Fan PQ curve testing - Part 1

[Eternalightwith]

I got one of those newer Scythe G series. 1900rmp. I'll sent it to you if you want to throw it in the mix

[Martinm210]

I got one of those newer Scythe G series. 1900rmp. I'll sent it to you if you want to throw it in the mix

Thanks!

I'd like to get caught up with what I've got on hand first, then I'd be interested in adding some more. The S-Flex series is one I'm interested in trying though, so I'll PM you later and check if you've still got one you're willing to give up.

[Martinm210]

Here is the San Ace running at 5V. I've heard many people say they like running these at very low voltages. They do undervolt well and due to the 38mm thickness they do have very good performance RPM vs RPM of a 25mm fan.

However, I'm not convinced of the motor noise. I think it is better than most fans in it's class (High speed fans), but I'm not sure it's as quiet as the fans designed to run in the Silent category.



Note that this was tested at 5.0V to keep it in the under 1500 RPM category.



It's impressive performance for the RPMs turning pretty well leading the pack and showing how a thicker fan blade is really helpful for pressure.

I'll have to leave it to round 2 for the noise check though. Subjectively, I don't think it's as quiet or smooth at 5V as the other fans. I would even suspect the Noise Blocker XL-2 is a better sound which is producing more flow than the San Ace at 5V.

The fan is however an awesome candidate for someone that need a large range of fan power (ie quiet for daily operations, but noise is ok and needs extreme fan performance for benchmarking/gaming) That's where I see this fan fitting in best.

[skinnee]

You're a machine! Excellent work!

[Martinm210]

You're a machine! Excellent work!

I got a little excited about trying something new, so I did spend a little time on this today.

Still ALOT less work than the water tests which is nice. And no running to the bathroom to dump a water loop.. I can run one in about an hour which is better than the 10-15hours per radiator you are all testing or the 5-10 hours per waterblock...

I'm just having fun...it's probably all wrong, but I'm learning and enjoying something a little different, so it's all good!

I'll try to work in another fan test in the evenings now and then for a little bit before it's time to go back to the closet again..

[skinnee]

Haha, I implemented a drain system in the custom res after the little aquarium heater incident. A ball valve, some tube and a gallon jug...perfect! The time thing...I still can't estimate start to completion properly, always something that pops up and takes longer than I expected. Like this radiator round 2, I am seeing results different than I expected. So, I have to test one or two of the first round radiators again to verify I am still consistent.

As for the fan testing you're doing, its great. I have always wondered, especially with radiator tests and setting RPMs what the actualy CFM and pressure curves were. Exactly how much air is being moved at 1000RPM on a D12SL12 versus a D12SH12 at 1000RPM? Excellent work my friend, excellent work!

[Power_VANO]

Excellent work Martin! Your tests are perfect quality, as always

[Brodholm]

nice! I really would like to see some S-flex in the mix. A 1200rpm version to compete against the yate SL in noise and cfm!

[hotdun]

Thanks Martin!

[-DC-]

NB rev.3 silent? hm, try 6 these fans on a radiator (in\out) and you won't tell they're silent) Magmas make much lower noise, but i can't argue... airflow of NB is nice

[Martinm210]

Thanks guys.

I'm currently running 8 YL D12SM12s on by dual quad rad setup, but they seem to be fine down under 1000RPM to my ears. We'll see after I'm done with all this, I may have to make a change..

And...I got one more done this evening....the Scythe Ultra Kaze 1000. I really had high hopes for this fans with it's unique RPM range for a 38mm fan. I bought this one myself a while back thinking it might just be the ticket for more pressure while retaining a silent RPM range.

Unfortunately it didn't turn out that way. This is my weakest fan tested so far, the 25mm 1000RPM noise blocker outperforms it and the NB is silent where this one has an obvious motor hum to it.

There are two things I'm noticing about the 38mm fans. Their larger hub/motor seems to be a cause of more motor noise than the 25mm types. And unlike the high performance fins of the San Ace....the Ultra Kaze fins and 38mm thickness really didn't seem to provide any noticeable benefit over 25mm across the curve where it matters. It did ok with static pressure at 0 flow rates, but it just didn't do very well across the curve.

The other thing I see similar to the San Ace 38mm test, these 38mm fans seem to have a much more prominent flat spot or stalling area.

Anyhow, I'm just not impressed with this one. The UK 3000 may be a very strong fan, but this UK1 just didn't perform very well for me..

The photo:


The Curve:


I'll update the OP with the summary/comparison chart in a bit..

