If you can't see the oscilloscope trace, try adjusting your KZread settings. The default refresh rate may mean you simply can't see the trace. For me, at 480p the trace is almost invisible, but changing the settings to 720p50 HD or 1080p50 HD makes it a lot more visible albeit with a lot more flicker. I suspect that's how it actually looks.

Note how the power factor of the LED when powered through the variac was better than it was when powered through the dimmer switch. That suggests that the LED almost certainly has a capacitive dropper, because the phase shift of theinductor (variac) is helping to cancel the phase shift of the capacitor. Large industrial users who are charged for apparent power as opposed to actual power because they use lots of power, mostly inductive loads (large motors to drive equipment) will install banks of capacitors to help cancel out the phase shift and bring their power factor closer to unity. This is also how those bogus "power saver" devices that Big Clive has debunked in previous videos sort of work. They basically contain a capacitor, on the assumption that most loads in a house are inductive, due to all the power supply transformers used in modern electronic devices. They perform negligible power factor correction- which is, of course, pointless for a residence, as residential customers are charged based on actual power used, not apparent power. That, of course, may change as resistive lighting becomes a thing of the past, but even still, the 30 or so uF capacitor in those bogus devices does very little to fix that problem.

Very interesting video, I learned a lot from this channel.

Cool. I always wondered how an LED bulb would react with a variac vs a standard dimmer, which chops part of the sine wave thereby reducing the average voltage. I realize variacs are very rarely if ever used anymore these days but many years ago they were used in theaters and opera houses for dimming large stage lighting since although physically large and heavy, much more efficient than the resistance dimmers used earlier, and safer than salt water dimmers (liquid rehostat) used in the early 20th century.

Class Video JW 👍

Thank you John .

Chuckle. When I was giving planetarium shows in the 80s, the main dome lights were controlled by large variacs, each one with about two dozen light bulbs around the rim of the dome. The knobs were around 6" diameter to make slowly reducing the light easier. The variacs themselves were about 12" x 12" x 12", maybe 2-3kW.

I’d be concerned the the PF Meter isn’t compatible with a Non-Linear Load, and therefore struggling to identify the phase shift :/ The TRIAC circuit will add some Reactance, but it be fairly small and constant so the Linear PF changes be fairly minimal. It will introduce Harmonic distortion sure, but it’s late and my brain is done..

@sdgelectronics

5 жыл бұрын

The harmonics are indeed reactive as they perform no useful work, but I agree - the PF displayed is not the true power factor. I will try the same thing with my home-made trailing edge dimmer on a Rohde & Schwarz power meter

Hi John I was wondering if you could discuss the difference between SELV and PELV devices and maybe give some examples of each. Thanks for your excellent videos. You have taught me loads.

Interesting video but wouldn't it be good to be able to see the current on the o-scope along with the voltage?

@PaulSteMarie

5 жыл бұрын

Current probes for o-scopes are rather pricey.

I think we're actually seeing a failure in the PF measurement on that device. The harmonics are reactive in nature, but the power factor on the display doesn't represent this correctly. I suspect the chip in the power meter is simply comparing zero crossing points rather than doing a full analysis of the AC waveforms.

@marct3928

5 жыл бұрын

I suspect you're right - would be interesting to know!

How is the meter measuring the power factor? Comparing the incoming mains voltage signal with the current that is being drawn and doing some sort of calculation on it? The way I see it, with a leading edge dimmer; the current is completely inline with the voltage (pf=1) until it gets suddenly cut off by the triac. It's just like switching the lamp on and off repetitively and very quickly, there is no reactive power where there's current lagging or leading voltage? Switching a linear load on and off should generate a PF=1?

@jwflame

5 жыл бұрын

Just switching a linear load is PF=1, but the dimmer isn't linear, it contains a fairly substantial inductor and some capacitors.

@sdgelectronics

5 жыл бұрын

I agree, the power meter will just be comparing the waveform zero crossing points giving a false reading

@sdgelectronics

5 жыл бұрын

I might try the same experiment on my channel with a Rohde & Schwarz power meter

@Mark1024MAK

5 жыл бұрын

A distorted sine wave also results in a poor power factor... as the supply system is designed to supply a current proportional to the voltage throughout the whole cycle. But with electronic loads, or electronically switched loads, the current drawn at different parts of the waveform are not proportional to the voltage throughout the whole cycle.

@koffibanan3099

5 жыл бұрын

Hmm. + switching off (or on) a load creates higher frequency harmonics, those will have an influence on the PF

The problem is that when the dimmer is used the current is only taken on the last part of the cycle. You could say it is still in phase but is in fact the power factor is not defined for a non-sinusoidal load current. As others posted it would be instructive to show the current trace on the scope as well. The fact that dimmers and lots of other loads take non-sinusoidal currents causes problems with the supply waveform as well as the harmonics heating the grid transformers. When I worked for London Underground the AC waveform was very poor as about 80% of the power was transformed to DC with 12 pole rectifiers. At times the AC voltage waveform had more than 2 zero crossings per cycle.

