

Then 8.7. Then 24.

Half a watt in at the top. 52 W out the bottom.

That's why load decides the stage count.

At 1 W, the five-stage is no colder than the four.

At 10 W, one plain module beats all of them.

Stages save power too. Holding half a watt at −40 takes 59 W with one stage, 16 with two, and 6 with four.

Two things set the floor. Bismuth telluride gets worse as it gets colder. By −100 °C its figure of merit is under half what it is at room temperature.

And the top stage is tiny, so any leak hurts. Still air costs a four-stage about 18 degrees. So they're sealed in a vacuum, or a heavy gas.

The coldest catalog part is six stages: 133 °C below its hot side, with nothing on it.

So who pays $216 for four stages, when one is $24? Anyone with a sensor and milliwatts of heat.

Two stages in a four-gram can cool the kind of X-ray detector used in handheld metal analyzers.

Four stages hold infrared detectors at −78 °C.

And six hold Hubble's infrared camera at −128, with no liquid nitrogen to run out.

So a stack is a ladder. Every stage carries the load, plus every watt burned above it.

Which is why two modules at 12 V get you the same −29.

---

Source: https://nearfieldnotes.com/videos/peltier-stack/