[Hondacity]

myth busted with the 38mm fans. with the heavier hub and thicker fins...do you think its dragging that much?

[Eddie3dfx]

Hmm 38mm panaflos?
Anyhow, thanks for testing this stuff Martin. I find these tests fascinating..

[Martinm210]

myth busted with the 38mm fans. with the heavier hub and thicker fins...do you think its dragging that much?

Yeah, the Ultra Kaze 1000 suprised me too. I expected good things from it. The fins are actually pretty large so I was thinking it may actually do better than the San Ace per RPM, but it wasn't even close. I'm guessing the Ultra Kaze was really designed for the high speed 3000RPM motor setup and the 1000 was just a minor tweak of some sort to get the RPM down for one more product option

The San Ace is still showing the stongest performance per RPM, but even with that, once you get down to 1000RPM, the difference is nill. I'm just not a fan of the 38mm fan noise so far.

Hmm 38mm panaflos?
Anyhow, thanks for testing this stuff Martin. I find these tests fascinating..

I do have a couple of medium speed panaflos, I'll be putting one up undervolted to some sort of level.

Glad you like it, I'm having fun with this. The suprises are always a nice reward...good or bad

[xpatar]

Martin will there be a test with SCYTHE DFS123812L-2000 ULTRA KAZE 120MM FAN? Do you believe that there will be a good balance between static pressure, airflow and noise?

[gillbot]

what about one of the 1600 rpm san aces?

Awesome martin, thanks!!

[Eddie3dfx]

I think the sony s-flex vs the zalman zf3 is in order
Those need db numbers though

Those san-aces are very smooth. No wonder they are so costly.

[xpatar]

And this is what that very bottom test looked like running it against the heatercore. The heatercore is very restrictive for airflow, so most rads will probably have noticeably less pressure drop and result in more air flow than this, but it give you an idea.

I'm measuring the pressure difference on each side of the fan with the gauge and the air flow on the inlet side.

It's alot like a pump test, measuring the pressure difference relative to flow rate. It seems my last test is showing more of a prominent step in the curve. That's the sort of thing I'm interested in seeing and should show up alot better with the stronger fans.

Martin, what would you think about the order fan-chamber-radiator-chamber in order for the pressure guage to measure values just before and after the radiator? Do you believe that we could have better results about pressure drop right after the radiator?

[Martinm210]

Martin will there be a test with SCYTHE DFS123812L-2000 ULTRA KAZE 120MM FAN? Do you believe that there will be a good balance between static pressure, airflow and noise?

I have some 3000RPM Ultra Kazes that I did my previous rad testing with. There is no doubt that they are very strong fans at high RPM. I'll at least being doing some testing with the 3000 which should give you a sense. Not sure about noise quality though. They are a good prices, but I wouldn't expect them to be as silent as some of the others. Subjectively they don't seem to have near the smooth motor quality of the San Ace, but I can't quantify that yet.

what about one of the 1600 rpm san aces?

Awesome martin, thanks!!

I've only got the one San Ace, but seeing that it's a strong performer, I'll include some more tests at higher than 5V which should give you some ideas.

I think the sony s-flex vs the zalman zf3 is in order
Those need db numbers though

Those san-aces are very smooth. No wonder they are so costly.

I've got some s-flexs in the works..

Martin, what would you think about the order fan-chamber-radiator-chamber in order for the pressure guage to measure values just before and after the radiator? Do you believe that we could have better results about pressure drop right after the radiator?

That is what I will want for radiator pressure drop, but testing the fans you want the pressure before and after the pump to measure the pressure rise so to speak.

So how does all these numbers and charts equate to keeping my LC loop cooler and by how much? Same thing with pump/top PQ curves (especially with PQ curves) and rad testing.

And I mean this with no disrespect to all the hard work, just trying to figure it all out practically speaking.

andyc

It's a good question, and I should try to explain. What I'm doing here is like measuring the PQ curve for a pump. By itself it doesn't tell you much, but combined with pressure drop curves for a system, you can determine the net flow rate.

Unfortunately I only have one radiator tested, the bonnie heatercore. With that pressure drop and with these curves, I can estimate the actual flow rate for any of the fans. Furthermore, if I have pressure drop curves for more than one radiator, I could estimate the flow rate for any radiator.

The cool thing about it is that pressure drop of a radiator is fairly linear and easy to measure.

Even better is how significant air flow rate is to the radiator performance. Unlike the rather minimal if any gain you get from water flow rate, air flow rate is HUGE! Remember that you can see anywhere from 3 to 5X!!! the performance difference on the same radiator depending on the fans you use. Obviously someone that likes 1000RPM fans isn't going to switch to 3000RPM fans, but they might go from 800 to 1200 if they see there is a big gain for them.

