Just keep the amp draw under what is said in the manual for that particular header.

 

5 fans per header.

 

12w/12v=1a

 

they've been measured to draw 0.36A at max speed: Phanteks PH-F120T30: The fan that pushes boundaries - Page 36 of 44 - HWCooling.net

I can check for you later how much mine draw in the 2000rpm max mode

 

All motors draw about 4 times more power starting up than they use while running. Typical computer fan specs are running amp rating, not start-up. Start-up time is the time fan takes to spin up to full speed, so for a second or 2 each fan is drawing about 4 times higher amps than it's running amp rating. This means your 0.20a fans will pull 0.8-1.0 amps when starting up. 12 watts 12 volts is 1 amp. So your headers are safe to run only one 0.20 amp fan. Ever notice peeps posting about burning out a fan header always seem to say they noticed fans were not running right after starting up their system?

 

How does it work then when phanteks fans are designed to be daisy chained together they click together you see? So is that then not safe?

 

Don't even need a controller. Just get a hub with external power and you're good to go. For example the Arctic Case Fan Hub.

 

I quit using my fan headers to power fans. I just use powered hubs now.

No regrats.

 

For me, I noticed my mobo lights were brighter when I had no fans plugged into the board. Voltages still showed the same, but whatever.. had a spare Molex powered hub, and my case came with a SATA powered hub. Makes it a bit cleaner in the main chamber too.

 

I'm guessing your mobo lights are very, very low power consumption, so are registering (brigtening) with slight increase in voltage and whatever is monitoring voltage isn't accurate enough to register it.

I have several custom made SATA powered PWM controlled hubs from years ago that I've used in several builds. I'm guessing very similar circuitry to Arctic, Thermalright, XSPC and others using mobo PWM fan plug and SATA power.

 

I got 6 d30-120 + 3 x 140 + pro watersys , PWM in bios 50% on all

 

Do you mean you just run them at 50 percent on the fan curve?

 

I run 2x sets of 4x Noctua A12x25, and 1x set of 3x Noctua A12x25 off motherboard fan headers (1 header = 4 fans, 1 header = 4 fans, 1 header = 3 fans) using splitters. I really wouldn't be afraid of this. I'm not going to debate the reality of the amperage/wattage usage during start-up, but to recommend only 1 fan per header is lunacy, and the amount of time it would take to burn-out the fan header would be years, and even then, maybe not. I've thought whether the motherboard spinning my fans to full throttle during POST would cause trouble, and figured I'd see what happens... it's been over a year, no issues. Years and years of motherboards in the past putting at least 3x fans per header without a single fan header burning out. For context, I don't run my fans full speed unless I'm benching, and I did plenty of that earlier in 2024.

If it's a cheapo motherboard, I'd be weary... but I dunno, it's not 1999 anymore. The headers can take the "abuse".

Going with a fan controller works too.

 

Most consumer fans are pretty sucky, and you should be able to run at least 3 from a header, probably more like 5 or 6.

 

But how do we know the fan header rating is peak draw or sustained draw? Because I'm willing to bet that the rating is just based on the amount of copper the trace has and the connector's physical design, so it's more about sustained amperage/ root mean square amperage draw. I personally just use the rating on the fan sticker... But, I just generally never have more than 3 fans on one header.

 

I have this little 90x38 DC Koala III rated for .31A and 4w, this little fecking fan will pop your header if you try to hot plug it. Only took me 3 boards to figure that out, one header for each board.

Kick axe fan though

 

Could you cold plug it?

Like this one?

Saw some in a used electronics shop, but never used one.

 

My whole argument is that start-up amperage is too quick in duration to thermally damage anything, so the rating in the manual is probably for continuous use.

 

I tend to agree, and experience aligns with this thinking. Even with the long POST time due to memory training causing all 11 fans to run full speed for a long time now. no issues.

Every day you use your PC, you add wear-and-tear to it .. as far as I can tell, even with 4x A12x25 demanding start-up power off a single fan header, it’s not causing damage to risk longevity/health of the header, at least not within reasonable time-frame. If it dies in 10 years rather than 12, I think that’s acceptable.

 

I've been trying to show how much many overload fan headers.

I admit I've baited a few peeps. It's just friendly banter with no malice intened.

As for amp rating of fan headers, they give a single rating, the safe limit for continues load.

Most of our fans (unlike Nidec & Noctua) don't give us anything but running load ratings.

Some users seem to think 3-5 times higher load at startup isn't a problem.

I've seen many burnt out fan headers and and many burnt traces who are still working and know that every one of them happened at startup load.

