What can cause tiny bubbles 'foaming' at high flow rates?

[Twinkee]

Rig:
Distilled water system. Silver coil in reservoir. Cleaned first with SysClean-HPD mixes using a little CP-701 pump. Pump at bottom of case, 90' compression fitting to GPU full board block, CPU block, snake fittings to create gentle U bend into triple rad outside and on top of case, very restrictive rotary extension into bottom of 100mm XSPC reservoir at top of loop, in-line filter at outlet of res, in-line (not T-piece) temp probe shortly before pump. 1/2" ID tubing throughout with no kinks.
No problems with DDC-T+ but inadequate cooling to overclock. Switched pump to a D5 Vario to get higher flow and that's where the trouble started.

Symptoms:
Filled system one gulp at a time then sealed the res and turned pump on to start bleeding the system. FOAMmmmmm! Stop that!
Let it settle and turn speed to lowest... bleeds fine. Eventually there's little bubbles that don't reach res bottom where it's outlet is.
Turn up speed a little... bubbles start reaching pump.
Turn up speed a lot... foam fills the res. Stop that!
Let it settle and been using it at slowest speed with little bubbles that don't reach the res outlet.

Situation:
It's usable at stock speeds. The GPU runs a little hotter than the CPU which wobbles about 45'C.

Overclocking:
Rig ready its overclocking time. Using ASUS AI Suite with a 5:8 ratio, Q9550 : DDR3 1333 system is stable at 3.5GHz. Runs Office, VLC, interweb... solitaire :S but try to play a game and after a while BSOD. Windows?! What do you mean the CPU isn't working properly, the inline water temp display was at 38'C. GPU display was 48'C, CPU display was 45'C. Speedfan consistently shows that the GPU block sensor tends to be 8'c lower than the chip sensor and agrees with the CPU block sensor. So WTF?

Fired up Speedfan to look at the core temps and started an intel burn test... the CPU temp shot up. Stop that! I saw 94'C before the cores promptly returned to 38'C.

No more OCing until this is resolved 🙁 I feel sure I need higher flow to get the heat away, but need to stop the FOAMmmmmm! Help please.

 

[toolmaker_03]

air is going through the pump creating the foam i have seen this a lot before i started using a res. are you fulling the res. with water after bleading the system of air?

 

[4Ryan6]

Also pure steam distilled water is getting harder and harder to find, you have to read the label regarding the process that was used and be positive it is steam distilled with zero additives, sometimes the additives in the water will cause a reaction with the Silver Kill Coil.

This can also be caused by a partially crimped tubing somewhere in the loop, or contaminants in the radiator from the solder process, you did flush the radiator before putting it to use right?

 

[Twinkee]

air is going through the pump creating the foam i have seen this a lot before i started using a res. are you fulling the res. with water after bleeding the system of air?

Following your advice I have filled the system as best I could. There's now a tic-tac sized bubble instead of an inch of air. This is the smallest I could get it after several tries.

It doesn't foam anymore, excellent! Thank you. However it is now, what I think is called, "vortexing".

At anything above the lowest speed there is a visible whirlpool of tiny bubbles which reaches down out of the reservoir towards the pump. The inline filter may well help a bit, but as I don't use clear tubing I can't really tell how bad these bubbles are outside the res.

When asked about these vortex micro-bubbles moderator manofchalk had this to say

I think your under the impression that bubbles are overly dangerous to the pump.
A pump can handle a bubble going through it just fine, a pump will only burn itself out when there is only air inside it. Which is why its always recommended to have the reservoir before the pump in the flow order and physically above it, so the pump isn't allowed to run dry during the filling process.

I'd appreciate confirmation that the little vortex of 'micro-bubbles' can be ignored. I'd rather not consign myself to the hell of "you didn't double check did you?"

 

[Twinkee]

Also pure steam distilled water is getting harder and harder to find, you have to read the label regarding the process that was used and be positive it is steam distilled with zero additives, sometimes the additives in the water will cause a reaction with the Silver Kill Coil.

Product "Distilled Water" closer inspection "Ingredients: aqua (Deionised)" How do these people get away with this stuff? Ah yes the "you didn't double check" fairy has visited me again.

This can also be caused by a partially crimped tubing somewhere in the loop, or contaminants in the radiator from the solder process, you did flush the radiator before putting it to use right?

No kinks, but I'm gonna replace that restrictive rotary fitting with a short graphics card fitting. same result but with a 1.5mm ish bigger bore.

SysClean'd everything except the D5 which hadn't arrived when I started cleaning. Once at high concentration once at low. Nice and clean.

Thanks for reply. Thoughts on vortex micro-bubbles are also welcome

 

[toolmaker_03]

sorry for the late replay, i had to sleep, the vortex issue is somthing that i have been fighting for years, the only answer that i have found was to get a larger res. i now use the tube style ras., and anything smaller then the 250mm allowes the vortex to form and reach the bottom of the res.
as far as micro bubbles going through the pump.
1) is the pump quite?
2)would you say that 90% of the liquid in the res is water?
or is there more air in there than that?

 

[Twinkee]

as far as micro bubbles going through the pump.
1) is the pump quite?
2)would you say that 90% of the liquid in the res is water?
or is there more air in there than that?

I was answering... but now that I've braved turning the speed up... the tic-tac has disintegrated and the pump is running clack-free at full speed.

SOLVED : Completely filled the reservoir after bleeding.
Thanks

 

[toolmaker_03]

glad to read that.