Regarding the function of the data, lets look at this first chart:

The straight red line represents the bonnie heatercore pressure drop, and the curves of the fans are the curves.

With this chart I can estimate the air flow rate of any of the fans. Just follow the pressure drop line up until it intersects the fan curve and drop down to the resulting flow rate. The important part of this is I only had to test the radiator one time, the fans could be tested completely independently to produce that first chart.

Scythe UK1=14CFM
Noise Blocker XL1=17CFM
YL D12SL12 S or San Ace @5V = 23CFM
YL D12SL12 C = 25CFM
NB XL2 = 27CFM

So you get double the flow rate using the NB XL2 vs the Scythe UK1. Double the flow rate will nearly double your performance (not quite, but a higher portion of that flow rate gain.

As a guess a system that had a 10C water delta using the Scythe UK1 would probably see something like a 5-6C delta changing over to the NB XL2 fans.


So...what I'm doing here may see a bit disconnected and hard to follow, but that's because it's only a few pieces of the puzzle (I need more radiator pressure drop curves). You can do more real world type of testing on a radiator (like my last test I'm doing for each fan), but that leaves you with a set of isolated results you can only use for that one comparison. It gives you an idea for that style of radiator, but you may get a different picture using a different radiator because the fan curves are not linear on the same slope. They have their own bumps and particular performance curves that don't allow you to estimate well without the full curve.

Not sure that makes sense, but what I'm really trying to do is see the complete performance picture of the fan. Performance of a fan is completely represented by the full PQ curve.

In a nutshell, it's a way of looking at the fan's performance in more detail and to understand the odd things a little better AND have the ability to create tools to estimate from.

If all radiators had the same air pressure drop, there would be no need to do this, you could simply test on a radiator and record the resulting air flow. But we have many types out there. So if you tested with only one radiator, it's possible you could find a different answer with a different radiator.

However, if you test and measure the entire PQ curve. You then only need to test the radiator one time to determine the pressure drop curve. The real beauty of it all is that you can then determine the performance for any radiator that you have a pressure drop curve. Can you imagine the number of fan/radiator combinations there are?

If there are 10 different radiator fin types out there and 50 fans, you'd have 500 sets of test doing it the real world way. But doing it with PQ and pressure drop, you reduce that amount of testing to 60 sets of tests.

Does that help?

I'll play around with looking at a means to create some sort of estimator out of the data collected, but for now I'm just enjoying a little testing. I've got my inspector's spectacles on and I'm trying to find the oddities in some of the fans that I have here. I'm seeing a few interesting things so far. I really got myself burned out on the thermal testing and have always enjoyed going after the areas that havn't recieved much attention...and air flow rates are something I know are important to performance unlike the water flow rates...especially for folks the either want to run smaller radiators or very low speed fans.

[alacheesu]

Thanks for your hard work. Very interesting.

[xpatar]

Martin with all the respect to your personality, what would you say after finishing all this great work, to start a series of tests that will show the real impact of static pressure on water temps especially with water coming out of the rad? It would be good to have all these results consolidated. Just a thought.

[Sadasius]

Martin with all the respect to your personality, what would you say after finishing all this great work, to start a series of tests that will show the real impact of static pressure on water temps especially with water coming out of the rad? It would be good to have all these results consolidated. Just a thought.

There is no solace for buying a fan that is not to be named here. BTW the rads help cool the entire loop, not just water coming out the rad (equilibrium). Which means water going in the rad will be cooler too on the rig with the better fans. So the important temps are all of them, not just out of the rad.

[gillbot]

I've only got the one San Ace, but seeing that it's a strong performer, I'll include some more tests at higher than 5V which should give you some ideas.

though you don't need to unless you want to. I appreciate the hard work!

[xpatar]

There is no solace for buying a fan that is not to be named here. BTW the rads help cool the entire loop, not just water coming out the rad (equilibrium). Which means water going in the rad will be cooler too on the rig with the better fans. So the important temps are all of them, not just out of the rad.

What I mean is what are the DT values for the temps right before and after the rads. Of course rads help cool the entrire loop, but wouldn't be better if we knew for every fan that is tested what are the gains in water temps because finally that's what we want is lower temps in our system. So what are the gains of using fans with higher static pressure in water temps? Generally speaking better fans wil give bettter temps is simply general. Why don't we record those numbers? How much do they really drop water temps?

[Sadasius]

I would like to see what worked best with whatever used on a test setup too in a very nice graph.