I routinely run 4-7x PWM fans on a single PWM signal, but have ran as many as 12x fans on a single PWM signal source.

I don't know how well built the Thermaltake 10-way PWM splitter hubi is. I quit uisng anything with their name on it long ago. Maybe they are okay. I know the Swiftech 8-way is good, and have heard the Phobya 8x PWM controlled fan hubs are also good. One of them with a simple 'Y' spliter on on header should work just fine. 8x PWM fans on a single signal almost always works, 10 sometimes do not. Only way to find out is try it and see.

But using same PWM signal does not mean all fans will all run the same speed. It only means all fans receive the same PWM signal. Different fans have different rpm range motors as well as different PWM% to rpm ratio programs in the fans' PWM circuitry.

Here are a few examples of different PWM % to RPM curves.


Original TR-140 & TY-143 are basically a straight line from idle to full speed, but newer models are not. TY-147A & TY-147A SQ idle at lower speed, then straight line to about 85% where they increase speed faster to 100%


All three are using same PWM % to rpm ratio. The difference is fans have different speed motors.


The Cryorig PWM % to rpm curve in fan is very different from Thermalright PWM % to rpm. Cryorig starts to increase speed at very low PWM signal % and reaches maximum rpm at 70-80% PWM signal .. while TY-140 / 147 and TY-143 stay at idle until PWM signal is up to about 40%, then straight line to about 85% where they speed up faster to full speed at 100%

Hope this helps.

 

Quote:

Originally Posted by doyll

But using same PWM signal does not mean all fans will all run the same speed. It only means all fans receive the same PWM signal. Different fans have different rpm range motors as well as different PWM% to rpm ratio programs in the fans' PWM circuitry.

Hope this helps.

Thanks, I always wondered about that. I understood that speed would be different, but I didn't know the PWM curves were different too.
Is that what's meant when it's said "Computer fans don't follow PWM standards."?
Or is there more "goofy" stuff too.

+Rep

 

Quote:

Originally Posted by billbartuska

Thanks, I always wondered about that. I understood that speed would be different, but I didn't know the PWM curves were different too.
Is that what's meant when it's said "Computer fans don't follow PWM standards."?
Or is there more "goofy" stuff too.

+Rep

I have no idea what "PWM standards" is supposed to mean. While we might think it's a 1:1 and/or linear ratio, as far as I know there is no such thing as a 'standard' PWM signal to 12v power pulse. It is simply whatever the makers decide they want it to be. Almost of my variable speed power and hand tools are PWM controlled now. Many even regulate the rpm so it remains constant regardless of load. For instance if I set a belt sander, router or shaper at a specific speed, that is what they run with or without load. For example, if shaper / router is set to 8000 rpm instead of 20000 rpm to run a larger cutter, even though the larger cutter requires more power to cut, he speed they run when not cutting material changes very little when they start cutting, and if they do drop in speed as they start to cut they pick that speed but up almost instantaneously. They are also 'soft starting' .. meaning when they are turned on they gradually apply power and pickup speed so there is no sudden load applied to them. Kinda like slowly apply throttle while slowing letting out the clutch in a car instead not slamming the throttle down and dropping the clutch.

 

NEMA, SAE, someone must have standards.

Intel

 

Quote:

Originally Posted by billbartuska

NEMA, SAE, someone must have standards.

Intel

Yeah, well, anyone can make up some 'standards'. You and I can make up some 'standard's but I doubt anyone will pay any more attention to them than they do all the so called 'standards' there already are out there.

We don't even have a 'standard' of what motherboard headers are PWM and what are not. Asus sold thousands of motherboards claiming their fan headers were 'PWM' when there was nothing on pin-4 at all. What 'standard' were they using?

I guess the point is I don't think there are any 'standards' saying what a pwm pulse signal rate is in reference to power pulse rate .. which supported by the PWM signal percent to 12v pulse creating so many different PWM / fan rpm curves for fans with similar minimum and maximum rpm. Good example is the Cryorig XF and XT curves compared to TY-140, 147 and TY-147A .. and TY-147A curve compared to TY-140 and TY-147. posted above. All were run on same Aquaero 5 PWM signal with same PSU powering them.

 

Thanks for the info guys. It is only a $20 product, so I guess I will take a leap of faith and try it out.

I would try the Silverstone one except that it only has 8 ports and I have 9 fans

. I did see that silverstone amongst my searches and I did like the build quality.

Doyll, thanks for the reminder on PWM curves. I saw these on ThermalBench for the Phanteks fans I wish to use on the build. I read the review thouroughly, and from what I understood it has a slight PWM Signal-to-RPM curve drop, however it was reported as "not that bad".

Based off the information in the reviews I believe the Phanteks MP series is going to be great.

Thanks for all the help guys, Im going to be placing orders this morning!

 

Quote:

Originally Posted by doyll

I routinely run 4-7x PWM fans on a single PWM signal, but have ran as many as 12x fans on a single PWM signal source.

I don't know how well built the Thermaltake 10-way PWM splitter hubi is. I quit uisng anything with their name on it long ago....

I can see why you made the decision to stop using their products. For the most part I cannot find products they make to be of any quality or real performance. I will never buy one of their cases, power supplies, or fans. However, weirdly, as I saw before, they do seem to be pretty good with odd-ball accessories. When my younger brother bought his MSI Gaming Laptop, I referred him to the Thermaltake Massive 23 GT for a good cooling pad. When it arrived we both realized that it was indeed a very good product. Very silent, very powerful additional airflow for his twin-fan cooling system.

