# Cooling Fan Testing and Comparisons

Discussion in 'Hardware' started by LennyRhys, 22 Apr 2020.

1. ### Dr. CoinActive Member

Joined:
13 Sep 2013
Posts:
995
243
With 2mm of water you might want to consider a inclined u-tube manometer.

Edit:

Last edited: 22 Apr 2020
cobalt6700 and LennyRhys like this.
2. ### LennyRhysOink!

Joined:
16 May 2011
Posts:
5,976
468
Ah very cool - yes I did encounter some inclined manometers when scouring ebay and didn't give it much thought. Turns out the is the perfect solution to measuring fans with low static pressure, and I'll see what kind of results I get!

3. ### adidanGuesswork is still work

Joined:
25 Mar 2009
Posts:
16,030
2,790
That's usually the first thing said by 9 out of 10 people who end up on Tattoo Fixers.

The_Crapman likes this.
4. ### The_CrapmanDon't phone it's just for fun.

Joined:
5 Dec 2011
Posts:
5,317
1,872
There's a lot of canines in magaluf

5. ### cobalt6700Active Member

Joined:
17 Oct 2005
Posts:
619
102
Pitot tube theory is pretty simple TBH - you measure the static and velocity air pressure at the same time, subtract static from velocity to find your head pressure, which can then be used to find the velocity in m/s: V m/s @ 20 DegC = 4.05√(Velocity p - Static p) mmH2O. A set of velocity readings which are averaged, then multiplied by the area in m will give you a flow rate in m3/s.

Just measuring in one place will almost certainly give you a duff reading, and you need at least a length 10 Diameters straight/free flowing upstream (can be reduced with flow straightener) and 5 Diameters downstream to get good readings. Measuring small flows in mmH2O can also be tricky, a 120mm G1238B running flat out is around 8.05m/s, 3.95mmH2O Δp, 0.116m3/s. At 1/2 the flow, you would be looking at 4.05m/s, 1mmH2O Δp, 0.0583m3/s. Pa would be a better unit, but as Dr.Coin said, you will need an incline manometer to measure in smaller units. I may have a couple of old ones hanging around if you wanted one

That's fair enough.

Cool - basically your water has a specific gravity of 1 (or very close to that at 20 degC its 0.9982, at 4 degC its 1) up to you if you want to chill it to 4deg. I'd stick with room temp for easy to ref results. Unless you are either very high or very low in relation to sea level, atmospheric pressure will have little to no effect on your readings. If you were measuring in Pa with calibrated manometer, sure.

I mean.... now that is what a call an axial fan! Wowzer that thing goes. I fear the googly eyes will not last long on that one.

LennyRhys and Bloody_Pete like this.
6. ### Mr_MistoffeleesThe Cat Lies Down on Broadway

Joined:
26 Aug 2014
Posts:
2,779
885
So, would suit my dog when he gets over-excited.

The_Crapman likes this.
7. ### lirathealSharing is Caring

Joined:
20 Nov 2005
Posts:
11,814
1,326
The massive child in me wants to put a piece of paper in it.

LennyRhys and Dr. Coin like this.
8. ### GoateeWell-Known Member

Joined:
19 Apr 2015
Posts:
1,669
599
I was thinking drop a grape in it, maybe a cheese string too would be cool.

LennyRhys, liratheal and Dr. Coin like this.
9. ### LennyRhysOink!

Joined:
16 May 2011
Posts:
5,976
468
That's very kind of you - I'll see how I get on with the DIY jobbie at the moment and may consider upgrading to a more serious measurement later!

Speaking of... I had a bash with inclining the u-gauge and the results are definitely a lot clearer as the water column moves a greater distance, however there are some issues with surface tension inside the tube and the water being a bit unpredictable. Tapping the tube a few times usually does the trick. but I've noticed that the water can sometimes "hug" the inside of the tube and simply go from being convex to concave on the intake side.

Here's the setup that I tested this evening, using three fans: Gentle Typhoon (2,250rpm), Noctua NH-A12x25 (2,070rpm), and Delta AFB1212M (2,060rpm). All the speed measurements are made with a laser tachometer.

