I am currently looking to build myself a new gaming rig, and have chosen the Corsair 620HX as the powersupply.
I did some googling on peoples experiences on this powersupply, and came across this page.
I am studying Electrical and Computer Engineering at UoA, and thought I might be able to shed some light on this high frequency noise some users appear to be having.
First off, huh? There isn't a degree in "Electrical and Computer Engineering" per say. That is the name of the department. You would have to be either taking a degree in "Electrical and Electronic Engineering" or "Computer Systems Engineering". You can't be doing both as its mutally exclusive. I am guessing you are year 2?
Some theory you have proposed only holds true to a certain extend and some is incorrect.
As an introduction, AC power is by no means sinusoidal at all. All AC devices that you connect to the grid are "seen" by your power generation company as either capacitive or inductive; Devices that are 100% resistive are very difficult to produce. This capacitive and or inductive effect of the devices on your AC supply modify the sine wave of your AC voltage, and add non-linear behaviour such as spiking on either the leading or trailing edges of the wave, clipping, or cupping on the extremes. This is due to the fact that capacitive loads cause the current to lead the voltage, and inductive loads cause the current to lag the voltage. A perfect AC sine wave would have the current exactly in phase with the voltage (A phase difference of 0 degrees). Any capacitive or inductive effects will cause the current to lag or lead in comparison to this ideal 0 degrees. Devices that are purely resistive do not effect the phase difference between the voltage and the current at all, and thus are the best kind of devices as they are more efficient.
Ok, I give you credit for stating the power factor correctly. HOWEVER, just because a load is capacitive or inductive DOES NOT cause the distortions as you have stated above. A pure capacitor itself or a pure inductor itself does indeed cause current to lag/lead the voltage respectively, yes but not distortions to the sine wave. The effects you have described however about the sine wave is true, but the cause is not. What causes those effects are due to A LOT of reasons, and would be too boring and/or tedious to list them all. It can also occur from various places. From the plant itself, the transmission line, distribution line, circuit breakers,compensators, industry plants, consumers, commercial places and so forth.
Power quality is of the utmost concern to power engineers due to 1.) Demanding increase of PQ due to sensitive equipment requiring them 2.) The sensitive equipment themselves is one of the main contributors to the power quality itself.
On the very basic concept of a Power Supply Unit (like the ones found inside a computer) it has 3 main functions. 1.) Step down the voltage 2.) Convert AC to DC. 2.) Regulate the DC voltage to "computer voltages" ala 12V, 5V.
PSUs has been around as long as electronics has been around. PSUs today serves the same function as their predecessor, however the method has changed. The method used to regulate the voltage has vastly improved on the sense of size reduction, power efficiency and wattage output. The method employed today has the same sense of PWM. The method is great on the microscale however on the macro it is one of the main causes of distortion being feed back to the mains. The PWM method "distorts" the waveform as part of the process.
Another point to note is the size of a transformer is inversely proportional to the frequency. A transformer is used to step down the mains voltage. If the transformer were to be run at 50Hz it would be as large if not larger than the whole PC itself! The mains may be running it at 50/60Hz, but the PSU has to be running at its own frequency. The higher the frequency the smaller the transformer has to be. The typical frequency is around and/or higher than the 10 kHz range. So why not run it at the highest frequency? Well you have to compensate, with all electronic products the smaller it is, the more costly it will be. So the electronic engineers has to compensate for the most cost effective. The frequency in a PSU is usually set by an RC clock on the on-board controller. A typical resistor has a tolerance of 10% (consumer grades) I'm pretty sure PSU companies would use industrial resistors with tigher tolerances. What that bascially means with tolerance taken into account the frequency would deviate from PSU to PSU. And this frequency combinded with the inductor toroids are the number 1 source of PSU buzzes. That's why some PSU buzzes, some don't, blame it on the tolarances!. A good PSU engineering company would take into account the extremes of the resistances use and made sure in the worse case scenerio the buzzing would not be in the audible range. However having said that, we do not live a a perfect world. There are spikes, swell, dips and what not in the power line and you would have to design low/mid/band pass filters to compensate for that. If the filter is poorly designed (or is at the extreme tolerances of the resistances to shift the frequency), buzzing could occur. Then you still have to design the overvoltage/overcurrent protection. Layout of the PCB to have the least EMI. Make sure it conforms to standards if you intend to sell it to that country yada yada yada. Can you start to see how complicated a PSU is and the myriad of places where the problem can occur?