Enjoyed this Mr JW, nice series of how crap old technology is when compared to new crap!, But isn't really too bad. Thanks for sharing and best regards from the Black Country.

@extrastuff9463

5 жыл бұрын

Well the power factor isn't ideal but at least the total power consumption is a lot lower for an equivalent amount of light. But it sure is a hassle, I'll have to get around to replacing quite a lot of GU10 bulbs with dimmable LEDs and replace my wall mounted dimmer in the living room and bedroom soon. Either that or just give up on dimming entirely, last time I checked those aren't for sale here anymore in the old halogen type and I'm all out of spares. Should have swapped them quite a while ago really but during winter I was both lazy and didn't care too much about "waste heat" since my bedroom only has a sparingly used resistive electrical heater as it is anyway and the ones in the living room see limited use. I do wonder though if it might one day become useful enough to have an efficient active power factor correcting DC power supply in the house that runs lighting circuits assuming they are all LED it should work. Could be interesting for the various low power small chargers etc too, but that would be complicated with outlets, compatibility and possible new plug types needed.

@bostedtap8399

5 жыл бұрын

@@extrastuff9463 yes, probably 95% of my BC, ES single bulbs are LED, with three dimmable. All the battens are LED. The CFL's are long gone, never liked those. Very good thought on the PF on lighting circuit's, I had wondered about this myself, but not technically competent to try it. As you state, the low power PSU's in addition to the various LED lamps, clip the peak to peak sine considerably, plus real VAR possible costs. Active PF controller?????. Enjoyed your reply.

@extrastuff9463

5 жыл бұрын

@@bostedtap8399 What I meant with active power factor correction is that in some regions for power supplies above a certain threshold above which passive correction with some capacitors etc isn't enough. Not at the pc right now so I'll be lazy and link something covering the basics: www.electronicdesign.com/power/what-s-difference-between-passive-and-active-power-factor-correctors It would deal with the observed power factor by the utility company fairly well. Not sure how it'd end up affecting net efficiency inside the house though (probably not worth the investment until it becomes mandatory or when they start charging for horrible power factors).

Is it possible to dim non-dimmable bulbs with variac?_

@jwflame

5 жыл бұрын

Sometimes. However it won't work will all types and for those that it does the dimming won't usually be linear.

This unavoidable power factor thing makes me wonder if it might one day become sensible to add an extra plug type and wiring to outlets providing DC from a powersupply that is fairly efficient (compared to standard run of the mill compact ones in most devices these days) possibly with active power factor correction as well. Running 5.2V USB compliant voltage to outlets seems insane due to losses in the wiring but something higher like at least 30V (but ideally higher than that) could be a way to avoid non-unity power factor issues and keep some of the horrible noise exceptionally bad appliances can generate off the grid. An extra socket type for low power consumption switch mode/capacitative dropper powered things would be a hassle but I could see a use for that if they ever do decide to start charging consumers based on power factor as well. If the DC voltage is high enough I suspect most switch mode power supplies designed today wouldn't even care about the difference (typically they are already rated for 100-250V AC).

@sdgelectronics

5 жыл бұрын

We'll probably end up with more issues with harmonics and EMC from poorly controlled loads.

@extrastuff9463

5 жыл бұрын

@@sdgelectronics I guess it'd be impossible to rule that situation out. Especially when substandard cheap options are used. Just a gut feeling but assuming the power supply is well designed and the LEDs are reasonable for the mains distribution side it shouldn't be worse than the combined effect of current terrible LED drivers now? How bad would it be within the residence assuming you'd only run lighting circuits on it and the rest just stays as is with normal AC mains.

Are there any standards for power factor, or recommended ratings for power factor, or is that down to manufacturers?

@icanrememberthis

5 жыл бұрын

In dk, if you wish to connect to the public grid, you must ensure your total installation has a power factor between 0,9-1,0 on the inductive side. This is for residential (and wherever else EN60364 applies). The power factor rule isn’t from EN60364, but from a localised agreement, translating to something like ‘common regulation (for connecting to the grid).

@sumilidero

5 жыл бұрын

industrial facilities have to keep power factor above certain level, otherwise they have to pay additional charges for reactive power consumed. In my country its about 0,92~. In terms of equipment, i dont think there are any rules (a device must consume certain amount of reactive power to work) and the compensation for that is in most cases done on the central level (in factory substation for example). Ofc, we would like every device to have PF 1 but its impossible, some high class appliances like welders etc, have some capacitors inside to help with PF but its still not sufficient (as the power factor changes depending on load and other things).

@puckcat22679

5 жыл бұрын

In the US, and many other countries, residential customers don't really need to worry too much about power factor. Industrial users, with lots of powerful electric motors, which are inductive loads, are typically charged for apparent power as opposed to actual power, so they will usually have power factor correction, in the form of large banks of capacitors.