Obviously these measurements are no longer mmH2O, but the comparative results are apples-to-apples and are very interesting indeed. The slowest fan of the bunch has (by far and above) the highest static pressure, but it's also the loudest.

Noctua NH-A12x25: 2mm
Gentle Typhoon: 2.6mm
Delta AFB: 4.3mm

I re-did each test and measured the tube at both points where the water level is, and I got the same results (unsurprisingly). I marked the measurements with tape and then measured the gap with my vernier calipers.

You can see from the picture below that the water is anything but level inside the tube, and I'm presuming this is surface tension / capillary behaviour? Water was at room temp, as suggested by cobalt6700!

Dr. Coin likes this.
10. ### LennyRhysOink!

Joined:
16 May 2011
Posts:
5,976
468
I have three 10,000rpm 92mm fans and I may be willing to set one aside for such frivolity in the name of science. I'll add paper, grapes, and cheese strings to the list.

cobalt6700 and Dr. Coin like this.

Joined:
3 Mar 2017
Posts:
3,412
1,507
and a carrot

12. ### GoateeWell-Known Member

Joined:
19 Apr 2015
Posts:
1,669
599
and a Peperami

The_Crapman and RedFlames like this.
13. ### Mr_MistoffeleesThe Cat Lies Down on Broadway

Joined:
26 Aug 2014
Posts:
2,779
885

The_Crapman and cobalt6700 like this.
14. ### Mr_MistoffeleesThe Cat Lies Down on Broadway

Joined:
26 Aug 2014
Posts:
2,779
885
Perhaps you could try a drip of washing-up liquid to reduce surface tension?

15. ### GoateeWell-Known Member

Joined:
19 Apr 2015
Posts:
1,669
599

Other than feeding your fans with stuff from the fridge, are you taking user submissions to add to your test roster?

I have a PC fan in mind (I actually have 6 in mind but I guess one would do) but they make your pride and joy look a bit weedy in the RPM department, so I wouldn't want you to have to upstage it.

Model PF40561BX-000U-S99
Size / Dimension Square - 40mm L x 40mm H
Width 56.00mm
Air Flow 31.7 CFM (0.888m³/min)
Static Pressure 3.600 in H2O (896.8 Pa)
Bearing Type Ball
Fan Type Tubeaxial
Features Auto Restart, PWM Control, Speed Sensor (Tach)
Noise 67.9 dB(A)
Power (Watts) 16.68W
RPM 18,000 RPM (the front fan is 21,000)

16. ### RedFlames...is not a Belgian football team

Joined:
23 Apr 2009
Posts:
13,089
1,978
If you're on Windows 10 [and not on an ancient version of same] - WinKey + . [or Winkey + ;], gives you an emoji/special character picker... not as quick as Alt codes, but better than nothing [also you're not haveing to remember the codes]

liratheal likes this.
17. ### Dr. CoinActive Member

Joined:
13 Sep 2013
Posts:
995
243
I didn't see the period at first and tied Win +
I don't see symbols just emojies when I use period or semicolon. My fall back for key codes I don't remember is to bring up the Character Map program (sorry a bit of ancient dating to NT 4.0). Tip, click the Advance View box, and the very bottom use the search box to find the symbol of your choice.

18. ### RedFlames...is not a Belgian football team

Joined:
23 Apr 2009
Posts:
13,089
1,978
This is what I see, Emoji, Kaomoji [stuff like '¯\_(ツ)_/¯' or '(╯°□°）╯︵ ┻━┻'] and then Symbols across the top, and categories across the bottom [unfortunately the search function doesn't really work that well with the symbols].

19. ### LennyRhysOink!

Joined:
16 May 2011
Posts:
5,976
468
Could do - I'd be concerned about the possibility of making unwanted bubbles at the surface though. No harm in trying I guess!

Absolutely - I was thinking of making it as comprehensive as possible and if there are fans that people want to suggest (or send to me) I'm happy to feature them. As a general rule of thumb, smaller fans can have higher rotational speeds because of lower air impedance, but their increased static pressure comes at the expense of vastly reduced airflow.

I've got a "fun" fan on the way from China which has already been in transit for a couple of weeks... fingers crossed it'll get here soon. It'll be the jewel in my crown of 12v fans.

Joined:
17 Oct 2005
Posts:
619