Fan Testing Round 5 (Working Thread)Are "NANO" Bearings the new sliced bread?

[Church]

Offtopic
Waterlogged: where do you buy C-strip to hide cuts? How thick is it?

[wez]

Offtopic
Waterlogged: where do you buy C-strip to hide cuts? How thick is it?

For high quality u-strip, try a motorcycle shop

[Martinm210]

Wow, Martin you are the man! I dont know how you have the time to do all this, but it's awesome to see it happen.

Do you have a YT channel? I'd just as well subscribe

Thanks!

Like any of my past testing obsessions, I just like learning and tinkering. I was never satisfied with my past test methods, so when I bought the new camera..it fired up an interest. Now with the increased ability with the H1 recorder and noise analysis software, I'm just having a ton of fun trying to figure things out.

The new software weighting seems to finally do noise some justice. I can also log dBA results now, so if I add in anemometer logging, I should be able to plot out some really accurate dBA vs CFM charts with little effort.

Almost there...having fun too. I do have a YT channel too..

[Waterlogged]

Offtopic
Waterlogged: where do you buy C-strip to hide cuts? How thick is it?

This is the stuff I get.

http://www.frozencpu.com/products/30...38&id=uJtz27Xz

They also have thinner stuff but I've never used it.

http://www.frozencpu.com/products/50...38&id=uJtz27Xz

[Martinm210]

FYI,

In an effort to improve my anemometer, I decided I was going to make a sacrifice of a high end fan to steal the bearings out of it.

I decided to chose the Gelid, since it appeared the bearing OD would be about right. There is basically just some sort of hard plastic sleeve bearing with one steel ball at the bottom. It doesn't have your typical ball bearing setup.

I kind of like this sort of sleeve bearing setup, but it surprised me. I guess the plastic sleeve is some sort of special plastic, because it is pressed into the frame. Anyhow, just thought I'd share. Now I have to go hunting for another bearing victim..

[Church]

Hmm, changing bearing type will change friction. It won't impact airflow readings because of that?

[Vinas]

Alright, I ran a mockup scope test video of just the GT..Ill post a video after it gets done rendering and uploading. Seems like this could be a usefull tool in identifying problem areas, although I suspect its more rpm related than voltage...I don't have my new kaze master just yet. Just want to see if this is of any use.

The video does at least confirm the odd harmonic is not coming from my radiator. I tested this open air just so I can eliminate those variables.

Martin, those audio scope tests are definitely capturing some good data.

I hooked up a couple of fans to my DSO just to see if I could get a reading depending on a couple factors. 1) voltage, 2) ground vs common, and 3) trying two different fans with the exact same test method.

Ultimately the electrical sine wave is changing when I put different fans and voltages to them... My general hypothesis is that the wave is directly proportional to the "fluidity" of the fan pulse from its windings, and possibly could affect rotor and bearing noise on the fan its self.

The first pic is a Scythe SY1225SL 12SH 120mm fan.

The second pic is a Thermaltake TT-9025A 92mm led fan.

The third pic is the Scythe SY1225SL 12SH with channel 1 amplified @12vdc.

[Brodholm]

Sorry for the late reply.

Yep, actually have to drill the holes out a little. There is a bit of a gap but C strip fills most (if not all) of it rather nicely and makes the hole look a bit more "finished".

Here's a couple of pics of how it looks on my current build.

[IMG]snip[/IMG]

[IMG]snip[/IMG]

It looks a little rough here because I cut the hole as large as I could for max air flow, it actually looks a bit better on a regular fan hole.

For screws, I usually use 6-32 or 10-32 machine screws depending on whether I'm mounting a rad (6-32) or fan(10-32). The only exception to that rule is the GT, it has a molded area between the holes in the open area for a small 6-32 nut to be inserted and a little hot glue holds it nicely in place.

No worries

Where did you get the "antivibb" stuff? (the blue rubber gromets).

[Martinm210]

Hmm, changing bearing type will change friction. It won't impact airflow readings because of that?

Oh yeah, starting over from scratch as usual. Actually I bought a used broken Omega HH-30 anemometer (A $500+ meter when working). I'm going to try and fix it first, but if I can't, I'm going to steal the vane probe out of it and morph the two together.