Another cause for the buzz/hum could be the internal transformer. This is due to the magnetic-domains in the ferrite core aligning and re-aligning themselves in a high frequency due to the changing magnetic field. This is perfectly normal.
The two I have listed above( internal frequency change due to tolarances and transformer hum) are the most likely candidates that would
cause buzzing. There are many many other reasons why a PSU buzzes.
Also, the AC supply is always affected by devices. Power transmission is a two way street, any effects from devices not properly designed, such as lack of RF sheilding for example, can travel back into the AC line and wreak havoc on other devices in your home.
The active power factor correction in modern power supplies is a type of technology that accounts for these effects on your AC power sine wave created by the power supply, and attempts to correct itself so that it appears to be a resistive load (Power factor of 1) as opposed to a capacitive of inductive load.
The active PFC in power supplies depends greatly on a number of factors, such as:
- Power factor of incoming AC signal
- 'Quality' of the incoming AC signal
- Load on the power supply
Incorrect in the sense that PFC does not compensate for sine distortions you have described. PFC is used for a "cleaner" power. (AC) Power is dictated by P = VI cos (Phi). Cos (Phi) is known as the the power factor. Thus in order to get maximum power Cos (phi) has to be 1. Thus Phi the angle has to be 0. Thus its win win. From the utility view point it seems to be a resistive load, and on your end the psu is running at its full potential. The PFC however does not solve the problem distortion caused by regulation, but instead probably contributes to it.
The electrical noise that most of you hear from your computers will be in the higher end of the audible scale (18,000 Hz and above most likely). This could be generated from several sources, such external frequencies travelling back into your power supply (From things like the motherboards clock generators, un-ferrite blauned cables, bad case design, just to name a few). Also, the active PFC possibly could be a source of this noise, as it uses a series of capacitors and inductors to couple across the supply to bring it's power factor as close as possible to 1.
ok, did you just contradict yourself by saying that the PFC is used to correct those problems?
A note on cables: All cables act as giant antennas to RF frequencies! And also, nearly all devices emit RF frequencies! Cables that have a cylindrical "bump" in the cable have an RF ferrite blaun on the cable, to help reduce the effects of RF being picked up by the cable.
The power supply noise can come from several components within the psu. Capacitors, Inductors are generally the culprits, and if you want to do further back-reading on the magnetic fields produced by these devices, do a google. However, inductor noise is sometimes created by the windings of the inductors vibrating on a minute scale. This is due to the magnetic field produced by inductors, which in turn inacts a force on the inductor itself, and depending on the frequency of the signal through these inductors, the coils may vibrate at a frequency within human hearing range. As for capacitors, I am not 100% sure yet as to what effects are generated by them, but I do know that they can often create resonating frequencies within the human hearing range.
These problems of musical electrical components are pretty much unavoidable.
However, I have noticed that there are people whom are having high frequency electronic noise issues when running an extremely low load on their power supply. 40-60W? For these people, go out and buy a 200W power supply, buying a 500W+ power supply for a load that small is bloody rediculous, as power supplies are extremely inefficient when loaded at such a small load! As a rule of thumb, your rig should pull around 40-50% of the total load supplied by the power supply when your rig is idle.
True to a certain extend. Every PSU has an optimal loading condition. don't quote me on this IIRC but it is usually 80% of the rated wattage.
Also, there are several things that might help this high frequency noise, such as:
- Using sheilded monitor cables (Look for ones with the cylindrical "bump" along the cord)
- Using sheilded/ferrited power cables
- Changing the load on your power supply
- Try changing the refresh rate of your monitor, the clock frequency of your cpu etc
- Try turning various devices on and off within your homes. (Other devices may be messing up the AC signal and the power supply is affected by these effects)
Also, ideally, the AC frequency is 50Hz, however in practice this can be anywhere from 45-65Hz in the most extreme cases, which also can affect your power supplies, and create unwanted electronic noise as your power supply attempts to correct this.
Incorrect, a utility will not let the frequency deviate more than .5Hz, (althought some countries such as India has been know to have very large deviations). More than .5Hz sensitive equipments will cease to work properly.
As for UPS', not all models have AC wave filtering and noise reduction. Very expensive models do, but for the most part, UPS' do not have corporate-grade AC smoothing.
Just my input
I hope that wasn't too confusing for some of you, and apologies about the length of my post. If anybody has more information than I do, feel free to correct me!
And here I am
. Although you had some of your wires mixed up, it was a good effort