Zitat von »eisbehr«

Zitat von »Jibby«

What are the 4 wires coming off the fans 1-4 on the aquaero 6 lt? There should be +12V, GND and what are the other two wires?

You ask for solutions and will "design an pcb" to solve the problem, but you don't even know the four connections of a pwm fan?! How will that end up then?! Hilarious.

I did not ask what the 4 connections of a PWM fan are, I asked what the 4 connections and hopefully an explanation of what is coming off of the aquaero 6 lt board so I can see if there is an easy way to modify. I couldn't find a schematic or block diagram of the aquaero 6 lt.

Can you please not respond to this thread? All you are doing is trolling.

I'm trolling?! You are the person not excepting the issue.
And your question makes no sense too. If you know the wires of a pwm fan, you know the connections of the connectors.

And how about looking into you manual you got with your product? Page 8. And because of your always harsh tone: rtfm.
https://aquacomputer.de/handbuecher.html…o_6_english.pdf

Zitat von »eisbehr«

I'm trolling?! You are the person not excepting the issue.
And your question makes no sense too. If you know the wires of a pwm fan, you know the connections of the connectors.

And how about looking into you manual you got with your product? Page 8. And because of your always harsh tone: rtfm.
https://aquacomputer.de/handbuecher.html…o_6_english.pdf

Thank you for showing me the information I needed. The manual that came with my Aquaero 6 lt did not have this information. I can post a photo if you'd like. Like I said, there must be something lost in translation because I have just been asking questions to try to find a solution, I did not mean to come off with a harsh tone.

I bread-boarded a simple circuit based on that information and I now seem to be able to control my Noctua fans. There is an easily fixable solution

Zitat von »Jibby«

It sounds like I can just redesign my own splitty9 and add a +12V to 5.5V switching regulator and then a 5.5V to 5V LDO and this should solve the issue.

I do not fully understand your plans but it seems to me to be nonsense. You can try to add a 4,7kOhm resistor or maybe also a 1k resistor between VCC (12V) and the PWM Signal pin. This could solve the issue but can also kill other fans so you need to be careful.
This is like adding a spring to the pedal I described.

Zitat von »Jibby«

Zitat von »Jibby«

Look, I just came here for facts and was met with rude words and hostility. I am waiting to hear back from Noctua but in the mean time, I wanted to see if there was a way to design a solution myself.

Instead of just explaining the exact issue, I was met with rude and hostile responses, coupled with condescending attitudes so now that I know how support reacts when I have an issue, I will be returning my entire aquaero setup.

Thanks

This is good. Thank you for supplying this information. What are the 4 wires coming off the fans 1-4 on the aquaero 6 lt? There should be +12V, GND and what are the other two wires?

*EDIT*

It sounds like I can just redesign my own splitty9 and add a +12V to 5.5V switching regulator and then a 5.5V to 5V LDO and this should solve the issue.

It sounds overcomplicated. As there is never more than a few mA sourced from pin 4, You will be fine with a cheaper and less problematic linear regulator (7805). I would rather generate 3,3 V (with a 78L33 for example) for pull-up, just to be on the safe side (the pin 4 may be internaly pulled to 3,3 V as it is recommended for newer designs). However, since Noctua suggests push-pull drive for pin 4, You should be fine with pulling up to 5 V anyway.

Actually, if You're no stranger to electric engineering you could open the fan and check exactly what they have done. Noctua says in their white paper that the pin 4 is indeed pulled up internally but they do not say to what voltage (in fact they say 5 or 3,3 V). The pull-up resistor value is certainly to big to act efficiently as a pull-up in this configuration (proably way over 100 kOhm - this is an indirect conclusion coming from observation of this forum and the Noctua whitepaper itself). Even worse, I suspect that the value of this internal resistor may vary between fan revisions.

The 4-pin fan specification is in fact very old and outdated so some things are not clearly specified, this is where all this mess comes from.

Zitat von »caronen«

The 4-pin fan specification is in fact very old and outdated so some things are not clearly specified, this is where all this mess comes from.

The last Intel PWM specification from 2005 is very clear. There's actually no room for special solutions like those from Noctua or EK.

Intel Spec: 1.3 from 2005
This signal must be pulled up to a maximum of 5.25V within the fan.

Zitat von »sebastian«

Zitat von »caronen«

The 4-pin fan specification is in fact very old and outdated so some things are not clearly specified, this is where all this mess comes from.

The last Intel PWM specification from 2005 is very clear. There's actually no room for special solutions like those from Noctua or EK.

