Psu Questions..

[dirtymike]

Current Set up
Msi B550 Gaming Plus motherboard
Crucial Mx500 1 tb
500gb SSD
Gigabyte Rtx 3060 Eagle
R5 5900x aio cooled cooler master ml240
Corsair RMx 750x
Just wanted to share my readings from HWINFO checking system for stability.

 

[dgianstefani]

You have yet to ask a question.

 

[HD64G]

All good there.

 

[Bill_Bright]

According to the ATX Form Factor standard, PSUs must maintain voltage tolerances within ±5% of required specifications.

Acceptable tolerance maximums:

12VDC ±5% = 11.4 to 12.6VDC
5VDC ±5% = 4.75 to 5.25VDC
3.3VDC ±5% = 3.14 to 3.47VDC
​

 

[dirtymike]

I appreciate the both of you, thanks.

Any SUS values here by chance?

 

[Bill_Bright]

None that I see.

 

[DoLlyBirD]

According to the ATX Form Factor standard, PSUs must maintain voltage tolerances within ±5% of required specifications.

Acceptable tolerance maximums:

​

12VDC ±5% = 11.4 to 12.6VDC​

5VDC ±5% = 4.75 to 5.25VDC​

3.3VDC ±5% = 3.14 to 3.47VDC​

​

Add to that, generally software readings when it comes to PSU rails can be completely wrong and if you have an issue that might be PSU related the only reliable way to measure those rails is with a HW multimeter, I've seen some crazy values in software that were completely false, not sure why they still present these values in SW tbh

 

[Bill_Bright]

Actually, even a multimeter is not 100% conclusive for several reasons.

(1) The load presented on a power supply by the computer components is constantly varying from a minimum amount at idle, to 100% load when the computer is heavily tasked. Conclusive testing requires measuring those voltages over a full range of expected loads to ensure the outputs are being properly regulated. This is a very difficult challenge, even for pros. Note this wide load range is exactly why the 80 PLUS certification program rates supplies at 20, 50 and 100% loads (plus 10% for Titanium certification).

(2) Most multimeters are incapable of measuring ripple and other anomalies that affect computer stability. This is done by a qualified technician using an oscilloscope or a dedicated power supply “analyzer” - sophisticated (and expensive!) electronic test equipment requiring special training to operate, and a basic knowledge of electronics theory to understand the results.

Therefore, conclusively testing a power supply is done in properly equipped electronics repair facilities. Many smaller shops often don't have the resources to buy the necessary test equipment. This is why swapping in a known good power supply, a tried-and-true troubleshooting method used by amateurs and professionals alike, is typically the best and easiest alternative.

No software based HW monitor program can measure ripple either.

I've seen some crazy values in software that were completely false

Me too. In fact, the popular HW info program, Speccy from the makers of CCleaner used to show the values for the +12V, +5V, and +3.3VDC voltages but so often were way off. For example, for the board in the computer I am using now, it used to report my +12VDC was measuring about 6.7V and my +5V was showing around 2V. If those were true, the computer would not be running. This happened so often that they removed the feature from Speecy.

 

[eidairaman1]

Current Set up
Msi B550 Gaming Plus motherboard
Crucial Mx500 1 tb
500gb SSD
Gigabyte Rtx 3060 Eagle
R5 5900x aio cooled cooler master ml240
Corsair RMx 750x
Just wanted to share my readings from HWINFO checking system for stability.

If the 12V, 5V, 3.3V are under or over 5% (4.99%- or 5.01%+) you should be concerned at that point.

So just do the 3 major values and add or subtract 5% to see what the absolute lowest or highest the volts on the psu should be.

 

[80251]

If you're only getting 11.4V on the +12V line how much does that affect your fans? How much slower do they run at 11.4V as opposed to 12V?

 

[P4-630]

If you're only getting 11.4V on the +12V line how much does that affect your fans? How much slower do they run at 11.4V as opposed to 12V?

Everything has tolerances, also fans and their rpm's.

 

[80251]

@P4-630

You're right of course, most fan manufacturers specify a range of acceptable input voltages. SanAce actually tests at the low end and high end of their input voltage range which create a whole different set of P-Q curves (e.g. the San Ace 140L). I'd rather have a better P-Q curve than a worse one.

 

[cvaldes]

If you're only getting 11.4V on the +12V line how much does that affect your fans? How much slower do they run at 11.4V as opposed to 12V?

