# High frequency mains transmission

Someone told me that if we used something like 1 kHz AC instead of 50 or 60 Hz AC, it would end up being much cheaper and efficient, because we could use smaller transformers. Is this true? It seems like it would probably cost more as the generators would have to spin much faster - also, it could lead to more losses due to the skin effect. (Just out of curiosity.)

• Skin effect has a massive effect at 50Hz, never mind 1kHz. Nov 22, 2010 at 20:35
• It's about 2 millimeters at 1 kHz in copper, which would effectively limit what you could safely carry to ~40 A. Nov 22, 2010 at 23:40
• @Nick T, normally you are pushing almost 0 current in high power lines as the current is what causes loss. Nov 22, 2010 at 23:46
• @Kortuk - a common misconception. The voltage is high to reduce current, but nowhere near 0. Thousands of amps will flow in a 400kV transmission line. High voltage brings problems and loses itself. Nov 23, 2010 at 0:17
• @Cybergibbons, I could believe that. No worries mate. I had always been taught that, but never in a detailed fashion. Nov 23, 2010 at 1:35

Yes, transformers could be lighter. 400 Hz power is used on aircraft for this reason.

The generator wouldn't need to spin faster if it is designed with more poles. Motors would also have to be redesigned to run at normal speeds with 1000 Hz power.

But how often are 60 Hz transformers used in modern electronics? There is a big one on the pole outside of my house, and in my boat-anchor hi-fi amps and tube guitar amps. All the other power supplies in my house use switching supplies that run their transformers at a much higher frequency.

edit: Another thing, transformers tend to vibrate slightly during operation. A 60 Hz transformer will "hum" a little bit. But imagine all the appliances in your house buzzing at 1 kHz. 1000 Hz is two octaves above middle C, and our ears are 30dB more sensitive to this frequency than 60 Hz:

Image source: http://en.wikipedia.org/wiki/Equal-loudness_contour

• Cheap alarm clocks I've found use transformers. Also, an older scope I had used one. But SMPS supplies are becoming cheaper than transformers nowadays so you see them more. Nov 22, 2010 at 17:21
• A lot of inexpensive clocks count the cycles from their mains supply to use as their timebase. Turns out that while not super-precise over short time spans, the mains supply is very accurate when measured over days. They have to keep precise frequency to aid in load balancing and synchronizing generators. Nov 22, 2010 at 20:19
• @Markrages, nice, most people do not understand that our ears bias certain frequencies. Nov 22, 2010 at 20:20
• read the chart more carefully. The Y-axis is SPL and the lines are equal loudness. So a 20 dB SPL at 1 kHz sounds as loud as a 57 dB SPL at 60 Hz. Nov 23, 2010 at 0:42
• Switch mode power supplies use frequencies 50kHz - 1MHz+ to reduce the size of transformers and inductors.
– W5VO
Nov 23, 2010 at 4:18

You would certainly see problems related to skin effect on long cables, but you would get huge capacitive losses.

It takes some energy to charge a wire to several hundred kV, so doing that more often will need more energy.

The longer the wires the higher the capacity and the higher the losses.

What is actually done is to switch to DC for long / high-voltage lines, with DC the ends of the wire get more expensive because you have to have high-voltage semiconductors to handle rectification at one end and to turn the DC into AC at the other end, but modern semiconductors have brought down the price on those stations a bit, so but for very long transmission lines it's a huge win.

• Wouldn't the losses be cancelled out by the fact that a capacitor ideally doesn't waste energy (electronics.stackexchange.com/questions/5355/…)? Therefore any losses would be resistive and irrelevant of frequency (ignoring for now the skin effect.) Nov 22, 2010 at 10:47
• Technically, the charging current isn't the loss, it's passing the current through the resistance of the cable (and everywhere else it passes on its way) which is where the loss is.
– user1844
Nov 22, 2010 at 12:13
• @dren.dk, I think you may have some misunderstandings about transmission lines, in real life they match all of the transmission lines, and if everything was lossless the power would be pushed down the line without losing any power. Everyone is correct that if you pushed the line to 1kHz you would lose power, but that would be to resistive and radiation losses. Nov 22, 2010 at 15:52
• @dren.dk - I don't agree - with perfect conductors/generators and a perfect capacitor, the reactive power would cost nothing to make, but that's all irrelevant, as if you have one lot of stuff impossibly perfect, you might as well specify free fuel while you're at it...
– user1844
Nov 22, 2010 at 16:31
• @XTL, Here is a list of projects, en.wikipedia.org/wiki/List_of_HVDC_projects, I read about this type of thing regularly in the magazines I get like EDN. I am sorry to say it has been almost 5 months since I have had time to read one and me finding which it has been is a lost cause. HVDC is one of the major steps being taken to allow many different renewable sources to connect. They all connect at DC, then the system is relayed and switched back to AC. Nov 23, 2010 at 15:34

I suspect losses due to capacitance would be somewhat more significant than skin effect

• Capacitance introduces no losses. Transmission lines could incorporate inductors to keep the power factor up, just like today at 60 Hz they incorporate banks of capacitors. Nov 24, 2010 at 3:30
• Capacitance itself doesn’t cause losses but if the power factor is poor then the resistive losses are significantly increased
– Frog
Jan 1, 2021 at 1:58

Not my field, but if we had say 1-5 hz across long lines and 5-20 for shorter height and distance lines (for safety, so the same effect of not instantly dying from the shock still applies) and then have decently high frequency converted at the entrance to buildings, that way syncing would be quick and easy for grid tie systems, but all the in home transformers would use the high frequency, or something along those lines.

• That makes a degree of sense but how did you derive those figures? Also the practicality of converting 1Hz to 50/60Hz is questionable since you’d need to store a significant amount of energy to maintain the output between cycles (less of a problem for 3-phase but still a consideration).
– Frog
Jan 1, 2021 at 2:02
• 1-5 Hz would not have any of the advantages of AC power at all. Might as well just go DC rather than such a low AC. Also, having multiple frequencies at different locations will prevent transformer use. So other techniques will have to be used at any place where the frequency changes. Basically, this sounds like a nightmare. Jan 1, 2021 at 3:37
• Wouldn’t the benefit of arc cancellation be there still? Jan 3, 2021 at 5:35