Eight months ago, on 21st October 2008, I converted my quad core computer to water cooling. I purchase a new Zalman ‘Reserator TX hybrid liquid cooling system’ I wrote and article on converting from air cooled to water cooled CPU.
Click here for the original article.
The following charts the progress and problems associated with the project .
The installation was very simple with the easy to follow instructions Zalman provided.
The Zalman kit was supplied with quarter litre of concentrated Anti Corrosion Coolant ZM-G300 which must be diluted with distilled water. The main problem was associated before installation: locating distilled water!
In the UK, it is not easy to obtain pure distilled water. Go into any motor shop and ask for distilled water and you will be given de-ionised water, not distilled. The distilled water available in supermarkets for use in domestic irons contains perfumes and other contaminants.
A pharmacy sold me a bottle of ‘Pure Water’ assuring me that it was distilled. I telephoned the manufacturers and spoke with their quality control pharmacist. He confirmed it was de-ionised, not distilled. However, discussing my needs etc, he assured me that it would be absolutely safe to use de-ionised water and no harm would become of my system.
So, I diluted the deep blue ZM-G300 according to instructions. I filled and commissioned the Reserator TX. All seemed fine. It worked like it said on the box. I was very pleased and wrote up the article on the installation etc.
However... During March 2008, the alarm would go off occasionally. There are three items that can trigger an alarm: Low coolant level, no coolant flow and over temperature. The alarm does not advise which triggered it.
Investigating the alarms, I determined at the time, that it was low coolant level that triggered the alarm.
During the interim period, I had located a seller of real distilled water on Internet and purchased 5 litres and another bottle of Zalman anti-corrosion coolant additive.
So, two months ago, I emptied the Reserator, removed the water block and pipe work and reverse flushed it out. During this process, I noticed that there were some small particles of blue ‘bits’ about the size of sugar crystals but didn’t take much notice or concern because of the cleaning.
Reinstalled the water block and filling with zm-G300/distilled water mix and all was well.
Yesterday, 15th May, the alarm went off permanently. I noticed that there was no liquid flow, indicating pump failure or blockage.
I disconnected the pipes from the CPU water block. I then connected the Reserator outlet & inlet pipes together. Powering the Reserator via an external 12v supply, full coolant flow was observed. Therefore, the blockage was most likely to be in the CPU water block.
I carefully removed the water block from the CPU. Cleaned off the thermal compound and stuck insulating tape underneath so as to protect the heat plane from scratching etc. I peeled off the warning notice: ‘Removing this sticker will void the warrantee’. Using the correct size allen key, I took it apart.
Caution: do not dismantle your water block unless you are prepared to write it off and have to buy another or be prepared to invalidate warrantee!
Extreme care must be taken when dismantling and handling the parts. There is a water seal ring that must not be touched. Likewise the sealing surface on the mating part must not get scratched.
What I discovered was astounding. The flow chamber matrix was blocked with a hard deposit or growth of some sort of crystalline material.
The blockage was in the centre of the matrix, coinciding with the inlet pipe. It was mechanically removed with a small probe tool and then immersed in a jar of vinegar (acetic acid) to help dissolve any alkaline deposits such as calcium. It was then scrubbed thoroughly with an old tooth brush, rinsed and put to one side to dry.
The top section of the water block had massive growths of the same crystalline material in and around the inlet and outlet pipes. The majority being on the inlet pipe aperture.
There also appeared to be some sort of blistering around the inlet/outlet aperture. I discovered that the internal surfaces of the water block are coated with a very thin layer of a plastic material. This had lifted and allowed coolant to ingress under it.
Side view showing the growth of crystalline material.
Close up
Removing the deposit/growth required a sharp knife to ‘slice’ away the stuff. It took some 30 minutes and another session with vinegar to remove nine tenths of it.
This is just a small amount of the dried out material removed from the top chamber of the water block. Crushing a small amount was very similar to crushing sugar crystals.
This shows the etching of the metal where the crystals were growing. The outline of the displaced plastic coating can be seen. (The oops is a scratch caused by careless handling of the removal knife!)
Eventually it seemed clean enough to re assemble. I offered a prayer that It would not leak all over a £600 CPU!
On reinstalling, everything appears fine – full flow is restored! (& my CPU remains dry!)
Conclusion.
I am now convinced that the earlier alarms were caused by restricted coolant flow. I am not sure of the cause of this problem. Two possibilities come to mind:
1) Using de-ionised water may be the culprit. Although I have been assured that it should not cause corrosion by an industrial chemist who specialised in de-ionised water. Because of my experience and unless it is proven otherwise, I do not recommend the use of de-ionised water.
2) I have a faulty water block that allowed coolant to make contact with the internal metal. I don’t favour this option.
Unanswered questions:
*The Anti-corrosion additive did not do its job! Why?
* What is the crystaline material - a precipitate? A chemical reaction with anti-corrosion liquid & de-ionised water?
I shall be forwarding these findings and photographs to Zalman and will report back their comments.
[all photographs copyright DonaldG