@puckcat22679

5 жыл бұрын

sumilidero the large capacitors in welders are actually there for voltage stabilization and helping to start the welding arc. Any power factor correction that comes out of them is incidental. Given that an electric welding machine is basically a giant transformer, followed by (on all but the absolute cheapest machines) a giant inductor, you'd need a lot of capacitance to cancel all that inductance. The 100,000 or so uF that good welders have for voltage stabilization isn't going to cut it.

@matthewday7565

5 жыл бұрын

The only thing (other than commercial gear) I can recall having a standard for, is the active PFC on modern PC power supplies - the input stage being a boost regulator that allows power to be drawn across the whole cycle, rather than just hitting the peaks hard like a FULL BRIDGE RECTIFIER with a chunky capacitor. The other bonus of the circuit being full range voltage input without the "go bang" switch

OMG I've got the exact same scope

Pity you did not show current signals. Idea for next video: a LED source with a dimmable PFC controlled driver.

Can you explain #PF for #noobs

@sdgelectronics

5 жыл бұрын

It's simply a measure of how effectively a load consumes electricity to produce work - i.e. the ratio of actual load power to the apparent load power.

@Mark1024MAK

5 жыл бұрын

It is a ratio between the voltage and the current being perfect sine waves and in sync with one another AND either the sine wave being distorted or the voltage and the current waveforms not being in sync, or both! A power factor of unity or 1 means the load is drawing current with a perfect sine wave and the voltage across the load is perfectly in step with / in sync with the current waveform. No unused current flows. Applied voltage and the current flow multiplied are equal to the power used by the load. Like a perfectly tuned petrol engine running under load at peak efficiency. A poor power factor (less than unity / less than 1) means the current flow to the load is greater than you would expect compared to the power used by the load. Or that the waveform is distorted. In the first case, that is more current flows. This difference is called apparent power. So applied voltage multiplied by the current flowing = actual power used plus the apparent power. This extra current / apparent power creates a magnetic field in transformers, motors, inductors or charges capacitors or similar. After the peak of the sine-wave, the collapsing magnetic field / discharging capacitor will cause some current to flow back into the supply. Hence the “apparent power” name, as this electricity did not do any ‘useful’ work. With a distorted waveform, current flows, then is cut or reduced part way through the cycle. The turbines at the power station can’t instantly respond to this. Same where little current is consumed and then the load current shoots rapidly up. Again, the turbines at the power station can’t instantly respond to this. So if these distortions constantly occur at each half cycle of the mains waveform, the supplier has to deliver the current that is demanded by the load when it does draw current, even though little or no current is drawn for the rest of the cycle. Like trying to run a four cylinder engine at full power with only two cylinders working properly.

The meter voltage was not changing so you are monitoring mains voltage not lamp voltage. You should have put the meter on the output of the Variac so effect of variac inductance is not measured. You should also invest in a better scope. The display is very bad with the sine wave dimmed for most of the trace.

@Mark1024MAK

5 жыл бұрын

The dimming of the display on the oscilloscope is due to the way the camera works. There is nothing wrong with the oscilloscope. All conventional oscilloscopes scan the beam across the display at a certain frequency. The camera captures the image at a certain frequency. The strange image with part of the trace missing is due to the interaction in the camera of these frequencies.

@peterdkay

5 жыл бұрын

@@Mark1024MAK Choosing a different sweep speed may fix problem. In this case the result was almost useless as 90% of trace was missing in video. A digital scope would be a better choice and can be purchased second hand on eBay for peanuts.

@johncoops6897

5 жыл бұрын

@@Mark1024MAK - Are you dumb or something? A 'scope where nobody can see anything is a TOTALLY USELESS scope. Hence Peter's comment that a BETTER 'scope be used for these videos. The one used in this video is totally useless (since there is nothing displayed), hence almost anything would be a BETTER scope, wouldn't it?

@Mark1024MAK

5 жыл бұрын

John Coops - There is no need to be rude. Of course it is easy to say now, after seeing the problem that either a storage oscilloscope or a digital oscilloscope may have been better. John may not have realised at the time that the picture that the camera was recording was not showing the whole waveform (what he could see on the oscilloscope directly with his eyes would have been the full trace of the waveform, you can tell this by the way he gives a detailed description of it). Without knowing details on the model of oscilloscope used and how it was configured, I can’t say if a different mode or scan speed could have been used. The trigger signal would have been 50Hz as the waveform is a 50Hz mains signal. I’m sure the video / camera guys can look at the recording and work out what frequency the video recording was using (maybe 60Hz). If the camera was recording at 60Hz, then it’s not surprising that a problem with the picture occurred. When recording a video of anything (that has a fast response time such as a LED) running at 50Hz you get problems (with a LED running at 50Hz, the LED appears to flicker/flash).

@marct3928

5 жыл бұрын

It may be your KZread settings. Watching at 480p the trace is almost invisible, but crank the settings up to 720p50 or 1080p50 and the trace becomes visible.