I had my vane probe and bearings out a dozen times yesterday, trying different lubricants and repressing the bearings in place. There was a fine line between low viscosity lubricants causing bearing noise vs greases that were too thick. All of them affected the calibration afterwards too.

I had to though, that anemometer vane rattling was driving me nuts. I've eliminated the wobble and noise is much better with the factory vane. Regardless, the vane probe on this anemometer is not the finest quality. It does have two ball bearings, but they are sloppy at best I'm not very happy with it.

It's all good though, anemometer vanes are not a very consistent instrument. Simple things like warm up time are critical. I generally don't trust any result that wasn't tested back to back with a good amount of warm up.

I'm also calibrating my microphone down, so any future tests will be stand alone for that reason as well.

Martin, those audio scope tests are definitely capturing some good data.

I hooked up a couple of fans to my DSO just to see if I could get a reading depending on a couple factors. 1) voltage, 2) ground vs common, and 3) trying two different fans with the exact same test method.

Ultimately the electrical sine wave is changing when I put different fans and voltages to them... My general hypothesis is that the wave is directly proportional to the "fluidity" of the fan pulse from its windings, and possibly could affect rotor and bearing noise on the fan its self.

The first pic is a Scythe SY1225SL 12SH 120mm fan.

The second pic is a Thermaltake TT-9025A 92mm led fan.

The third pic is the Scythe SY1225SL 12SH with channel 1 amplified @12vdc.

Interesting, and nice tools and setup you got there. There is definitely some spikes that can get missed with large voltage increments. I've been experimenting with some logging type methods.

I just need to figure out a way to very slowly and consistently increase voltage. Using my watch to slowly turn the voltage dial at a constant rate is proving a bit tricky..

I'll post a couple of charts to show you what I'm working on. New methods again as usual.

[Martinm210]

This is what I've been playing with, trying to find a way to take a more detailed CFM vs dBA look. The combination of the Zoom H1 plus software weighted dBA is feeling much more accurate to my ear, so I think having a good logged type method would be good. The only problem I see with this currently is the anemometer lag. I really need some way of increasing voltage very slowly and consistently...

Anything to take the manual interpretation and manual logging out of the method would be good too.

[Manicdan]

that last chart looks like you need to sort things by dBA, instead of what i think is time? it might get rid of having multiple Y values for a single X value

[Waterlogged]

No worries

Where did you get the "antivibb" stuff? (the blue rubber gromets).

I get mine from Mcmaster directly but you can find them in the "Silence", "Noise Reduction", "Quiet PC" or similarly named sections at E-tailer's that have such sections, look for "HDD silencing" kits. They come in different colors too from what I've seen.

If you want to try and get them from Mcmaster, you can use a company called MyUS.com.

For Radiators and direct mounted GT's I use # 9311K138

For direct mounted fans other than GT's (that get 10-32 screws) I use # 9311K139

[Martinm210]

that last chart looks like you need to sort things by dBA, instead of what i think is time? it might get rid of having multiple Y values for a single X value

Time is just the common tie of the actual test. I think I am getting some bumps in dBA from not adjusting voltage smoothly enough though. I suppose there could be bumps from harmonics and or just fan characteristics as well. I would have to retest a couple of times to sort out if it's just testing error or if there is actually something going on there.

I know most fans have a flat spot in their PQ curves, so I would think it's possible to find that same sort of flat spot in a voltage sweep like this. It should theoretically go vertical when the pressures go to that point.

My biggest problem with this method is the need for a way to very very slowly increase voltage. Something like 3-13V in a linear progression over a 5 minute period. Maybe something like gearing up a clock motor to the fan controller knob..

I'm not sure, but it seems like any sort of logging method is going to require a loooong slow test to do any good. Maybe even change the dbA averaging to 5 seconds or longer. This quick test, I ran at 2 second smoothing.

[Manicdan]

just do a few quick tests, one where you adjust it across 10 seconds, 30 seconds and 3-5 minutes (good luck moving that slow by hand, lol). and see how different the results come out. if they look near identical, thats probably going to be more than good enough

theres potentiometers that are in a box with a flathead screw on the side and can turn alot, i think built for very high range of values, maybe you can get something like that, and just hook it up to a motor that runs 1-5 RPM?

