Vapochill PE

Extreme cooling phase change CPU cooler – Joe

FIRST LOOK: Well made, very good software controls.


Vapochill comes in two colors (pic courtesy of Asetek).

The good guys at Asetek were nice enough to send a sample of the Vapochill PE. Asetek is one of the first to manufacture a phase change (refrigerator) case and is the winner of an Overclocking award we gave out about four years ago. I’m going to first preview the system and then fire it up for testing next week.

Vapochill comes in two flavors – the SE (Standard Edition) and PE (Premium Edition). The primary difference between the two is the ability of the PE to handle 160 watts of “maximum stable load” while the SE tops out at 130 watts:

SE vs PE

Graph courtesy of Asetek.

This is due to a beefier compressor for the PE, which translates to potentially higher overclocking. Note that temps average about 5C lower with the PE in addition to its extended CPU range:

What’s Inside?


The Vapochill houses the compressor unit at the top of the case. The air intakes for a 120 mm case fan are OK, although I would have preferred a totally unobstructed 120 mm circular opening instead.

Stripping off the sides and top reveals the chassis:

PE Side

It’s equipped with bays for 5 HDs, 3 CDs and one floppy. I was impressed with the compactness of the refrigeration assembly. If desired, another 120 mm fan could be added to the rear – plenty of room for it. Note that the Power Supply sits in front of the CPU and cooling head (red dashed area).

I removed the motherboard tray (two screws) to better show the back side of the case:

PE no Tray

The CPU cooling head is routed through a hole in the case top; the “bristles” you see of the head are nylon ties used for shipping. On the left hand side, a little less than midway up the front, is the controller PCB:


This is the heart of the unit’s functioning.

A view towards the front of the case shows the drive bays – note that a 120 mm front fan will cool hard drives very nicely. There is a fair amount of metal and plastic in front of the case fan – removing some metal will increase airflow a bit. The plastic front bezel is screwed onto the metal frame – no pop-off here.

Case Inside

A closer look at the refrigeration unit:

Top 1

The radiator is not bolted on and actually “floats” a bit above the case floor – there is a strip of sticky tape under it, but it did not hold it in place (note: this is NOT the way systems are shipped – Asetek bolts the radiator to the case; apparently this was a test that mistakenly was shipped to me). A look from the other side

Top 2

shows the capacitor (extreme right) – this is needed to boot up the compressor from a cold start. This is a Danfoss BD50F – a 12 volt unit with a variable speed motor. This is significant as once the unit hits a preset temp, the motor can reduce rpms while holding the temp.

The cooling head


features a copper core that contacts the CPU’s core. Asetek has an exploded view of its assembly:

Head Diagram

Pic courtesy of Asetek.

The “clamshell” encases the copper cooling core in foam for insulation. On top of the core clamshell is the bracket used to bolt the cooling head to the motherboard.

Shipping with the Vapochill are all the parts required:


Included are:

  • Foam insulation for the CPU socket and back
  • Cooling Head mounting hardware
  • Two heating elements (for socket moisture control)
  • Communications cable for changing system settings
  • Boot Disk for system setting software
  • Power cable for PCB
  • Case hardware
  • Thermal grease
  • ATX connector plate
  • Additional thermal probe (optional)

In addition there are three manuals:

  • System Manual – 29 pages
  • Chill Control & Configuration Utility – 17 pages
  • Installation Guide Socket 478 – 26 pages

The manuals are clearly written, well illustrated and absolutely required reading.

Of special interest is the “Chill Control” PCB. This allows the user to customize many features of the system, such as:

  • Change temperatures for start-up and shut-down
  • Setting to keep a steady temperature on the CPU
  • Fan speed adjustment: 30 – 100% power
  • Compressor speed Adjustment
  • Adjust ChillControl shut-down temp
  • Adjust heating elements power

To access these features, you plug in its serial cable (included) to the PCB and plug it into a COM port on the PC. Insert the Boot Disk, power up and it boots into a software program which programs the Chill Control unit. For example, the factory pre-set is for the PC to power up when the -5C; you can change this either higher or lower, depending on your preference.