The best person to answer this is you.

Boot into UEFI/BIOS and head over to the fan configuration section.

Pick a fan header to test on (CPU, CHASSIS, SYS1, whatever). Switch from PWM control to DC control. Set to 12V and note fan RPM. Then switch to 11.4V (or closest value) and then note new fan RPM. Depending on your motherboard and fan circuitry, there should be a slight difference although some systems may not register such a slight change.

You can switch to 6V and 9V (50% and 75% voltage respectively) and note a change. It probably won't be exactly proportional because DC fan control isn't as a precise as PWM control. It should be fairly close. The most important thing is that you will learn how much such chances affect YOUR system.

When you are done playing with this you can discard all changes as you exit UEFI and reboot your system on the previous settings.

Best of luck.

 

[80251]

The only fans that I have that would probably be affected by less than 12V are ones that you can't connect to motherboard headers (i.e. they draw far too much current [especially on start-up]).

I rather get the +12V I paid for than something less than that.

 

[cvaldes]

The only fans that I have that would probably be affected by less than 12V are ones that you can't connect to motherboard headers (i.e. they draw far too much current [especially on start-up]).

I rather get the +12V I paid for than something less than that.

Clearly you have a different philosophy on the operation of fans in your PC case.

I hate noise.

My goal with the fans in my various builds is to reduce noise as much as possible, mostly by reducing fan speed without greatly sacrificing cooling performance. I don't ever run any PC fan at 100% speed because of the noise. Fans at 100% speed also use more electricity.

Remember that the greenest electricity is the electricity you don't use. At least in my neck of the woods electricity isn't free (or notably cheap).

So keeping fan revolutions low helps my ears, the planet, and my wallet.

3-pin fan connectors control fan speed by adjusting voltage. If you want to run your fans at 12V, why do you care what voltage regulation does?

Noctua markets a Low-Noise Adapter pigtail. This is simply an inline resistor on pin 2 (+12V DC) that knocks off ~15% of voltage and corresponds to a roughly similar decrease in fan rotational speed.

If you want powerful and loud 12V PC fans, check out Noctua's industrial-PPC 3000 models. They provide at a lot more at +12V than the Noctua models I've installed in my various builds.

 

[eidairaman1]

The only fans that I have that would probably be affected by less than 12V are ones that you can't connect to motherboard headers (i.e. they draw far too much current [especially on start-up]).

I rather get the +12V I paid for than something less than that.

Atx spec allows for tolerance and sometimes mobo sensors are trash, best to check volts on the psu using a DMM

 

[Bill_Bright]

If you're only getting 11.4V on the +12V line how much does that affect your fans? How much slower do they run at 11.4V as opposed to 12V?

Of course all fans are different but essentially, there would be very little, insignificant, negligible difference. For sure, if the RPM and CFM differences between 12.0 and 11.4V impacted cooling (and thus system stability), your cooling was already woefully inadequate!

It should be noted that most computer fans are designed to operate with much lower voltages. In fact, that is exactly how their speeds are commonly adjusted/controlled. My Fractal Design case, for example, has a built in 3-speed case fan speed controller. It is even labeled as 12V, 7V, and 5V. I keep mine set at 7VDC. This ensures adequate cooling in absolute (to my ears, anyway) silence. A good thing because, like cvaldes, I hate fan noise.

 

[cvaldes]

If you're only getting 11.4V on the +12V line how much does that affect your fans? How much slower do they run at 11.4V as opposed to 12V?

For giggles, I played with this for a few minutes on my daily driver PC.

Indeed my mainstream MSI B550 motherboard does accept the 11.4V input and the fan speed decreased about 4%. As previously discussed DC voltage control of fan motors isn't precise but it's reasonably close.

Based on my previous experience using DC voltage control on fan headers, I was unsurprised by this quick test.

Naturally my results are specific for my hardware combination (motherboard and specific fans) and your mileage may vary if you use different gear.

Once again, you are the best person to answer your question since motherboard performance and fan manufacturing tolerances vary from unit to unit.

Fans especially as mechanical devices will show a noticeable amount of performance variance. I see this using the exactly same model of fans on different fan headers.

I would not be surprised if the results were slightly different on my other builds but I'm not going to spend any time on that.

