dimmer harmonics

[jpdavis]

hi :)

 

Can someone tell me what dimmer harmonics are? Ive done a search but can't find the answer and my electronics teacher cant explain them.

 

Thanks

[Ike]

Any type of harmonic is simply a wave with a frequency of a whole number times the original wave. For example when we're talking about harmonics in power distribution, dimmers and the like we take the main frequency or as we call it the fundamental or first harmonic as being 50Hz the second harmonic is two times the fundamental or 100Hz, the third is three times or 150Hz, the forth is four times or 200Hz and so on and so on.

 

Now why are they important? Well in a ideal AC system the voltage and current would be a perfect sine wave, this is nice and easy to describe and do maths with. In reality this is hardly ever the case and things like dimmers and electronic goods are particularly good at messing up the waveform. Luckily a French bloke called Jean Baptiste Joseph Fourier came up with the idea that you could describe any wave as a fundamental and a series of harmonics, this gives us a way of describing the distorted waves mathematically so we can do calculations with them and make predictions.

 

To try and make it a bit clearer I've created the following example, it's not totally accurate but it's near enough what the current drawn by a mobile phone charger or other small appliance might be. I haven't used a dimmer as the waveform changes depending on how high you have the faders and the design of the dimmer used but the theory is the same.

 

http://i29.tinypic.com/29lfm6v.jpg

 

As you can see there are big spikes where lots of current is drawn then quite horizontal wobbly bits where very little is drawn. Mathematically this waveform would be described as a fundamental of 50Hz with 60% 3rd harmonic, 40% 5th harmonic, 20% 7th harmonic, 10% 9th harmonic, 5% 11th harmonic, 2.5% 13th harmonic and 20% 21st harmonic. The mathematical way is obviously much more accurate as it describes the wave exactly and you don't need to know what I would class as 'quite horizontal', 'wobbly' or 'very little current'. The wave above has a lot of harmonic content and was purposefully chosen as it's an extreme case, even if something looks like a perfect sine wave on an oscilloscope it probably has a bit of harmonic content in there.

 

So why do we care? Well if you're familiar with three phase you'll know that the neutral currents from each phase cancel each other out to a certain extent which means we only need one neutral conductor for the three phases, waveforms with high triplin harmonic (thats those that are multiples of three - 3rd, 6th, 9th etc) content don't so we sometimes need to think about beefing up the neutral conductor to cope with this which can be expensive in terms of extra cabling. Transformers generally work best with pure sine waves and we therefore sometimes have to down-rate these if there is a lot of distortion in order that they don't overheat. Cables and distribution equipment are all subject to the peak currents of the waveforms but only the average actually does any work therefore the more peaky the waveform the less efficient the system or the lower the power factor. Heavy distortion can upset sensitive equipment and cause problems with audio systems. It can upset electricity suppliers who can demand you remove the cause or charge you extra to cover their costs in dealing with the problem.

 

Hope that doesn't just confuse you more...

[rooftop]

oh go on, write the formula for the fourier transform to the 21st harmonic - I dare you :P

 

(brings back memories)

[Dmills]

Here is a thread I started a few years back when we were planning the power distribution for our new dimmer room, that discusses the issue at some length and that also discusses moving light harmonics (which is a related problem).

 

Regards, Dan.

[KevinE]

I occasionally do electronic work for the local water authority where motors supplied via inverter drives (off rectifiers)around the 500kW range (each) can cause havoc to the supply harmonics as you can imagine. Where the maximum supply harmonics stipulated by the elec company apply to a particular area, it's usual to fit Phase Shift Transformers to essentially provide extra phases, and then feed the diode or thyristor bridges equally across the existing and new phases. Can this be pressed to good use in theatre applications with conventional dimmers?

[Ike]

Extra transformers and wiring wouldn't in my opinion be practical in the industry but neither should they be necessary. If manufacturers pulled out their fingers and started investing more in developing low cost sine wave dimmers power distortion problems could be a thing of the past.

