[SOLVED] Do Asus Motherboards have overcurrent protection on fan headers?

[objecttothis]

I called Asus technical support and they were useless. The motherboard manual lists the fan header current limit (1A for most, but two headers with a 3A limit).

I just bought an Asus ROG Crosshair VIII Hero Wi-Fi and I have three daisychained Silverstone Slimmer 120mm PWM fans. Those fans each draw 0.32A, so a total maximum power draw of 0.96A. I would like to connect them to one of the 1A PWM fan headers on the board. Does anyone know if, for some reason, the fans draw more than their rated 0.96A and exceed the 1A limit of the header, does the board have overcurrent protection to prevent damage to the header, or am I risking damaging the fan header by running fans capable of coming so close to the current limits?

 

[Lutfij]

Why not just pick up a PWM fan hub and then have them on that, split up the fans, then in the BIOS or Asus's app set fan curves/profiles?

As for your question regarding the amperage/overcurrent protection, so long as you're under 1A, I'd say you're fine though the cautious part of me would advice not having all eggs in one basket.

 

[objecttothis]

I've been running a Corsair Commander Pro and Commander Core XT and they have many drawbacks
1- Corsair iCue does not play well with others, causing fan and lighting issues.
2- If iCue isn't running, my fans ramp up to full speed because they are based on the water temp sensor connected to the commander pro.

PWM hubs take up a lot of space. The Crosshair VIII Hero has flow and 3 temp headers as well as two pump headers (1 being 3A) 5 1A PWM headers and 1 3A PWM header. I'm moving all my fans and sensors to the board so that I can get away from fan hubs. I plan to use a PWM fan distribution block to run 6 fans off the 3A PWM header.

I think I've just decided that since my water pump has a SATA power cable and a PWM header, I'm just going to connect it to the AIO pump header (1A) and use the 3A pump header for my 3 fans which are daisy chained. I'm assuming that the Asus BIOS will allow me to set my own fan curve on that header.

That said, my question still stands. 0.96A is very close to the 1A current limit. I understand the risks. I'm just wanting to know if the headers have overcurrent protection to cut off power to the header should fans on a particular header draw too much current.

 

[drea.drechsler]

I've been running a Corsair Commander Pro and Commander Core XT and they have many drawbacks
1- Corsair iCue does not play well with others, causing fan and lighting issues.
2- If iCue isn't running, my fans ramp up to full speed because they are based on the water temp sensor connected to the commander pro.

PWM hubs take up a lot of space. The Crosshair VIII Hero has flow and 3 temp headers as well as two pump headers (1 being 3A) 5 1A PWM headers and 1 3A PWM header. I'm moving all my fans and sensors to the board so that I can get away from fan hubs. I plan to use a PWM fan distribution block to run 6 fans off the 3A PWM header.

I think I've just decided that since my water pump has a SATA power cable and a PWM header, I'm just going to connect it to the AIO pump header (1A) and use the 3A pump header for my 3 fans which are daisy chained. I'm assuming that the Asus BIOS will allow me to set my own fan curve on that header.

That said, my question still stands. 0.96A is very close to the 1A current limit. I understand the risks. I'm just wanting to know if the headers have overcurrent protection to cut off power to the header should fans on a particular header draw too much current.

A 1A current rating gives the header a 12W power rating so check the power ratings on your fans. Most modern fans are very low power draw and you should be able to run two at least, quite likely three, even on the 1A header.

Running the pump on a SATA cable is a good way to run it since it needs to run full speed all the time anyway. But does it offer a way to monitor it's RPM? That might be a single wire on a 3-pin fan plug that you can connect to any fan header. That way you an set an alarm to monitor the pump RPM in case it should fail. Pretty important unless the system provides it's own means of alerting you to a pump failure.

 

[objecttothis]

A 1A current rating gives the header a 12W power rating so check the power ratings on your fans. Most modern fans are very low power draw and you should be able to run two at least, quite likely three, even on the 1A header.

Running the pump on a SATA cable is a good way to run it since it needs to run full speed all the time anyway. But does it offer a way to monitor it's RPM? That might be a single wire on a 3-pin fan plug that you can connect to any fan header. That way you an set an alarm to monitor the pump RPM in case it should fail. Pretty important unless the system provides it's own means of alerting you to a pump failure.

As noted in the OP, the fans in question are 0.32A PWM fans. At 12V, that means they are 3.84W fans. Three of them together gives a stated maximum current draw of 0.96A with 11.52W of electricity. Both numbers are under the stated maximum of 1A and 12W for the fan header, but only just which is why I was asking about whether the motherboard header had overcurrent protection to keep from damaging the motherboard should the fans try to draw more than their rated power when being run at full speed.

Regarding the pump, sorry I was not more clear. It's a D5 pump with a SATA for power AND a 4-pin with just the PWM and RPM wires connected. I will control the pump from the 1A AIO header since it doesn't need the 3A "W_PUMP+" and use that header for the fans unless someone can say for sure that the board has some overcurrent protection. The only thing that will prevent me from using that header is if the Asus software doesn't want to allow me to set a fan curve because it thinks that header should only be for pumps... given that it shouldn't be able to tell the difference, I'm probably fine, with perhaps the only annoyance being the software mislabeling the fans as a pump.

 

[drea.drechsler]

.... The only thing that will prevent me from using that header is if the Asus software doesn't want to allow me to set a fan curve because it thinks that header should only be for pumps... given that it shouldn't be able to tell the difference, I'm probably fine, with perhaps the only annoyance being the software mislabeling the fans as a pump.

Speaking to that concern...I can tell you I was dissappointed in my Asus TUF B550m PLus for that reason. That being: the CPU Aux (IIRC) header, intended for a liquid cooling system pump, is fixed at +12V constant output, no change allowed. That may not translate to a Crosshair board but it does indicate that Asus isn't above gimping boards in the pettiest of ways.