Like I keep saying, what you do is your business. I'm not going to run more than a couple, maybe 3 on single header. It's too easy and cheap to get a SATA powered fan header PWM controlled hub to risk burning out my or my clients fan headers.

 

And by start-up amperage, it's not 100% fan speed when BIOS is training memory. It's from when the fan speed lags the input voltage/ signal due to inertia. That is what causes the spike in start-up amperage. All electrical motors do this. That is why your light can dim when starting a vacuum... not that the vacuum is running at 100%.

I'd also blame mobo makers or even AMD for blasting the CPU fan at 100% every time you boot or resume from sleep. One nice side effect of using a lot of headers is that you have better control over beating.

 

A12x25 is a weak fan to begin with.. so shouldn't be a problem.

 

And by start-up amperage, it's not 100% fan speed when BIOS is training memory. It's from when the fan speed lags the input voltage/ signal due to inertia. That is what causes the spike in start-up amperage. All electrical motors do this. That is why your light can dim when starting a vacuum... not that the vacuum is running at 100%.

I'd also blame mobo makers or even AMD for blasting the CPU fan at 100% every time you boot or resume from sleep. One nice side effect of using a lot of headers is that you have better control over beating.

Indeed!

Start-up spike / surge is why big buildings like shopping centers with many HVAC units (often roof) have them start-up in timed delay cycle. 10 units take a couple of minutes to all be up and running because all starting at once would seriously overload system, the power grid supplying complex.

A12x25 is a weak fan to begin with.. so shouldn't be a problem.

I assume a noisy litlte beast.

 

https://en.wikipedia.org/wiki/Power_factor

Basically when the voltage and amperage are out of phase due to changing load/ high load... ie when the fans are spinning up, not necessarily when they are at constant speed, although unless there is a correction... well, what ever, just read the wiki.

 

No idea what your quote is about. Okay, it's about voltage/amperage phase from changing load/high load.
You need to be more specific and tell us where in wiki article the part you want us to read.

Like this quote from Champion

"When using an appliance that has a motor, you will need to compare the running current (amps) with the starting load. For instance a 1 hp capacitor start electric motor typically runs at about 1500 watts, but has a starting load of about 4500 watts." ​

What is the difference between running watts and starting watts?

When using an appliance that has a motor, you will need to compare the running current (amps) with the starting load. For instance a 1 hp capacitor start electric motor typically runs at about 1500 w…

help.championpowerequipment.com

.

Their example clearly shows 3 times as much power to start capacitor motor. Capacitor start motors adds capacitor power to line power to supply startup power surge motors draw at startup. This means 4500w plus 4500w capacitor supplied to start 1hp capacitor start motor that typically runs at about 1500w. In this example motor could be pulling 9000w starting up.

Above example of power surge at startup in capacitor motor happens in our fan motors at startup. Similar increases apply to other motor load increases, like bad bearings (especially at startup) or something gets pushed into impeller stopping it from turning and stalling motor. Something I didn't mention before is header failure at startup often involve old fans with stiff bearings that make motor work much harder / longer to spin fan up to speed.

 

If the fans take awhile to rev up to speed, or they are not making it to full speed, the header is overloaded.

Easiest way to tell.

Also, just for clarity, when you blow a header, the header is usually fine, but the PWM controller is what you actually hear pop.

 

Not sure what you mean here, but I think you have some wires crossed up here, no pun. When PWM first came out I thought the same but that's not the case, at least not with the PWM fan headers I've used.

PWM signal from PWM fan header to fan (pin-4) is control signa to PWM pulse control in fan. 12v constant power (pin-2) is pulsed in fan by PWM fan PWM circuitry with signal from motherboard PWM fan header. There is no pulsed power from PWM fan header to fan. So how can it "pop" PWM in motherboard instead of 12v circuit?

If PWM fan header sent pulsed power to fan, than non-pwm 3-pin fans would not always run full speed on PWM headers.

I don't know if PWM circuitry is killed when motherboard PWM headers are overloaded or if it's just the 12v power, but I know for sure a 4-pin PWM fan headers does not pusle 12 power to fan. Maybe some 3-pin fan headers pulse power to 3-pin / 4-pin fans, but there would be no reason to make "PWM fans" if motherboard was pulsing 12v power to fan .. and 3-pin fans would be speed controlled instead of only running at full speed when plugged into PWM headers.

Seems I read that some newer motherboards have PWM pulsed power on 3-pin fan headers instead of variable voltage, but I could be wrong about this.

There are PWM variable speed control units on the market that do what you said. They usually have a knob to turn speed up or down. They send constant power to give motor full speed, and pulse it as different rates to lower motor speed. Before they came rheostats lowered voltage for slow motor rpm.