1
\$\begingroup\$

[California, USA, regular 110 V power.]

I'm reading the power factor of my "Late 2013" 27-inch iMac using a P3 P4400 Kill-A-Watt. It reads 0.95 when operating normally (60 to 120 W), but when sleeping, even with Power Nap off and power bouncing between 0 & 2 W (to power the USB mouse & keyboard to allow user wakeup) the PF bounces around 0.50. The K-A-W has a low update rate, so it's hard to read with the computer in this state.

If I "Prevent computer from sleeping when display is off", with screen off it draws about 22 W and the PF sits at 0.50.

In both of these cases (just screen off or computer sleeping), an Ikea LED bulb and my 15+ year old Sony ICF-CD831 CD clock radio's LED display flicker. Usually a few seconds between flicks, sometimes at about 3 Hz. The Ikea bulb runs very hot in a non-enclosed fixture. It might be failing and I'll throw it out when I'm done using it as a diagnostic. A GE bulb on the same circuit does not flicker.

As soon as I wake up the computer, the flickering stops.

My house has a 40 year old 100 Ampere Zinsco panel that the landlord's going to replace. There are only 2 20 Amp breakers for all outlets in a 3 bedroom house and some 3 prong outlets don't have the ground connected.

How can poor power factor on the computer cause devices with poor regulation to flicker? Is oxide on circuit breaker contacts a possible contributing factor?

I checked the line voltage with a DMM & saw it dipping from ~120 to under 110 with a 10 W LED bulb switched on giving an ESR of about 165 ohms, but an 1800 W hair dryer caused the same sag, so there doesn't seem to be a simple non-reactive loose wire.

Diagram added: enter image description here Load A is iMac causing interference, B is clock radio or 10 W LED light fixture being sensitive to that interference. X and Y are the 100 V loads around the house (refrigerator, lights), Z is lumped 220 V loads (stove, clothes dryer). The only way to eliminate X, Y and Z would be to get a 200+ W inverter and battery and run A & B from that in isolation of other loads in my house & neighbors on the other side of the pole transformer & I only have a little 50 W.

\$\endgroup\$
  • 3
    \$\begingroup\$ You should make a drawing of your setup, there is simply too much text to visualize your situation quickly. Also you bring in the low power factor (do you know what a low power factor means?) into the problem while there is no reason to assume the PF is an issue. Same about the "oxide in circuit breaker", why would that be an issue? You over complicate your question with random rogue conclusions that have no basis. \$\endgroup\$ – Bimpelrekkie Dec 14 '18 at 21:44
  • \$\begingroup\$ Laptop psu is almost certainly running in a "burst " mode, drawing a short pulse of energy occasionally. This is for less than a mains cycle so waveform is arbitrary non sinusoidal with random shape so many harmonics. Adding a load - maybe via a USB port (fan / light / coffee warmer (higher power) ...) if they stay alive, may allow the unit to waste a few watts and keep the psu in non burst mode. \$\endgroup\$ – Russell McMahon Dec 15 '18 at 10:27
  • \$\begingroup\$ XY problem! What is the problem you are experiencing? PF isn’t one. \$\endgroup\$ – winny Dec 15 '18 at 15:16
  • \$\begingroup\$ Right: the problem is flickering lights. This only occurs when the computer is sleeping or screen-off. That the power factor falls from 0.99 to ~0.50 when it's sleeping is simply a sign that might suggest how the computer's load is causing the flickering. I should have chosen a better title/question. \$\endgroup\$ – Morgan Dec 16 '18 at 1:53
2
\$\begingroup\$

Low power factor is often an indication of harmonic distortion of the load current. Every type of electronic device connected to AC power has the potential for harmonic distortion. Electronic loads may or may not have built in harmonic mitigation. The percentage distortion is like to me greatest with light load conditions, but the total distortion will be low unless all connected loads are generating harmonics.

Harmonic distortion and other types of electromagnetic interference generated by electronic loads could be interfering with some other electronic devices.

