Scythe Ultra Kaze not running at rated 3000rpm

[Paradoxxx]

Hi all I purchased some case fans namely "Scythe Ultra Kaze 120mm Case Fan 3000rpm" and sad to say they are not operating at the maximum fan speed as indicated by my fan controller when turned to "max". The max fan speed should be close to 3000rpm and I'm currently operating at 2640rpm.

Fan Controller = Scythe KAZE MASTER ACE 5.25" Bay Fan Controller - Black

Here is a youtube video I found with my fan on the same fan controller albeit I have three such fans on my controller each on different channels.

http://www.youtube.com/watch?v=PNIxFM0sG1k&feature=related

In addition I would like to have three fans operated by a single channel, I can't seem to find a cable adapter to assist with this goal, can anyone direct me?

[jmn2519]

Measure the voltage going to the fans. I'll bet you they are not getting the full 12 volts. Waterlogged should be in here any time now to explain how the fan controller can't seem to supply a full 12V.

[Paradoxxx]

Measure the voltage going to the fans. I'll bet you they are not getting the full 12 volts. Waterlogged should be in here any time now to explain how the fan controller can't seem to supply a full 12V.

While I don't have tools to confirm the voltages going to the fan, it does seem like a logical assumption. I await Mr. Waterlogged response.

[RCG_Bex]

It won't be getting enough power. Also you can't run 3x kaze 3000s on that controller, it'll blow under load. Those fans are 7w each!

[Martinm210]

Vdroop, the evil consequence of resistance between your PSU and fan.

I got over 3,000RPM on my sample when giving exactly 12.0V to the fan plug. It's pretty surprising how much Vdroop occurs in simple wiring, and it all depends on the resistance of the wire and the current being drawn. For example, I've got a PMP-450S on the test bench right now with only 1' of molex wire between my PSU and the pump plug. I'm drawing around 2 amps at 18V, and I have to turn my PSU to about 18.5V to see an actual 18V at the pump plug. Most fan controllers are even worse, as it's a bunch of circuit full of resistance.

Only way around that is to use a transformer based fan controller, they can actually increase voltage and make up for any losses due to resistance.

Also fan controllers are generally not very accurate, I've seen many as much as 100RPM off or more, using a multimeter in HZ mode and muliplying HZ x 30 is much more accurate. I have yet to find a fan controller that is very precise in reporting RPM.

Finally there is that RPM spec, which usually says +-10%. Expect sample variance...it's real.

Here is that UK3 test I ran...
http://img718.imageshack.us/img718/2920/r6a4scytheuk3.png

[Church]

Imho it's less bad for fans, as given choice and available at han controller most (i'm guessing) would prefer running fans downvolted with having extra rpm-s just as reserve.
Voltage drop might hurt pumps more, as their head pressure goes down much more relative to voltage drop, and as most run pumps at full speed.

[Waterlogged]

Measure the voltage going to the fans. I'll bet you they are not getting the full 12 volts. Waterlogged should be in here any time now to explain how the fan controller can't seem to supply a full 12V.

:sofa::para:

While being far from an electronics guru (see my sig for proof:)), I can say that the Kaze Master does "steal" a bit of the power. IIRC, when I was doing my "conversion" on a Kaze to full auto control, it only gave something like 11.7V to the headers.

[Johnny87au]

Ahh, going off martins list your only running them at 10v which results in 2640rpm, u need. Full 12v.. Best bet would run them off moles with adapters, at least you'll be hitting the full 12v..

[Paradoxxx]

Thanks for all the replies. Seems my problem may be rectified by either plugging the fans directly into a 4pin/3min molex or by getting a fan controller that has a built in transformer...

Are there any intelligent fan controllers that change fan speeds based on temperature sensors? I ask because it would be nice to have the fans on my rad increase speed when my water temperature rises beyond a certain temperature.

[thegcpu]

Are there any intelligent fan controllers that change fan speeds based on temperature sensors? I ask because it would be nice to have the fans on my rad increase speed when my water temperature rises beyond a certain temperature.


This unit will do just what you’re asking for.

http://www.koolance.com/water-cooling/product_info.php?product_id=751

[waterkooled]

Are there any intelligent fan controllers that change fan speeds based on temperature sensors? I ask because it would be nice to have the fans on my rad increase speed when my water temperature rises beyond a certain temperature.

Another option would be the Aquaero, very nice product.

[Paradoxxx]

Hey guys it looks like I overloaded one of the channels on my fan controller. I had the channel connected to four Yate Loons 140mm (http://www.jab-tech.com/Yate-Loon-140mm-Case-Fan-High-Speed-D14SH-12-pr-4764.html) and now the controller can't even power one fans.

Anyways I'm in the market for a replacement,

Ultimately I would like to have a fan controller that can allow me to adjust fan speeds automatically based on CPU temperatures purchased at a reasonable price. I can't seem to find exactly what I'm looking for.