Just reading the material and looking over the system, you get the impression of quality and well thought out features. Considering that Asetek had been around for a while, it does appear that they’ve learned some things and developed a nice feature set.

I am intentionally writing this before testing to see if my initial impressions hold up – stay tuned!

NOTE: North American pricing has been reduced; prices now range from $469 to $559:

North American Pricing


Email Joe


The directions for setting up the Vapochill are very clear and explicit – each step is explained and there are plenty of pictures to guide you through the process.

Setup involves a number of distinct steps to ensure that there is no moisture condensation in the CPU area. Asetek includes foam and heaters for this purpose – setting up these components is not difficult nor particularly time consuming. I took about two hours to do it, taking my time, reading the directions and taking pics along the way.

First set of steps involves placing insulating foam around the socket area:


This involves removing the P4 retention bracket, replacing the usual plastic motherboard mounts with screws and nylon nuts, and then placing insulating foam. You also insert the two steel bars through the bracket’s holes – these are what hold the cooling head to the CPU. Easy enough to do.

Next, place one of the heaters on the motherboard’s back


and then stick a foam pad over it – one side of the pad has sticky tape – peel off the backing and place the pad over the back heater; you get this:


Next, there’s another foam layer that’s placed around the CPU socket and a second heater:


Both these heaters are designed to ward off moisture and can be configured through the Chill Control unit.

What I did not do is grease up the socket pin holes with the “Heat Conduction Compound” that comes with the Vapochill – I’ll do that for the long-term test, but for now I skipped it. Users should not skip this step. I used Cooling Flow thermal grease between the cooling head and CPU case top.

Having done all this, next step is to mount the motherboard into the case. I do not like the standoffs for the motherboard:


They are very easy to use – that’s not the problem. If you want to remove the motherboard, you have to go to each standoff and pinch each one while moving the motherboard up a bit to clear the top part of the standoff (so it does not lock). Not a show stopper by any means – just a tedious way to remove a motherboard – but I would much prefer screws.

Once the motherboard is in the case, you then have to align the cooling head so that it mates with the CPU core. This may involve bending the stalk to get things to line up. I probably spent more time doing this than any other step.

Once the cooling head is lined up, then you bolt it onto the CPU with two screws which thread into the two steel bars in the retention bracket. This is the tricky part.

In order to ensure that they are exerting equal force on the CPU, the Vapochill manual states the “the screws should protrude 2-3mm on the other side of the bars”. Now look closely at this pic:


There is no way I could see the screws as they threaded through the steel bars. I finally used a small mirror (my wife’s) to look at the underside of each bar to see where the screws were. Maybe some others are physically more flexible than I am, but not a show stopper, just something to plan for.

The installation overall I found relatively painless – this is not rocket science, just take your time and follow the directions. I can’t see more than three hours, starting cold.

And then…

A funny thing happened while performance testing….page 3

Email Joe

Having followed the steps, I then hooked up the power supply and case switches to the motherboard, per instructions, and fired up the Vapochill. I was pleaseantly surprised that the compressor is virtually silent. I doubt you would hear anything even if the 120mm radiator fan was off.

The Vapochill did what the manual said it would do – since it was starting for the first time, it would take a couple of minutes to reach the motherboard start-up temp of -5 C. This is to allow the system to power up to its operating range before the CPU fires up.

The LCD display on the case blinked for a while, then showed head temps declining as it reached -5 C. and then…


The motherboard did not fire up; it received power, as I see that the fan on the video card powers up, but no boot.

Hmmm… What did I do wrong here? Now what follows is an extended debugging period. Remove the cooling head to make sure it’s on the CPU – no problem there – the CPU imprint is fine and it’s cold.

OK, maybe something with the motherboard? I hook up a UXD diagnostic card, power up and see that BIOS does not start when the Vapochill hits -5 C. At this temp (which you can change), the Chill Control unit signals the motherboard to boot up. Somehow, the motherboard is either fritzed or not listening to the signal.