It is difficult to know if you should replace the harmonic generator or the items being interfered with. A third option would be to just put up with the annoyance.

\$\endgroup\$
  • \$\begingroup\$ It seems that unless the iMac is faulty, there must be an exacerbating factor or we'd hear about more sleeping iMacs causing devices to flicker. I will check the PF on some other Macs. \$\endgroup\$ – Morgan Dec 14 '18 at 23:12
  • \$\begingroup\$ Checking power factor with a K-A-W is a rather limited diagnostic procedure. An oscilloscope would be better. Also, the distribution system helps determine the harmonic level due to any given load. It is difficult to determine how much harmonic or EMI disturbance indicated a "faulty" product. \$\endgroup\$ – Charles Cowie Dec 14 '18 at 23:36
  • \$\begingroup\$ Agree, the switching power supply inside any PC may do some "pollution"to the grid. Some switching power supply schema will generate more interference at light load. That's why PC supplies have EMI and PFC stage at the input. \$\endgroup\$ – diverger Dec 15 '18 at 0:48
0
\$\begingroup\$

Likely problem: The computer psu is almost certainly running in a "burst" mode, drawing a short pulse of energy occasionally. This is for less than a mains cycle and/or not synchronised to the mains waveform, so the waveform is arbitrary & non sinusoidal with random shape, so many harmonics are produced and power factor is poor.

The filtering on the other devices is not adequate to remove the harmonics.

Possible solution: Adding a load - maybe via a USB port (fan / light / coffee warmer (higher power) ...). If the USB ports are powered up in this mode then you can probably tailor the load to keep the psu in non burst mode - and waste a few watts in the process.

\$\endgroup\$
0
\$\begingroup\$

How can poor power factor on the computer cause devices with poor regulation to flicker? Is oxide on circuit breaker contacts a possible contributing factor?

Not directly.

Power factor is the ratio between real power and apparent power. When PF is 1, real power and apparent power are the same, which means current is proportional to voltage all along the AC cycle, as would be the case if the load was a simple resistor. This is the ideal load.

PF lower than 1 occurs for example when the load is reactive (capacitor, inductor, motor...) and there is a phase shift between voltage and current, or with switching power supplies.

Your iMac has Power Factor Correction (PFC) which allows it to behave more like a high-PF load. However, it may be a passive PFC circuit which will only work well at high load, or an active PFC circuit which uses some power to work, and this means it is more efficient to turn it off when the Mac draws very little power. This is why the PF you measure varies according to the power it uses. This is normal.

This isn't very important anyway: the amount of harmonics injected on the mains by the device depends on its power factor, but it also depends on the power it uses. When the computer is asleep, it draws very little power. A high proportion of the current it draws will be ugly high harmonics, but the total current is tiny, so it doesn't cause trouble.

Now I agree with Russell that the iMac may draw power in bursts which make your lamps flicker. The interesting question is: does the Mac draw large occasional current bursts which can legitimately be expected to disturb other appliances, or is it a problem with excessive resistance in the wiring in your home turning these current bursts into excessive voltage drops?

Try plugging an incandescent light in the same outlet as the Mac. Does it flicker at the same 3Hz too?

Try plugging the iMac and the incandescent light in another outlet, maybe in another room, preferably on another breaker.

The Ikea bulb runs very hot in a non-enclosed fixture. It might be failing and I'll throw it out when I'm done using it as a diagnostic. A GE bulb on the same circuit does not flicker.

Its internal smoothing capacitor is likely cooked.

I checked the line voltage with a DMM & saw it dipping from ~120 to under 110 with a 10 W LED bulb switched on giving an ESR of about 165 ohms, but an 1800 W hair dryer caused the same sag, so there doesn't seem to be a simple non-reactive loose wire.

Out of curiosity I did the same test with lamp having two 8W LED bulbs and could not see any drop of voltage, with a multimeter resolution of 0.1V.

I suspect there is a problem somewhere in your wiring.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.