Its weird really, I have owned Thermaltake Flagship products, and I hated every last one of them, but then every small accessory Ive ever tried from them is either excellent or good enough to not complain about it. Am I the only one who finds this odd? LOL

 

I'm actually currently working on building my own custom fan controller and was planning on doing a write up in the near future. The fan controller will be aimed at keeping the water temp of the custom water loop within a specified temp spec (or delta t). So far I've implemented a basic design with Arduino and Raspberry Pi, but i'm still working on which one i'll end up using in the long run. Most likely the Arduino due to its very low cost and simplicity. The RPi has a higher entry cost as well as requiring several external modules to implement A2D for the thermistor to read water temps as well as PWM modules to drive the fans. But with the RPi you could implement a pretty nifty GUI on a touch screen built into the case for some manual control.

Right now i'm trying to find a flow meter so i can monitor the water flow and use that as input as well. Since i'm not going to be monitoring cpu or gpu temps with this device, i need to make sure i'm not running the pump at too low of speeds while not getting a high water temp reading. While this is not necessary, it would be more as a failsafe. My idea was to possibly incorporate some sort of notification system that would let the user know that the water flow has stopped (in case of pump failure) or the water temp is at unusually high temps (fan failure). Another aspect I would like to implement is to incorporate an ambient air temp sensor as well since the water temps will be associated with the delta t. So the pump and fan speeds will not only account for the current water temp, but for a target delta t.

As of now the system is working with 1 temp sensor and 1 pwm fan. just need to get it all into a neat package and add more fan headers. Also it seems that my PWM DDC3.2 pump wasn't accepting the PWM signal, but I need to do a bit more testing to weed out some other variables as to why it wasn't working as intended.

And of couse I want/could use this to control some custom LED lighting.

Long story short, if you have some technical ambition, you could build your own to your exact specification at a very low cost. Or wait for my step by step write up LOL

 

Quote:

Originally Posted by IMI4tth3w

I'm actually currently working on building my own custom fan controller and was planning on doing a write up in the near future. The fan controller will be aimed at keeping the water temp of the custom water loop within a specified temp spec (or delta t). So far I've implemented a basic design with Arduino and Raspberry Pi, but i'm still working on which one i'll end up using in the long run. Most likely the Arduino due to its very low cost and simplicity. The RPi has a higher entry cost as well as requiring several external modules to implement A2D for the thermistor to read water temps as well as PWM modules to drive the fans. But with the RPi you could implement a pretty nifty GUI on a touch screen built into the case for some manual control...

Hmmmm... This is a very neat idea, but it sounds like your interface will probably be through a 5.25 Touch Screen panel, wont it? Im not planning on installing my 5.25 Drive bays, unfortunately, so I think I will have to pass on this occasion, however I do very much like the sound of your idea. Do be sure to share with us or send us a link to your write-up, I'm gonna have to see this thing for myself, it sounds pretty awesome!

 

Quote:

Originally Posted by WallacEngineer

Thanks for the info guys. It is only a $20 product, so I guess I will take a leap of faith and try it out.

I would try the Silverstone one except that it only has 8 ports and I have 9 fans

. I did see that silverstone amongst my searches and I did like the build quality.

Doyll, thanks for the reminder on PWM curves. I saw these on ThermalBench for the Phanteks fans I wish to use on the build. I read the review thouroughly, and from what I understood it has a slight PWM Signal-to-RPM curve drop, however it was reported as "not that bad".

Based off the information in the reviews I believe the Phanteks MP series is going to be great.

Thanks for all the help guys, Im going to be placing orders this morning!

The PH-F140MP is good. Here is graph of of it with some others

 

Quote:

Originally Posted by doyll

The PH-F140MP is good. Here is graph of of it with some others...

Yep, that's pretty much exactly the same results posted in the "Phanteks PH-F140MP and PH-F120MP Review" I found on ThermalBench. That weird dip and then gets pretty much linear and finally at the top it suddenly decides to make up for the drop at the bottom.

As such, I have just placed my orders. 5 X PH-140MP, 2 X PH-120MP, 2 X PH-140HP II, DeepCool RGB 360 Lighting Kit, Thermaltake Commander FP PWM Hub, and 2 X Phobya Sleeved SATA Power Extention cable - 12".

 

Quote:

Originally Posted by WallacEngineer

Hmmmm... This is a very neat idea, but it sounds like your interface will probably be through a 5.25 Touch Screen panel, wont it? Im not planning on installing my 5.25 Drive bays, unfortunately, so I think I will have to pass on this occasion, however I do very much like the sound of your idea. Do be sure to share with us or send us a link to your write-up, I'm gonna have to see this thing for myself, it sounds pretty awesome!

Thanks, hopefully I can get it all put together and working pretty soon.

So the general idea consists of 2 different controller types. One would be a much more simple Arduino controller that would have a program already written that the user can go in, look at, and change some variable values in the beggining for their specific set up. Write the code to the arduino and forget about the controller while it does its thing.

A second more advanced version would be done using a RPi and could/would feature some sort of interface touch screen for doing manual on the fly adjustments and monitoring. The interface i had pictured in my head would be done using a 7" RPi touch screen that could be molded into the case for example. But there are virtually unlimited ways of doing this. But many cases are ditching the 5.25" bays and i'd like a design that didn't need one either.

A final option i was looking into was to create a GUI/program for the PC, similar to lets say the Corsair Link software but much more basic, for making manual changes to the controller. This could apply to both the arduino or RPi configuration.

But anyways once I get the Arduino implementation working, i'll clean up the code so that it can be easily modified and customized depending on the user setup and share it with everyone. Then maybe look into how far the rabbit hole can go with the RPi.