I unhook the Vapochill cooling head, install a heatsink on the CPU, hook up the PS to run it without the Vapochill and BAM! It boots up no sweat.

OK – not the mobo, the problem is with the Vapochill.

At this point, I should mention that the motherboard is an ABIT BD7II.

If you cruise around Asetek’s site, you’ll see a page about Supported Motherboards. Asetek has found that ASUS motherboards seem to work fine with the Vapochills and recommends using them. As to others,

…we have been reported problems on certain motherboards. If you want to be free of these possible errors you should go for the [ASUS] motherboards above.”

Emails back and forth to Asetek about this. I also check into Asetek’s Forum (very valuable – a must if you are considering a purchase) and post an inquiry about my problem. Cruising around the Forum, it turns out that other users have experienced similar problems with Abit and Epox motherboards; the user workarounds are varied and interesting in their ingenuity.

I try different workarounds, but to no avail. Finally what I did was use two motherboards – one hooked up to the Vapochill with its own power supply. This mobo has nothing on it – no RAM, CPU or Video – a “blank”.


Circled in RED are the Vapochill Power-On and Reset plugs.

I hooked another PS to the motherboard in the Vapochill with a separate Power On switch.

The Vapochill boots up per normal when hooked up to the “blank” motherboard while chilling the CPU in the case. Then when it hits -5.0 C, I power up the mobo in the Vapochill using a separate Power-On switch. Works fine! Not a long term solution, but it will enable me to do detail testing.

Asetek is working very hard on this problem as you read this, so we may have some resolution on this shortly. However, if you are considering a Vapochill purchase, a non-approved motherboard may be an issue.

Stay tuned – I will post performance tests shortly. I can say right now that it does what Asetek states – the cooling head is in the -25C to -30C range.

Email Joe


For testing, I am using an ABIT BD7II with a P4 1.6 A; I’m waiting for general availability on the P4 C1 steppings and will buy a 1.8 to get something like 3+ GHz.

During my testing, I booted up the Vapochill, set up a test, left for a while and came back after about 10 minutes and found that the fan was not working. The compressor was warm to the touch and the LCD showed a temp of -10 C – not bad considering I was running Prime 95 and this was passive cooling!

The odds of a fan dying are very low; curiously, I saw in the LCD display that the fan is turning 2000 rpm – obviously something was askew. I quickly shut down and try to figure out what was going on, checking connections, etc. I found this:


The red wire is to the fan, the white to the Power-On switch.

The insulation on the red + wire to the fan was cut. The hole through which it passes from the top section to the Chill Control unit does not have a rubber grommet around the edge, and as a consequence its sharp edge cut into the insulation. I also found that the Power-On wire to the case switch was also cut.

Power to the fan was shorting out on the case, sending a false reading to the Chill Control unit. I taped it over and it works fine. I would suggest to Asetek that a rubber grommet would be a nice addition to future Vapochills, and strongly urge purchasers to check these wires and retrofit a grommet to be on the safe side.

On to results. I ran the RAM at conservative settings to see how far I could push the P4 1.6 A. This is one of the very first ones to come out and has never done much more than about 140 MHz stable with water. A look at the CPU database shows upper end results in the 3+ GHz range.

I was, of course, interested in CPU temps and was able to do the following:



LCD Temp

CPU Temp

Temp Delta

IDLE, 1.66v, 87.4w


-26.5 C

-3 C

23.5 C

Prime 95, 1.66v, 87.4w


-17.0 C

10 C

27 C

Prime 95, 1.44v, 43.1w


-23.0 C

0 C

23 C

CPU Temp per MBM P4 on-die.

This is a 50% overclock – best I’ve ever done with this CPU by about 160 MHz. Note the difference between that Cooling Head and P4 on-die temps – this is due to the location of the Cooling Head thermistor vs the P4 diode and thermal losses due to thermal grease interfaces – about 25 C.