Intel Spec: 1.3 from 2005
This signal must be pulled up to a maximum of 5.25V within the fan.

As for "special solutions" I totally agree, even though I'm still curious what their motivation was.

As for Intel specification, I can agree that it is not ambiguous, but I would not call it clear.

It is indeed specified that the fan should have a pull-up to its internal logic voltage but only the maximum current is specified. In my opinion, the allowed input impedance should have been specified as well. Furthermore, even if a compulsory pull-up resistor implies an open collector, this is only indicated in the section for motherboard manufacturers.
I am perhaps too used to read specifications of scientific instruments so maybe this is why I find Intel’s specification not strict enough.

And there is also a question why it is Intel that should make standards… I will not go into this debate but even though an informal standard is better than no standard at all, from moral standpoint Noctua is as entitled to propose standards as is Intel

In fact, when you look more closely to things, Intel’s standard was not so closely followed – there were mobos with inferior current rating on fan pins, pull-up on output is not so uncommon and even push-pull drive can be found. If I'm not mistaken even Aquaero 5 has a pull-up resistor on its sole PWM fan output…

Anyway, would not be useful to make a sticky thread about all this pull-up story since this is not the first time the issue arises ?

Zitat von »caronen«

And there is also a question why it is Intel that should make standards […] from moral standpoint Noctua is as entitled to propose standards as is Intel

Totally! But intel came up with this standard and mainboard and fan manufacturers adapted it. So they all (more or less) rely on that standard that intel invented. Nobody forced them, it was just a practical solution.

But Noctua/EK didn't created their own standards, as they could do. They want to create pwm fans for a widely available system, but not keep on the specifications. So i would love to see an own standard of them, but it would be hard to convince manufacturers of a second system for fan control.

I assume the main problem is: They simply didn't know about these issues. I can only guess, but i assume that both companies did not put their own development into electronics but simply bought in China.
Anyway Intel was the first company with a PWM fan, they specified the Pin out, the connector, the frequency, the kind of control. So anyone in the PC business should comit to this specification.

Anyhting else is just providing problems and we have to loose lifetime for solving this.

Zitat von »eisbehr«

Zitat von »caronen«

And there is also a question why it is Intel that should make standards […] from moral standpoint Noctua is as entitled to propose standards as is Intel

Totally! But intel came up with this standard and mainboard and fan manufacturers adapted it. So they all (more or less) rely on that standard that intel invented. Nobody forced them, it was just a practical solution.

But Noctua/EK didn't created their own standards, as they could do. They want to create pwm fans for a widely available system, but not keep on the specifications. So i would love to see an own standard of them, but it would be hard to convince manufacturers of a second system for fan control.

To be well understood, I have no problem with Intel imposing a standard since none existed before. And it is well defined from an engineering point of view (even though the document is not clear enough for my taste). Well, Noctua at least published a document that kinda explains that they did not follow the standard

Believe me, I have to deal with multiple engineering standards in my work, so I can appreciate an universal standard whether it is formal or not; and I agree totally with Aquacomputer on this. I do not intend to defend these clumsy guys in any way. It's just the situation is not so pleasant for us, the users, as there is no warning that not all PWM fans are not made equal.

But I am also curious why EK and Noctua (but not only) did what they did. I find it hard to believe that they just have bought the stuff in China without looking even if it is not impossible. What kind of money can one spare getting rid of a single resistor? I would like to know if only for the sake of knowing

Zitat von »Stephan«

Zitat von »Jibby«

It sounds like I can just redesign my own splitty9 and add a +12V to 5.5V switching regulator and then a 5.5V to 5V LDO and this should solve the issue.

I do not fully understand your plans but it seems to me to be nonsense. You can try to add a 4,7kOhm resistor or maybe also a 1k resistor between VCC (12V) and the PWM Signal pin. This could solve the issue but can also kill other fans so you need to be careful.
This is like adding a spring to the pedal I described.

My solution is not nonsense. It fixed the issue. It probably is overly complicated but I can provide a clean +5V signal myself. The resistor is a good idea thank you. Just need to make sure that the resistor is the right size so that it does not exceed the +6V intel PWM spec.

Zitat von »caronen«

Zitat von »Jibby«

Zitat von »Jibby«

Look, I just came here for facts and was met with rude words and hostility. I am waiting to hear back from Noctua but in the mean time, I wanted to see if there was a way to design a solution myself.

Instead of just explaining the exact issue, I was met with rude and hostile responses, coupled with condescending attitudes so now that I know how support reacts when I have an issue, I will be returning my entire aquaero setup.