 

[80251]

The Noctua i-PPC fans are weak sauce in comparison to the industrial 127x38mm and 127x50mm fans I use, which both have a maximum air flow of between 179.75CFM to 200 CFM and draw 2.7A+. They're only on when I need to cool my videocard.

 

[cvaldes]

Sounds noisy.

On my gaming build (see System Specs) the Noctua fans on the CPU radiator max out at 1025 rpm. The fans on the GPU radiator max out at 975 rpm and due to their placement, they are especially difficult to hear.

For me having to use noisy high-speed fans would indicate a flawed build design from a holistic perspective (acoustics, power/environment, money).

I don't need to use industrial grade fans because the entire system component choices add up to excellent cooling capacity at very low sound levels.

I actually use one of the Noctua 140mm industrial-PPC 2000 PWM fans on my NXZT H1 but again the fan curves are adjusted so it maxes out at 1300 rpm which is audible but not overbearing. Thankfully this is not a build where I max out the CPU in normal use. I've tried that fan at 100% power and the acoustics are unpleasantly prominent.

 

[caroline!]

For me having to use noisy high-speed fans would indicate a flawed build design from a holistic perspective (acoustics, power/environment, money).

*hides 38mm Deltas and jap Sunons* y-yeah, heh
Most of these "non-consumer grade" fans work fine in the 5-14V range. I went for these because I already had them but also because I didn't want my new PC to overheat, I've never liked auto fan speed or curves except for the CPU so I hooked up the rest to a bunch of 3P switches to manually control the speed or turn them off.

The Noctua i-PPC fans are weak sauce in comparison to the industrial 127x38mm and 127x50mm fans I use, which both have a maximum air flow of between 179.75CFM to 200 CFM and draw 2.7A+. They're only on when I need to cool my videocard.

Let me guess, a fellow AMD user

Noctua stuff isn't actually industrial they just call their fans that way so users think they're better than the standard ones, but hey they have to milk those juicy gamers' wallets somehow.

 

[eidairaman1]

*hides 38mm Deltas and jap Sunons* y-yeah, heh
Most of these "non-consumer grade" fans work fine in the 5-14V range. I went for these because I already had them but also because I didn't want my new PC to overheat, I've never liked auto fan speed or curves except for the CPU so I hooked up the rest to a bunch of 3P switches to manually control the speed or turn them off.

Let me guess, a fellow AMD user

Noctua stuff isn't actually industrial they just call their fans that way so users think they're better than the standard ones, but hey they have to milk those juicy gamers' wallets somehow.

Noctua is overrated

 

[80251]

I like the noctua iPPC-3000 NF-A14 (but I'd rather have a SanAce or Delta 140x38mm or 140x51mm fan) and I also like they're new NF-A20 which seems to have finally broken away from the pack of low performance 200mm fans.

The delta 127x38mm won't run on less than 8.5V and the nidec 127x50mm won't run on less than 7V, but I'm not sure if either fan will start at such low voltages and if they do the current draw must be pretty high, since I need the airflow though it doesn't really matter and that's why I want >= 12V on the +12V line (all my other PSU's even going back to my enermax 650 Watt model) provided > 12V on the +12V line, why should I settle for less?

 

[cvaldes]

Noctua is overrated

I’m okay with that.

PC fans are commodity parts. I already have a selection of fans from the various cases and AIOs I’ve purchased in recent years including some parts like Akasa and Sunon.

Some of the bundled fans I use (e.g. Arctic), others I dump into a box that I shove into a closet.

Fans are so inexpensive relative to other PC components it makes zero sense from a time perspective to buy 10-15 fans, run them through a bunch of time consuming tests to save five bucks.

And the amortization period would be what? Seven, ten years?

Now if I were a system builder selling hundreds of assembled PCs I might have a different perspective but I’m just a consumer with a few extra bucks in my hobby budget.

Same with thermal grease. I’m not going to buy twenty different tubes to end up with something that might be 1 °C better.

And this isn’t just for PCs.

If I owned a bakery I’d keep a very keen eye on the cost of ingredients or supplies like parchment paper. But I’m just a hobbyist baker so if I use farmers market eggs and imported French butter for a cake I’m serving to friends and family, I’m okay with the additional expense.

 

[Bill_Bright]

Noctua is overrated

LOL

I do NOT disagree. I just wonder what popular brand is not overrated these days? The fan boys overhype with exaggerated claims of over-excellence while at the same time, bash with equal vigor and misleading claims about the competition.