Of particular interest is the difference between running at spec and running flat out while radiating twice the power: A CPU temp difference of 10 C. I compared these results to those published by Asetek:


The red X’s mark my results.

Pretty much right on what Asetek claims for Cooling Head temps. Based on my results only and Vapochill PE readings from the graph above, I estimate P4 CPU temps as follows:

Estimated P4 On-Die Temps

CPU Watts @ Load

Vapochill LCD Temp

Est. P4 On-Die Temp

Temp Delta

40 watts

-23 C

2 C

25 C

60 watts

-19 C

6 C

25 C

80 watts

-16 C

9 C

25 C

100 watts

-13 C

12 C

25 C

120 watts

-10 C

15 C

25 C

If you have a Vapochill, I would appreciate sending me your P4 results so I can compare them to my results and report on them.


No question the Vapochill is going to wring the most you can get out of highly overclockable CPUs. Whether or not this is an economic decision is up to each user to decide. On straight economics, a Vapochill will allow users to get high end performance from low end CPUs at low end prices.

The Vapochill PE’s compressor is virtually silent, and there is no discernable vibration from it. The only noise you will hear are from fans.

Motherboard compatibility is a major concern which potential buyers must address. If you have a motherboard that’s not on Asetek’s approved list, odds are you will have a problem.

Overall, the build quality is good, the case OK and noise levels are low. Asetek’s warranty is two years. I have used a Kryotech for almost three years without a hitch – refrigeration technology is very mature, and with a quality compressor, a system such as this will be around for a long time.

Finally, closely examine your unit by removing the top cover to check for possible frayed wiring, per my comments above.

I am going to get more familiar with the Chill Control unit and its ability to vary its settings and will report back on this shortly.

Email Joe

Moisture Control

As I indicated earlier, I did not use the thermal grease Asetek includes to insulate the CPU’s pins form moisture. I DO NOT recommend you ignore this step, I only did so to allow quick changes to the CPU.

As a “moisture torture test”, I ran the Vapochill for 12 hours at idle. The P4 basically powers down and is radiating very little power. At the end of 12 hours, the LCD read -29.5 C and the P4 on-die temp read -7 C, per MBM. I shut down and removed the cooling head to examine the CPU for any signs of moisture.

I found very little – two very small drops on the back of the CPU. The foam was not moist at all. Remember – this is without any thermal grease to insulate against moisture intrusion. Clearly, follow Asetek’s instructions to guard against moisture.

Temperature Gradients

I found about a 25 C difference between P4 on-die temps and Vapochill Cooling Head temps. There is an interesting thread about thermal grease HERE in Asetek’s Forum.

To throw a bit more light on this, I used a P4 1500 that I have modified to incorporate a thermocouple in its case top. This gives me temp readings about as close to the Cooling Head as possible (except for embedding one in the head itself). The difference between the two is the thermal resistance between the head and the CPU case top.

The sum total of this resistance is due to two items:

  • Thermal Grease
  • The fit between the Cooling Head and CPU Case Top

The best thermal grease in the world will not due much to correct uneven surfaces.

I ran the modified P4 at Idle and Stress with the following results:


LCD Temp

CPU Case Temp

MBM CPU Diode Temp


-28.5 C

-25.4 C

-10 C

STRESS (Prime 59)

-20.0 C

-12.2 C

2 C

Clearly the difference between the Cooling Head and CPU Case Top is less than between the Cooling Head and P4 On-die temp. There is an additional thermal interface (CPU chip to heatspreader) that accounts for some of this, as well as the location on the die of the temp diode.

Users interested in increasing performance might consider removing the P4 heatspreader, although this affects the clearance between the Cooling Head and CPU which must be compensated. If anyone has tried this, I’d like to hear about your experience.


Vapochill Fixes

As I tested the Vapochill PE, I found a couple of problems:

  • Two wires wire cut ( Details HERE)
  • Motherboard incompatibility (Details HERE)

Wiring Fix

Asetek is now putting protective plastic edging around the hole between the case top and the interior:

Hole Fix

Pic from Asetek.