[MagisD]

Time is just the common tie of the actual test. I think I am getting some bumps in dBA from not adjusting voltage smoothly enough though. I suppose there could be bumps from harmonics and or just fan characteristics as well. I would have to retest a couple of times to sort out if it's just testing error or if there is actually something going on there.

I know most fans have a flat spot in their PQ curves, so I would think it's possible to find that same sort of flat spot in a voltage sweep like this. It should theoretically go vertical when the pressures go to that point.

My biggest problem with this method is the need for a way to very very slowly increase voltage. Something like 3-13V in a linear progression over a 5 minute period. Maybe something like gearing up a clock motor to the fan controller knob..

I'm not sure, but it seems like any sort of logging method is going to require a loooong slow test to do any good. Maybe even change the dbA averaging to 5 seconds or longer. This quick test, I ran at 2 second smoothing.

That new lamptron Fc-t has voltage as an setting a display and touch slider seems perfect to replace your kaze with. Maybe you can get them to put a custom firmware on there for you tester edition

Sent from my X10a using Tapatalk

[Vapor]

I know most fans have a flat spot in their PQ curves, so I would think it's possible to find that same sort of flat spot in a voltage sweep like this. It should theoretically go vertical when the pressures go to that point.

I don't think you'll see the 'flat spot' in the PQ curve by doing a voltage/RPM sweep. Every RPM has a unique PQ curve, but once you know a single PQ curve at any RPM, you can figure out the PQ curve at any other RPM because pressure scales with the square of RPM and flow scales linearly with RPM.

Long and the short, you'll only get the flat spot if you do a pressure sweep. By varying voltage/RPM, you're just scaling the PQ curve and where it intersects with the pdrop curve of the rad+chamber. If you think of it the PQ curve as a stretchable shape (that scales with RPM), I'd guess you're hitting the same region of the shape at all tested RPMs as opposed to changing where on the shape you intersect. Not sure if that makes sense typed out, but it does in my mind

You up for testing more fans? I have a spare San Ace 109R1212H1011 and am curious about the new Zalman SF3 (and am placing an FCPU order shortly).

[Martinm210]

I don't think you'll see the 'flat spot' in the PQ curve by doing a voltage/RPM sweep. Every RPM has a unique PQ curve, but once you know a single PQ curve at any RPM, you can figure out the PQ curve at any other RPM because pressure scales with the square of RPM and flow scales linearly with RPM.

Long and the short, you'll only get the flat spot if you do a pressure sweep. By varying voltage/RPM, you're just scaling the PQ curve and where it intersects with the pdrop curve of the rad+chamber. If you think of it the PQ curve as a stretchable shape (that scales with RPM), I'd guess you're hitting the same region of the shape at all tested RPMs as opposed to changing where on the shape you intersect. Not sure if that makes sense typed out, but it does in my mind

You up for testing more fans? I have a spare San Ace 109R1212H1011 and am curious about the new Zalman SF3 (and am placing an FCPU order shortly).

That makes sense, I guess I was thinking the flat point would be sliding around a little, but I have not seen it yet in this sort of fixed restriction testing.

The only odd bumps I'm seeing are the harmonic ones. I'm still suspect of this particular GT fan, attached below is the logged sound level over a rolling voltage sweep type test. Perhaps this fan has been sitting around too long.

Anyhow, I sure would be interested.. Those are a couple I've really been interested in as well.

[the finisher]

I'm working on processing the noctua, and it also had some similar resonance spots. I suspect it's the RS120 radiator just resonating with certain fans since I've now experienced it with two different test rigs. I guess I wouldn't worry too much about those resonance spikes that come and go, they will probably vary on each and every setup anyhow. Just because it happens doesn't mean it will or wont on other setups. Just one of those things that happens with vibration and sound.

This is the main reason I'm testing on a radiator. Fan specs and fan testing in open air scenarios just don't complete the picture for me. This, while limited as well, at least gives me some awareness about the things that can happen. Sound and vibration induced noises are interesting, and much more complex than a simple dB or dBA number..

Thank goodness for fan controllers so we can tune those things out. That's also why I'm recording these, so I can share the little details...good and bad.