I've explored two options

Lamptron Touch

Pros


30w per channel
6 Channels
Auto mode
Affordable


Con


Doesn't interface directly with CPU temp.



Aquaero 5

Pros


Interfaces with OS to determine CPU Temp.
20w per channel
Lots of fancy display info.


Cons

Extra channels take up an extra bay
Cost - Expensive!!!


Perhaps its due to my experience with the Scythe KAZE MASTER ACE but 10w doesn't seem to cut it.

I require advice from the guru's of XS to help me meet my goal.

[Shoggy]

That's wrong!

Each channel can take a load of 1.65A which equals to 20W per channel. The overall load of all channels combined should not exceed 60W.

The so called extra channel is a very small controller (named poweradjust 2) which can be placed anywhere in the PC ;)

http://shop.aquacomputer.de/images/product_images/info_images/53082.jpg

[Church]

10W per channel .. that sounds like old, Aquaero4, and even that with powerboost upgrade could handle 24W on first channel and upto 15W on others.

[Paradoxxx]

That's wrong!

Each channel can take a load of 1.65A which equals to 20W per channel. The overall load of all channels combined should not exceed 60W.

The so called extra channel is a very small controller (named poweradjust 2) which can be placed anywhere in the PC ;)

http://shop.aquacomputer.de/images/product_images/info_images/53082.jpg

Thanks Shoggy, I mistakenly read the Wattage rating for the Aquaero 4.

Hmm this puts a new spin on things, I would be able to attach three fans to one channel with ease.

Shoggy while I have your attention


Am I correct in assuming that the Aquaero5 LT is exactly the same as the PRO without the interface

How does the LT mount? I only see a PCB in the pictures..

Regarding the software, can it monitor my CPU temperature (as read by the motherboard sensors) and automatically increase the fan speeds on CH1,2.. etc?

[Martinm210]

The sunbeam rheosmart is a low cost option. I just bought one to try an hooked it up using speedfan as the software. It does it's job in converting one pwm channel into several dynamically adjusted analog voltage channels, but it's a bit limited. It's only as good as the pwm feed. In my case I've only figured out how to ramp up at one temp point, so it's not very smooth of a ramp up down, more of an on/off switch.

But..if you had a gigabyte board with easy tune 6, you could select lower and upper temp point and get a better result.

Anyhow, it's not going to do anything as nice as the others, but for $30 it's a good way to at least get some dynamic noise control put in place.

[Shoggy]

Yes the main features of LT, PRO and XT are all the same. Maybe you want to get the passive block or the waterblock for it to get the maximum power.

The LT variant comes only with 4 standoffs so it can be mount somewhere in the case. It is likely that we will offer some accessory so it can be mounted in a drive bay.

The aquaero can use 4 software sensors. It will get its values from a software like Open Hardware Monitor. You can use these sensor values like any other value so there are no limit for the control. Please not that the software sensors will not be available from the start.

[Vinas]

Sunbeam rheostat extreme = best in terms of power, and fairly cheap too. The only problem is that it is blue, but you can mod it like I did and change all the leds to red. Other than that the Lamptron FC5 is pretty nice, although over time mine has had a problem keeping power fed to four 3,000rpm delta's. The sunbeam rheobus extreme did that with ease. Keep in mind you need to dedicate one channel to each fan.

[Paradoxxx]

The Rheosmart is a nice option however how reputable is the Sunbeam brand? Would I expect the device to "loose" power over time as compared to the Aquaero?

If I attach 3xScythe Ultra Kaze 120mm Case Fan 3000rpm to one channel of the Aquaero would I be on the "edge" of the power supply for the channel and thus why the Heatsink/waterblock will be recommended?

Just a suggestion, it would be nice if I could buy the LT and later mount it with the "interface" that the pro/xt has at a later date.

Thanks for all the feedback.

[Shoggy]

The channels are monitored in several ways and also the power draw is monitored. So if it goes beyond 20W* the channels will be deactivated so I guess you will not be able to run three of them on one channel.

*the aquaero allows a controlled overload for a short moment so that pumps etc. will start even when they need much more power when spinning up.

[Church]

Shoggy: how then things changes with waterblock? Shouldn't then AA5 handle more then 20W? Can that power draw limit monitoring be disabled, so that deactivation occurs only on overheat?
BTW, what were overheat values for AA5?

[Shoggy]

No, the limit is always 20W and can not be disabled. We will also never offer this option since we had enough fun with burned aquaeros in the past because people believed in the (completely wrong) manufacturer ratings for fans, pumps or whatever.

The only reason for the waterblock is to get the maximum from the device which is 60W (everything combined). Without any cooling at all you will only able to run a few fans, with the passive block a pump and a few fans and to push it to the limit you will need the waterblock.