Easy enough and a good fix.

Motherboard Compatibility

This problem was more troubling and not limited to Abit – a number of folks have encountered similar problems with other brands.

The problem I had was that after the Vapochill hit its “boot up” temp, the motherboard would not boot. I and others have done workarounds on this issue, but (IMHO) an expensive product should not burden users. Asetek spent the time and energy to modify the ChillControl software to fix this issue – you can modify its software.

Let me speak to how this upgrade works:

When I evaluate a product, I take the “Man from Mars” approach – nothing is obvious and I want specifics, not generalities, on something so critical as upgrading firmware. If you’ve ever updated BIOS, you probably remember the first time you did it – most likely with some trepidation and happy it worked.

The directions on upgrading the ChillControl firmware are not as detailed as I would like. It assumes you know that:

  • You have to modify the ChillControl Boot Disk
  • The ChillControl screen will not load system data when upgrading firmware

For versions up to 1.2, you need a second PC to change the firmware; versions 1.3 and later can be updated using the Vapochill alone. For those interested in updating up to version 1.2, the specific steps I took were as follows:

  1. Download the latest firmware update from Asetek
  2. Unzip – you will find a file named “Asetek.exe”
  3. Open the “ChillControl Configuration Utility” disk
  4. Rename the current “Asetek.exe” file to something like “AsetekOLD.exe” and copy the new “Asetek.exe” to the disk
  5. Power down the Vapochill and disconnect the Power Supply’s 120v plug
  6. Locate dip switch S1 on the ChillControl board
  7. Move dip switch S1 on the ChillControl board to ON


  8. Connect the Communications Cable’s RJ11 plug (looks like a big telephone plug) to the ChillControl board


  9. Connect the other end to the second PC’s serial port
  10. Connect the PS plug and start the Vapochill
  11. The Vapochill will not boot the motherboard – the front LEDs should show 4 steady lights
  12. Boot up the second PC with the ChillControl disk
  13. Select the appropriate COM port (1 or 2)
  14. The bottom screen will stop at “Com1 [or 2] Opened! – Waiting for ChillControl to connect”
  15. Now hit F8 – this starts the firmware upload
  16. The LED lights will flash sequentially and you will see a progress indicator in the bottom screen
  17. After 100% complete, the firmware has been updated
  18. Shut down the Vapochill and second PC
  19. Disconnect the Vapochill PS 120v power plug
  20. Disconnect the communications cable from the ChillControl board
  21. Move dip switch S1 on the ChillControl board to OFF

  22. Connect the Vapochill PS 120v power plug and boot the VapoChill

After these steps, the Vapochill should boot up with the firmware updated. Note that if you have ever updated the BIOS on a motherboard, the Vapochill update is different – there is no “bin” file; the update is in the Asetek.exe file.


I have to thank Asetek for being responsive to the issues I identified above – others who have had the motherboard boot problem are now running without workarounds.

So far, I have to say that the Vapochill clearly benefits from having been around for a while – the build quality is very good and the compressor is whisper quiet, a testament to quality components. Installation is fairly easy and the pieces are well thought out and do their jobs nicely. I found that removing the cooling head was no big deal, so that changing CPUs is not a daunting task.

The case is OK – my one complaint is that the floppy is located in a deep recess, but that’s me. The overall look is quite nice.

Cooling performance is quite good and will keep CPU temps well below ambient. How far depends on the CPU used, watts radiated, thermal grease, contact pressure, etc.

Moisture control is absolutely key – follow the directions and you should be OK.

The major question any consumer addresses: “Is it worth it?”

On straight economics, a case could be made that over time, users can buy something like a P4 1.8A, overclock it to 3+ GHz, and in the process enjoy superior performance on the cheap. Over enough CPUs, it could be argued that the Vapochill pays for itself.

However, I would argue that economics aside, the Vapochill is a thrill ride for the hardcore overclocker – and that ain’t all bad.

Email Joe

Be the first to comment

Leave a Reply