Maybe someone in US has a San Ace 9G1212H101(1) they can lend you? They are quite popular. In any case, just a suggestion.

Not that I know of, but I'll keep an eye on that one.

I have several, in different speed's as well. Happy to loan them out, I'll pay shipping.

Want to try a 140L? You'll need 24 volts for full speed, they start @ >5vdc, nice @12.

PM me, out of town for 2 weeks though

[Vapor]

If the finisher can send 9G San Aces, that will compliment the 109Rs I'm sending and we can finally get a good comparison

Also sending Martin a few other fans other than the 109Rs and the ZM-SF3, should be interesting

Martin, can you swap the axes from the bottom chart?

[Martinm210]

Martin, can you swap the axes from the bottom chart?

Sure.

I was kind of waffling on which way to go on the dBA vs CFM. I think I've done it both ways. The first way makes higher better, and this way makes lower better.

Which is easier to read?

FYI, having troubles with interpolating in Excel. For some reason the trendlines are not checking as they should. Can't figure out why, sometimes it works, sometimes it's off a little, sometimes it's way off. Maybe a significant figures thing. Anyhow, seems like actual logging is the only way to really do this, and I'd still need a means to dial voltage up very slowly and consistently. I was thinking of using a big wall clock, taking the minute hand and gearing it up like 40:1 or so...and attaching it to a fan controller. Consistency on this voltage sweep is going to be critical. There is probably an electronic way to do this too, but I'm clueless about that.

Still several areas I'd like to improve on this test method, but I do like the logging of dBA, seems to be fairly worthwhile and would be nice to start a test and walk away....come back to a pile of logged data.

[the finisher]

If the finisher can send 9G San Aces, that will compliment the 109Rs I'm sending and we can finally get a good comparison

Also sending Martin a few other fans other than the 109Rs and the ZM-SF3, should be interesting

Martin, can you swap the axes from the bottom chart?

If you can let me know where to send them, I'll box 'em up now.

Leaving for San Fran/San Diego on Sunday PM, my Wife can ship them next week.

You want a H speed, and M speed? I swear the M's under volt more quietly than the H's. Be cool to see for sure
[9G1212H101_2600 rpm]
[9G1212M101_1950 rpm]

2600 rpm is too much for my taste anyway.


I also have some M speed Type S [9S1212M401] They have almost identical specs to the 1850rpm Nidec GT's, be an interesting comparison, especially to me, bought quite a few.
You can actually get these easier than the GT's, although they are a bare fan.

http://db.sanyodenki.co.jp/product_d...master_id=2017

These Fan's are new, I've just tested that they work. Let me know soon, I can pin & sleeve them before I go.


Edit:Got your pm, they will be sent.

[Vampiyer]

Oh hell yes!

Are you sending the Sanyo Silent series "M" fans? I'd love to see those tested. I've suspected that those were the fans to beat the GT.

Martin: Thanks for all the new tests...geez, that GT of yours is a whiny brat... The odd thing is that my GT fans do not exhibit the same resonance pattern when taken through their voltage range. There is some resonance excitation, but not as high in amplitude nor as offensive (but it IS there).

I think you were right when you suggested that the fans have been optimized for 12v operation. I use mine at 9.0 - 12.0v, and there seems to be at least one small region of resonance within that range that I have to avoid. I go above or below it, but I had assumed it to be an artifact of my fan controller. It looks like you've proved it to be the fan itself.

I suppose that makes sense because my San Ace fans (109R) do not whine at all, but they do tick a bit at lower voltages (the apparently common trait of 38mm thickness fans).

[the finisher]

Oh hell yes!

Are you sending the Sanyo Silent series "M" fans? I'd love to see those tested. I've suspected that those were the fans to beat the GT.

Sending And it might beat the GT, I have both and like the sound of the San Ace better.
I'm unsure if it has as good of pressure on a rad? Awesome case fan though.

[Hondacity]

i never liked the m401s....just me

[the finisher]

I've only used the M101's on rads, be interesting to see how the 401 stacks up to the GT with actual testing.
As Martin has stated, a fan's sound can be a personal preference thing

I do like the GT I have, I usually tune my fans rpm to avoid any harmonic spikes on a given application.