First temperature limit is 85°C where the fan channel is set to 100% to decrease the temperature, back at 65°C (takes only a few seconds) your settings will be used again and the second limit is 95°C where the channel gets deactivated if the temperature still goes up.

[MagisD]

I'm thinking of moving to this but just iffy on how to do 28 1850 gt's on 7 sets of 4.

Sent from my X10a using Tapatalk

[Church]

Ouch. I initially had impression that 20W per channel and 60W maximum was basic AA5 capability, w/o passive heatsink or waterblock, and with waterblock it should noticeably scale up. And now you say it's hardcoded/without way to disable :(.
Imho you really should update description in AA5 thread and product description in shops with note that those numbers are attainable only with waterblock on, and very preferable - with additional numbers for average incase temps for what to aproximately expect with passive heatsink and with no heatsink at all, for others to not get their expectancies too high like i had :/
I myself can only hope that someone hacks firmware to remove that limit as i find it hard to believe that liquid cooled controller maximum is that low. 10 NB PK-3 can use upto 42W on startup, and 24W when running (measured in independent test 4.2W & 2.4W running), definitely above limit of 20W :/. +I'll need another 8 on another channel. It makes it 75.6W startup/43.2W running, seemingly OK, but not from 2 channels with limits on. (and no, as i planned to use other two channels for monitoring 2 pump rpms, i won't be able to spread out fans on more chans (as pumps are PWM ones with separate direct PSU molex for power, they won't use any extra from AA5, but will occupy 2 channels).
Pitty that for 4 watts extra from single channel one should buy poweradjust. Even AA4 had 24W from at least one of chans. And by reading that 60W are only with waterblock, i'm kind of thinking now that without it AA5 might result being even less powerful then AA4(45W total).
EDIT
Hmm, but now i remembered you writing about AA5 handling Laing D5 pump when liquid cooled. So limit on channel is higher then 20W? :/ I'm officially confused now. :shrug:

[Shoggy]

I myself can only hope that someone hacks firmware to remove that limit as i find it hard to believe that liquid cooled controller maximum is that low.
You can be sure that this will never happen since the firmware and also USB communication when flashing a firmware is encrypted and secured against attacks in several ways. The firmware represents more than two years of development which we do not give away for free as readable code so anyone can alter it and much more important: see how it's done.

[Church]

Shoggy: But how about D5 pump (24W nominal power usage) from AA5? Was it only with in-house non-standard firmware without 20W limit in place?

[Shoggy]

We used the D5 with a regular firmware and were able to slow it down to 40%. Any lower setting only caused the pump to stop. So from 40 to 100% you had full control over the pump.

[Paradoxxx]

No, the limit is always 20W and can not be disabled. We will also never offer this option since we had enough fun with burned aquaeros in the past because people believed in the (completely wrong) manufacturer ratings for fans, pumps or whatever.
The only reason for the waterblock is to get the maximum from the device which is 60W (everything combined). Without any cooling at all you will only able to run a few fans, with the passive block a pump and a few fans and to push it to the limit you will need the waterblock.

First temperature limit is 85°C where the fan channel is set to 100% to decrease the temperature, back at 65°C (takes only a few seconds) your settings will be used again and the second limit is 95°C where the channel gets deactivated if the temperature still goes up.

So the 0.6A rating on http://www.jab-tech.com/Scythe-ULTRA-KAZE-120-x-38-mm-Case-Fan-High-pr-3940.html is not accurate?

Considering my fans don't run at the full voltage anyway (as indicated earlier in the thread) I was hoping the math would be

10V * 0.6A * 3 fans= 18W?

[Shoggy]

No idea how much power these fans draw but in most cases the ratings provided by the manufacturer are wrong. I can remember a customer whose aquaero burned down completely and was very annoyed that we would not exchange it since for us it was obvious that the damage was causes by a massive overload. He promised again and again that his fans could not be the problem because in theory everything was within the allowed range. We asked to send us the fans and we measured their power consumption. In the end it was almost three times higher than rated by the manufacturer and while spinning up it was even worse.

I don't say that this is the case with any fan on this planet but I would not blindly trust values from the manufacturers. It could be also less power.

Your little calculation will not work for reality. When you slow down the fans, the temperature of the voltage regulator will go up, so in the end the fans might need less power but because of a higher resistance the unused power will result in a higher temperature. That is also the reason why the aquaero sets an overheated channel to 100% at first since this cools down the voltage regulator significantly within a few seconds.

[Paradoxxx]

With that said, whats the maximum number of fans (model outlined above) would you recommend to be attached to 1 channel of the AA5?

[Shoggy]

Hard to tell because I don't know their real power consumption. If we take the numbers as truth you can only use two of them per channel and would end up with 8 fans.