Liquid Cooled CPU Project

After wanting to liquid cool my CPU for years and after finding I knew what I had to do!

This project was going to be a lot of work so I wanted to make sure that I had direct evidence that the liquid cooler was actually doing something. So, I put together a digital temperature sensor array that I could use to record the performance of my liquid cooler. I used Dallas Semiconductor ( DS1820 sensors and a Microchip ( PIC16c55 microcontroller to do the job. I recorded temperatures at multiple points at 10 second intervals.

At this time I have pictures and brief descriptions of my Phase I and Phase II units as well as lots of cool graphs of my temperature data.

I managed to cool a temperature sensor to -44.5F during my cascade experiment. Check out the link below. Watchout Kryotech!

Phase I

Phase II - 30F (-1C)

Phase IIa - 11.75F (-11.25C)

Phase IIb - Jet-Dry

Phase III - In Design Stage ** NEW 11-12-98 ***

Temperature Data

Refrigerator Experiments - [8-10-98]

K6-2 Overclocking Predictions

I've attempted to compute the wattage and target CPU temperatures for the K6-2 running at different core and bus speeds. Check out my predictions.

Next Step:

Collect Temperature Data for:
  1. Cooling without Peltier
  2. 40mm (1.57") 3.9 amp 33.4 Watt Peltier
  3. 40mm + 30mm Peltiers Cascade Configuration
  4. 2 40mm Peltiers in Parallel
  5. 2 40mm Peltiers in Parallel + 1 30mm in Series

Question? Comment? Suggestion? E-mail me:

Totally off-topic: Two pics from my back yard in La Jolla, CA: here. And if you REALLY want to get off-topic you can check out my NUKE'O'TRON Nuclear Weapon Effects Calculator and my Titan ICBM page @

All images and data Copyright 1998 Larkin Lowrey