 

oh go on, write the formula for the fourier transform to the 21st harmonic - I dare you :** laughs out loud **:

No chance. It's something I've been meaning to swat up on for about four years and I still haven't got round to it. Ah well there's always next year... :P

[dbuckley]

If manufacturers pulled out their fingers and started investing more in developing low cost sine wave dimmers power distortion problems could be a thing of the past.

 

Is this not more down to patents and licencing? Having said that, the patents must be getting old now...

[Ike]

I'm not to sure what's actually patented but even the companies with the patents don't seem to be doing much with them and instead are just churning out the same old products and phase angle IGBT dimmers. Hardly very forward thinking is it? What's worse is they'll most likely expect our sympathy when a company on the over side of the world nicks 'their' ideas and starts manufacturing the products they were to scared to.

[KevinE]

a company on the over side of the world nicks 'their' ideas

 

Beat me in saying that!

 

No doubt the topic is already being studied by some designer far away!! Maybe the IGBT's are on the breadboard already and will be on the container next week ;-)

[David A]

Lite-Puter TW have just registered a patent on sine wave dimmers.

[jpdavis]

Im still confused. Whats the power factor?

[Ike]

Basic electricity theory states that power is equal to voltage multiplied by current. Now in AC systems the voltage and current are always changing so we use an average figure for these known as the RMS or Root Mean Square value. Mains here in the UK is 230V RMS, a 13A plug is designed for 13A RMS, a 15A for 15A RMS and so on. This however gives us a problem because for the maths to work the current has to be directly proportional to the voltage however as you can see in the above example this isn't always the case. The current and voltage must also be in phase which again doesn't apply with some loads.

 

For the majority of loads we therefore need to make a distinction between the voltage multiplied by the current which we call apparent power and the actual power the device is using which is referred to as the real power. As you'd expect we measure the real power in Watts but the apparent power is just a meaningless number so we use the units VA or Volt-Amperes which simply tells whoever is reading it we've multiplied the voltage by the current.

 

Now in order to answer your actual question the power factor is the ratio of real power to apparent power written either as a percentage or number between zero and one. The higher this number the better, a low power factor is a sign that we have current flowing around that isn't actually doing any work but still creates heat in the cabling and unnecessary losses.

 

For the above example it can be taken that the current and voltage are in phase so all you have to do is use the following formula to calculate the power factor:

http://www.blue-room.org.uk/brwiki/images/6/67/Power_factor.jpeg

The THD(%) is the total harmonic distortion percentage which is simply the sum of all the individual harmonic percentages obviously excluding the first/fundamental. In the real world we tend to only bother about the first 21 or so, in the case of the example you have no choice as I only made up figures for that many.

 

If you do it right you should find the example has a power factor of around 0.54 which is very very poor.

[dbuckley]

we use an average figure for these known as the RMS or Root Mean Square value.

 

Pedant - average and RMS values can be very different!

 

What is beyond argument (is that a foolish thing for me to claim?) is that the RMS current has the same heating effect of a DC current of the same magnitude.

[Ike]

Pedant - average and RMS values can be very different!

The RMS (or quadratic mean) is still a type of average it's just rarely referred to as one.

[matt c]

I feel a calculus moment coming on.

 

I have vague recollections of calculating RMS values for all sorts of nasty looking (but repetitive) waveforms.

 

I had a wonderful program on my TI85 calculator where you input the relative amplitudes of harmonics, or a known equation for a waveform, and a few other parameters) in at one end and it spewed out all sorts of wonderful figures including power factor and RMS values for the current.

 

Unfortunately, when the office move took place, I was on holiday, and no-one thought that the little black cased thing on the windowsill was worth anything, so it went in the skip, along with the £200 soldering station I'd bought a couple of months previous. Am I allowed to swear? B*******.

 

Matt.