 

[objecttothis]

Speaking to that concern...I can tell you I was dissappointed in my Asus TUF B550m PLus for that reason. That being: the CPU Aux (IIRC) header, intended for a liquid cooling system pump, is fixed at +12V constant output, no change allowed. That may not translate to a Crosshair board but it does indicate that Asus isn't above gimping boards in the pettiest of ways.

Balls. I did see in the manual where it says those headers have a default speed of "full speed", but I assumed that meant that it ran at full speed until or unless you set a fan curve to prevent the pump not running while in boot. I really hope Asus didn't screw this up by doing something stupid like only allow the pump to run at 100%. Pumps should not be constantly running at 100%. That's just dumb.

 

[objecttothis]

If that does end up being the case, I'll just stick the pump on one of the chassis fan headers and use a splitter on two of the other fans. They say you can't fix stupid, but usually you can work around it.

 

[drea.drechsler]

... Pumps should not be constantly running at 100%.

For an AIO? I disagree...100% is needed since they're pretty weak pumps (in terms of flow volume) anyway. A closed loop system might be different, with much higher capacity pumps and liquid volume. A decent AIO is going to be nearly silent at 100% too.

 

[objecttothis]

For an AIO? I disagree...100% is needed since they're pretty weak pumps (in terms of flow volume) anyway. A closed loop system might be different, with much higher capacity pumps and liquid volume. A decent AIO is going to be nearly silent at 100% too.

You're probably right. I am using a D5 pump in a custom loop, so I run it at 30% until water temp starts to pick up and then ramp it up from there. I may adjust this slightly if the flow sensor is reading less than 0.5GPM, but I still wouldn't want to run the pump at full speed. Above 1.5GPM the performance gain in negligible.

 

[drea.drechsler]

You're probably right. I am using a D5 pump in a custom loop, so I run it at 30% until water temp starts to pick up and then ramp it up from there. I may adjust this slightly if the flow sensor is reading less than 0.5GPM, but I still wouldn't want to run the pump at full speed. Above 1.5GPM the performance gain in negligible.

Something almost everybody forgets (or doesn't understand) about liquid cooling is the thermal properties of water (the primary liquid in AIO's and CCL's).

Water has immense capacity to readily absorb heat, and so keeps cooling the CPU well after the point an air cooler becomes thermally saturated. But water also gives up that heat very slowly. The loop stays warm well after CPU activity, and core temperatures, have dropped. Since the motherboard fan headers follow CPU core temps pump speed also drops dropping the rate of water exchange in the radiator along with fan speed and that means it's not able to use the full capacity of the radiator to cool it back down. That's another reason I like to keep fan speeds higher on the rad, and pump speed at max on my AIO.

Also, you won't really be able to measure that performance as CPU temperatures. You'd have to create an all-new metric nobody reports or measures. Something I'd call recovery time: time period after which the cooling system has returned to full cooling capacity. With air cooling it's almost instantaneous, which is it's best advantage.

 

[objecttothis]

This is exactly why I run radiator fan and pump curves against water temp and not CPU or GPU temps. I hope Asus allows me to select the water temp as the metric for deciding the fan curve. It's what I'm doing in Corsair's iCue software.

 

[objecttothis]

I just found this article (https://www.technewstoday.com/asus-fan-control/) that shows screenshots of where you can change the temp source in the manual fan curve and of a fan curve on the AIO_PUMP header other than 100% constant. Asus might not be completely dumb after all.

 

[drea.drechsler]

I just found this article (https://www.technewstoday.com/asus-fan-control/) that shows screenshots of where you can change the temp source in the manual fan curve and of a fan curve on the AIO_PUMP header other than 100% constant. Asus might not be completely dumb after all.

Is this a feature of the Hero boards? I don't know: so a TUF board line may not have it. A way to upsell to the STRIX line, or even Hero, if it's an absolute must-have perhaps? Another thing my TUF board lacked was DIMM voltage and VRM temp readouts: you had to go to STRIX if you wanted that.

No, Asus isn't dumb by any means. At least their marketing folk aren't if this is what they do to force an up-sell.

 

[Misgar]

It might be interesting to find out if the fan motors have a high startup current (similar to hard disk motors at switch on) and then the current drops back to the specified level when the fan blades are up to speed. On multi-disk arrays, there is sometimes the option to stagger drive startups, to avoid momentarily crowbarring the PSU. I doubt the fan header is susceptible to such short term overloads.

You could experiment and plug in two fans and watch how they spin up, then try three fans and see if there is any noticeable lag or slow down in powering on.

If you've got an IR thermometer, you could aim it at the MOSFET transistor associated with the fan header. If its temperature exceeds 100C with three fans connected, I'd be inclined to drop back to two fans.

 

[objecttothis]

I'll find out soon enough. The tech specs on the Noctua NF-F12 IndustrialPPC 3000 PWM fans I have (https://noctua.at/en/nf-f12-industrialppc-3000-pwm/specification) state max power draw of 0.30A and typical of 0.26A. I think they are including startup power draw. I have an IR thermometer, so I'll see what the mosfet temperature looks like with the fans connected.

 

[objecttothis]

To follow up on previous posts:
- I connected a splitter to the three 0.26A fans and connected those to the W_Pump+ header which has a 3A limit. I was able to tell the BIOS to control those fans with Q-fan rather than 100% all the time.
- I connected the D5 pump to AIO_pump since that is just a PWM signal and that too was able to indicate in the BIOS that it should be controlled by Q-Fan rather than run at 100%.

Presently, none of the headers are even close to using the rated limitations on the current, so it's good.