Active PFC; responsbile for high-pitched noise?

[Cattykit]

http://www.directcanada.com/products...anufacture=BFG

2 reasons I bought that PSU was because it has 140mm fan and efficiency rate of 85%. I thought 140mm fan would be quieter than 120mm one but I was wrong. The fan itself was a lot louder than the 120mm Cosair 550VX one.
I ended up buying a quiet 140mm fan replacing the original one. While the fan noise is gone, I notice this nasty high-pitched noise. Far worse than the fan noise this is. It drives my ears nuts and probably my cats too.

So, I'm in the market for a different PSU that'd fulfill my desire. I came across this PSU:
http://www.directcanada.com/products/?sku=13220AC1145

650W 80+ silver rating and it's only $64.15 (after $20 that is but still!). I'm thinking about buying this one but I have two questions beforehand.

1. I read Active PFC is reponsible for high-pitched noise. Is that true? I doubt it because Corsair 550VX does not make any high-pitched noise at all.

2. Should I buy 550W version instead?

I have i7 860 @ 3.7ghz. However, I have all the energy saving features turned on and I use offset Vcore. So, 95% of the time, my computer runs @ 1680ghz with 1.056 vCore. Also, I use nVidia GT 240 which uses only 10W in idle.

Thus, I believe my system eats up very little power for I mostly use it for surfing internet. Now, I know PSU is most efficient when in 50% load. I wonder if using 550W PSU on my system would be closer to that 50% load case.

 

[Emulex]

only if you use a crap battery back up. active pfc requires a true sine wave input to avoid those nasty things.

 

[HOOfan 1]

Several things...that listing is absolutely wrong.....the GS-550 does not have Active PFC. Either they screwed up and said it has active PFC...or that they are marketing the LS-550 as a GS-550.

If you got a PSU with the red voltage switch on it...then it does not have active PFC.

You didn't actually replace a Corsair VX550 with that BFG GS-550 did you? There is absolutely no comparison between those units...the Corsair is far superior. BFG is rated at 25C and peak output....Corsair is rated at 50C and continuous output. The Corsair truely does have active PFC.

The high pitched noise is probably coil whine...it could be from the PFC coils...it could be from other coils.

APFC units do not absolutely require true-sine wave UPS...it is fine to run them on a stepped wave for a very short period...although that might make it produce a noise.
A non-sine wave UPS will only be producing that non-sine wave power when the AC mains cuts out...otherwise all it is doing is passing through and possibly conditioning the AC mains power....which is true sine wave.

 

[johnnydrinkawater]

Poorly wound transformers have a tendency to whistle. (built a hand-wound boost converter)

Power factor correction engages extra capacitive load to bring your power factor within reasonable value to 1. Needless to say, capacitive loads aren't wound wire components.

 

[HOOfan 1]

Poorly wound transformers have a tendency to whistle. (built a hand-wound boost converter)

Power factor correction engages extra capacitive load to bring your power factor within reasonable value to 1. Needless to say, capacitive loads aren't wound wire components.

PFC uses toroidal coil chokes....

 

[johnnydrinkawater]

PFC uses toroidal coil chokes....

I see, sorry didn't know. When I think capacitive, I think electrolytic or ceramic...

 

[HOOfan 1]

I see, sorry didn't know. When I think capacitive, I think electrolytic or ceramic...

I'm no engineer...but from reading about PSUs for a while I know that passive PFC uses an inductor...I suppose a toroidal coil.

 

[johnnydrinkawater]

I'm no engineer...but from reading about PSUs for a while I know that passive PFC uses an inductor...I suppose a toroidal coil.

I don't know much about PFC in power electronics, excuse me if I'm sounding ignorant.

 

[johnnydrinkawater]

But in regard to the OP, I would suspect a transformer more so than a toroidal inductor, especially if it is sleeved in rubber.

 

[HOOfan 1]

I don't know much about PFC in power electronics, excuse me if I'm sounding ignorant.

no you don't sound ignorant. Capacitors are used in PFC as well.

 

[Emulex]

show me the documents from APC,HP, DELL, (any other ups manufacturer) that lists the supported battery backups for active PFC power supplies.

It is well documented.

 

[HOOfan 1]

show me the documents from APC,HP, DELL, (any other ups manufacturer) that lists the supported battery backups for active PFC power supplies.

It is well documented.

Of course they also want you to buy a more expensive sine wave UPS.
It is also a CYA thing. With 220V power it might not be a good idea to use a stepped wave...with 120V power, it is perfectly fine.
As I already stated...you only get non-sine wave AC from the UPS when the mains power has been cut...so you are either shutting your computer down or the UPS battery will be drained in less than an hour anyway.
Every firearms manufacturer also states that their guns are not made for hand loads and that you should always use factory loads....yet if you follow a load guide then you can fire hand loads in most weapons.