Thanks

This is good. Thank you for supplying this information. What are the 4 wires coming off the fans 1-4 on the aquaero 6 lt? There should be +12V, GND and what are the other two wires?

*EDIT*

It sounds like I can just redesign my own splitty9 and add a +12V to 5.5V switching regulator and then a 5.5V to 5V LDO and this should solve the issue.

It sounds overcomplicated. As there is never more than a few mA sourced from pin 4, You will be fine with a cheaper and less problematic linear regulator (7805). I would rather generate 3,3 V (with a 78L33 for example) for pull-up, just to be on the safe side (the pin 4 may be internaly pulled to 3,3 V as it is recommended for newer designs). However, since Noctua suggests push-pull drive for pin 4, You should be fine with pulling up to 5 V anyway.

Actually, if You're no stranger to electric engineering you could open the fan and check exactly what they have done. Noctua says in their white paper that the pin 4 is indeed pulled up internally but they do not say to what voltage (in fact they say 5 or 3,3 V). The pull-up resistor value is certainly to big to act efficiently as a pull-up in this configuration (proably way over 100 kOhm - this is an indirect conclusion coming from observation of this forum and the Noctua whitepaper itself). Even worse, I suspect that the value of this internal resistor may vary between fan revisions.

The 4-pin fan specification is in fact very old and outdated so some things are not clearly specified, this is where all this mess comes from.

Good idea. I was trying not to void my fan warranty by opening it up but I guess providing my own clean +5V signal also probably voids the warranty so I have nothing to lose.

Zitat von »sebastian«

Zitat von »caronen«

The 4-pin fan specification is in fact very old and outdated so some things are not clearly specified, this is where all this mess comes from.

The last Intel PWM specification from 2005 is very clear. There's actually no room for special solutions like those from Noctua or EK.

Intel Spec: 1.3 from 2005
This signal must be pulled up to a maximum of 5.25V within the fan.

Do the Artic fans work properly with Aquaero pwm?

https://www.arctic.ac/us_en/p12-pwm.html
https://www.arctic.ac/us_en/p14-pwm.html

We should build a fan database for aquacomputer compatibility on pwm and argb.

Zitat von »eisbehr«

We should build a fan database for aquacomputer compatibility on pwm and argb.

That would be a great idea.

Zitat von »Jibby«

Zitat von »sebastian«

Zitat von »caronen«

The 4-pin fan specification is in fact very old and outdated so some things are not clearly specified, this is where all this mess comes from.

The last Intel PWM specification from 2005 is very clear. There's actually no room for special solutions like those from Noctua or EK.

Intel Spec: 1.3 from 2005
This signal must be pulled up to a maximum of 5.25V within the fan.

Do the Artic fans work properly with Aquaero pwm?

https://www.arctic.ac/us_en/p12-pwm.html
https://www.arctic.ac/us_en/p14-pwm.html

Yes they do, I'm using one of each in my rig. Both attached via Quadro

If you supply fixed +5V to the PWM signal there are two things that can happen:

1. You are not able to control the fan as the controller will not be able to pull down this voltage to GND. So the fan can not be controlled.
2. If the controller is not protected by a inline resistor it can burn.

If you like to add a external voltage you need to supply it with a resistor (4k7 for example) to the circuit.

As we would like to be as compatible with our products as possible we are currently developing a solution for this problem that can adjust to the wrong fan behaviour.

Currently these issues have only be confirmed with Noctua fans. And it looks like Noctua has provided different kind of PCBs at the same fans so this is really somehting very strange. I still believe they did a mistake and then did this whitepaper to describe the issue. But that is something you need to discuss with Noctua. I don't find any engineering reason why to change this. Even at industrial devices (for example CNC, SPS) it is common to control devices with NPN signals why the device supplys the control voltage.

At EK fans we found issues that if you combine the RPM signals of the fans (which happens at many splitters like SPLITTY9) the PWM control of the fans will fail. We haven't investigate deeper in this issue. This can simply be solved by removing the pin from the connectors.

In total > 99,9% of all PWM fans will work with our hardware perfectly.

Zitat von »Stephan«

If you supply fixed +5V to the PWM signal there are two things that can happen:

1. You are not able to control the fan as the controller will not be able to pull down this voltage to GND. So the fan can not be controlled.
2. If the controller is not protected by a inline resistor it can burn.

If you like to add a external voltage you need to supply it with a resistor (4k7 for example) to the circuit.

As we would like to be as compatible with our products as possible we are currently developing a solution for this problem that can adjust to the wrong fan behaviour.

Currently these issues have only be confirmed with Noctua fans. And it looks like Noctua has provided different kind of PCBs at the same fans so this is really somehting very strange. I still believe they did a mistake and then did this whitepaper to describe the issue. But that is something you need to discuss with Noctua. I don't find any engineering reason why to change this. Even at industrial devices (for example CNC, SPS) it is common to control devices with NPN signals why the device supplys the control voltage.

At EK fans we found issues that if you combine the RPM signals of the fans (which happens at many splitters like SPLITTY9) the PWM control of the fans will fail. We haven't investigate deeper in this issue. This can simply be solved by removing the pin from the connectors.

In total > 99,9% of all PWM fans will work with our hardware perfectly.

Ok thank you for the information and thank you for working on a solution.

Just to clarify, I was providing 5V to PMOS/NMOS configuration described in the noctua white paper. I agree, the resistor is a good idea for noise reduction/limiting current.

Noctua is taking their sweet time getting back to me on a potential fix so I may just go with the artic fans. They look great and are 1/3rd the cost

Zitat von »Jibby«

Zitat von »Stephan«

If you supply fixed +5V to the PWM signal there are two things that can happen:

1. You are not able to control the fan as the controller will not be able to pull down this voltage to GND. So the fan can not be controlled.
2. If the controller is not protected by a inline resistor it can burn.

If you like to add a external voltage you need to supply it with a resistor (4k7 for example) to the circuit.

As we would like to be as compatible with our products as possible we are currently developing a solution for this problem that can adjust to the wrong fan behaviour.

Currently these issues have only be confirmed with Noctua fans. And it looks like Noctua has provided different kind of PCBs at the same fans so this is really somehting very strange. I still believe they did a mistake and then did this whitepaper to describe the issue. But that is something you need to discuss with Noctua. I don't find any engineering reason why to change this. Even at industrial devices (for example CNC, SPS) it is common to control devices with NPN signals why the device supplys the control voltage.

At EK fans we found issues that if you combine the RPM signals of the fans (which happens at many splitters like SPLITTY9) the PWM control of the fans will fail. We haven't investigate deeper in this issue. This can simply be solved by removing the pin from the connectors.

In total > 99,9% of all PWM fans will work with our hardware perfectly.

Ok thank you for the information and thank you for working on a solution.

Just to clarify, I was providing 5V to PMOS/NMOS configuration described in the noctua white paper. I agree, the resistor is a good idea for noise reduction/limiting current.

Noctua is taking their sweet time getting back to me on a potential fix so I may just go with the artic fans. They look great and are 1/3rd the cost

I have 15 of the Artic P12 PWM PST fans in my build right now and I have to say I am very impressed with them. At max speed they are very quiet so they can be ran at a high RPM and they move a good bit of air.
I also decided to pick up 3 of the Infamous Noctua NF-A12x25 pwm's and compared to the artics I don't think they are worth it. They are louder than the Artics by a good bit once you cross over the 1800rpm threshold and as far as moving air through a radiator I feel the Artics do a fine job and considering the price of them it's a no brainer especially if you want to do push/pull.

Zitat von »Jibby«

It sounds like I can just redesign my own splitty9 and add a +12V to 5.5V switching regulator and then a 5.5V to 5V LDO and this should solve the issue.

Any chance you could explain this fix in laymans terms? I am an engineer, but not the electrical variety. I have a good multimeter, and I'm pretty handy with a soldering iron, but I don't posses a scope, or any more advanced equipment. Is this something I could do with my existing tools and a few components ordered online?

Zitat von »Stephan«

You can try to add a 4,7kOhm resistor or maybe also a 1k resistor between VCC (12V) and the PWM Signal pin.

With full understanding that Aquacomputer takes no responsibility if I wreck my equipment or burn down my house, would you mind explaining a little bit more in detail what you mean?

Are you talking about bridging the +12V line (pin 2, yellow) and PWM signal (pin 4, blue) with two resistors in series, one 4.7K ohm and one 1k Ohm?

I almost $500 in new Noctua fans for my new build which are past the time window for returns. I ahve reached out to their customer support, but I have no idea what I can expect from them. I may be on my own and have to figure out a solution.

Consider me pretty annoyed at Noctua right now.

I ahve also posted a news post in the HardForum community about this issue.

I'm hoping the added attention helps bring about a solution, but if it doesn't, I'd appreciate any and all suggestions on workaround in laymans terms.