Go ask people like Jonny Guru or Paul Johnson...they will tell you it is fine.
A non-sine wave UPS will work fine with an APFC unit at 120V

http://www.jonnyguru.com/forums/showthread.php?t=3964

 

[johnnydrinkawater]

no you don't sound ignorant. Capacitors are used in PFC as well.

Well, I deal more with utility scale power engineering. I find this subject very interesting and would love to learn more. I see that in power electronics, PFC is more a matter of killing extra harmonics than strictly analyzing the system as a matter of L vs. C impedances.

Cool stuff

 

[Mark R]

Well, I deal more with utility scale power engineering. I find this subject very interesting and would love to learn more. I see that in power electronics, PFC is more a matter of killing extra harmonics than strictly analyzing the system as a matter of L vs. C impedances.

That's exactly it. The problem with power electronics is the very distorted current waveform caused by charging capacitors from rectified AC. This isn't such a problem with heavy duty electronics, running from 3 phase supplies, as the 3 phases result in cancellation of the most troublesome harmonics. Still, for heavy industrial electronics (e.g. heavy duty motor control inverters), the preferred option may be to use phase-shift transformers to create a 6, 9 or even 12 phase supply, as you get much better control of the harmonics with more phases.

However, most electronics are single phase powered, and harmonic distortion is widely recognised as a serious problem. In many areas of the world (notably not the US), it is now a legal requirement that 'high impact' electronic devices (i.e. popular ones which take a relevant amount of power - e.g. PCs, LCD or plasma TVs, etc.) must strictly control their current waveform.

Passive PFC was common with weaker PSUs - up to about 300W. This consists of a network of capacitors and inductors, and would essentially act as a low-pass current filter, that would present a very high impedence at the harmonic frequencies, thereby reducing the harmonic currents. However, the demand for higher power levels, required a new approach, as the inductor and capacitor networks rapidly grew unmanageable in size, cost and power losses.

The modern technique is to use an active PFC system. An active PFC circuit consists of a step-up DC-DC converter controlled by a feedback loop. This step-up converter sits between the rectifiers and the storage capacitors. However, the feedback circuit doesn't simply regulate the voltage on the capacitor - in addition, it regulates the input current to the step-up converter, so that the instantaneous input current is proportional to the instantaneous line voltage. The result is a compact, efficient circuit with exceptionally good harmonic reduction. There is a side benefit too, which is because the PFC circuit is a voltage regulator, the circuit can trivially be made 'universal voltage' and because the regulator is a step-up regulator, the higher DC voltage allows the use of smaller reservoir capacitors, and improves efficiency of the main power transformer due to lower primary currents.

The problem with poor quality UPSs comes from the very stepped waveform that they produce. This causes problems with active PFC circuits as they attempt to regulate the DC bus voltage and current - when there may be (relatively) prolonged periods of zero voltage. This can result in very abnormal levels of ripple in the reservoir capacitors and correspondingly a severely distorted current waveform when the line voltage changes in a stepwise fashion. This can lead to all sorts of problems, including uncontrolled oscillations in the PFC circuit, or the UPS inverter, as well as poor voltage regulation. The use of 'modified sine-wave' inverters to power active PFC circuits is not generally recommended for this reason. A similar phenomenon can be seen when powering active PFC circuits from a ferroresonant transformer, and the use of these 'voltage regulating' transformers is not recommended with modern active PFC PSUs.

 

[HOOfan 1]

So you are saying the extended 0V state of a stepped wave is the problem...while Silverstone's engineer in the thread I linked above stated that the problem was that the peak voltage of the stepped waves was higher than sine waves and thus the voltage may exceed the capacitors' capabilities....especially with a 220V input.

APC's explanation in that thread was total BS though, talking about changeover time.

This seems to be one of those completely gray areas where not even engineers can 100% agree.

I'll just leave it at...I'm not an engineer.

Of course the truth remains that the PSU in the original post is not an Active PFC unit, so the whine he is hearing from his unit has nothing to do with APFC and a square wave UPS

 

[Cattykit]

Whoa, a lot of terms I don't understand!

Anyway, thanks for all the inputs guys. I ended up buying 650W version. It runs great! OCCT test revealed, compared to the BFG one, vCore fluctuation is a lot less as well as other voltage fluctuation.

 

[HOOfan 1]

OCCT is useless for evaluating a PSU. Only a hardware multimeter can be trusted to measure voltage.

 

[faxon]

OCCT is useless for evaluating a PSU. Only a hardware multimeter can be trusted to measure voltage.

actually for testing a PSU through and through you really need a programmable load meter hooked up to an oscilloscope, so you can measure current ripple and voltage regulation ant different load scenarios. that's how all the cool people do it don'cha know! :awe: