Will AC power distribution become obsolete?

[Marco]

Another reason why a single low voltage DC net in your house is not feasible.

The point was not a low voltage DC net, the point was proliferation of low voltage plugs for most appliances.

DC microgrids are generally dreamed of being 350V DC, but with DC buck is trivial and reliable so adding some at the socket is less of an issue (zero-crossings are a bitch, even when you ignore power factor).

[coppice]

AC-DC radios  which, as their name suggests, can operate on ac or DC mains, were widespread across North America & Europe during the 1950s/'60s, mainly, in the NA case because the manufacturers could save a few cents on power transformers.

Saving a mains transformer saves a lot more than a few cents, but cost saving was not the only reason for avoiding a mains transformer for a radio intended to be compact and easily moved. It saved considerable weight and space. In the case of TVs, the saving was enormous. If you've seen a mid 50s TV that still used a transformer, you'll know about this. The mains transformer in those could be massive.

[pcprogrammer]

Another reason why a single low voltage DC net in your house is not feasible.

The point was not a low voltage DC net, the point was proliferation of low voltage plugs for most appliances.

It is what the OP was on about. 30VDC throughout the house.

DC microgrids are generally dreamed of being 350V DC, but with DC buck is trivial and reliable so adding some at the socket is less of an issue (zero-crossings are a bitch, even when you ignore power factor).

And with that comes the need for a lot of replacing these sockets when lightning strikes nearby. A simple AC transformer is much more robust than electronics devices.

The way it is now at least allows you to pull all the plugs when a thunderstorm comes your way.

Luckily I'm old enough to be dead by the time this would ever take place. Sounds like a lot of trouble and waste of money for nothing.

[Marco]

One of our buildings runs entirely (every light and every socket, every appliance) through a large UPS, which has the same effect.[/color][/font][/b]

Goes to show, AC introduces its own host of complexity and inefficiency.

If HVDC becomes common, some new industrial and services location will want in on it. Large electrolyser plants will lead the way.

[madires]

DC microgrids are generally dreamed of being 350V DC, but with DC buck is trivial and reliable so adding some at the socket is less of an issue (zero-crossings are a bitch, even when you ignore power factor).

There are already data centers with a HV DC distribution to power tons of servers, meant to increase the efficiency a bit. However, it's a special application, not for general use. A DC PSU is basically the same as an AC one, but you can remove the bridge rectifier (or replace it with a single diode for reverse polarity protection) and also do away the active PFC. This makes it a bit more efficient and cheaper. But it doesn't really change the overall design or reliability. On the other side, a HV DC distribution is more expensive and makes only sense when you're running a large data center full of servers (with DC PSUs).

[SeanB]

Yes biggest issue with DC is circuit protection, 230VAC a cheap and relatively simple fuse or breaker will act as circuit protection, and for overheating a simple thermal switch with either bimetal contacts or a melting link works well, as all of them will break arcing with each zero crossing. DC over 48V requires very specialised fuses, and the breakers need to have magnets to divert the arc to chutes to cool it, and then you cannot use the simple reliable thermal breakers to act as last ditch protection. Under 50V, AC or DC, you can use simple fuses, and not have much of a difference in clearing time.

[johansen]

What industry is left in your country?

None at 60Hz. Like the majority of the planet.

I think you need to leave the office more and take a walk outside...lol

[Marco]

A DC PSU is basically the same as an AC one, but you can remove the bridge rectifier (or replace it with a single diode for reverse polarity protection) and also do away the active PFC.

No electrolytics, on average a third less high voltage switches and no electrolytics.

[madires]

Nope! On the primary side you want to keep the HV electrolytic caps as filter/buffer caps. One reason is the inductance of the HV DC power wires. On the secondary side you'll still need the low-ESR electrolytics, which are much more likely to fail than the HV ones on the primary side.

[nctnico]

A DC PSU is basically the same as an AC one, but you can remove the bridge rectifier (or replace it with a single diode for reverse polarity protection) and also do away the active PFC.

No electrolytics, on average a third less high voltage switches and no electrolytics.

Sorry, but you are very wrong about this. A DC grid needs to stay DC which means you'll need to filter large current ripples. This is a requirement for DC powered telecom equipment for example. A couple of years ago I oversaw the design of a piece of telecom equipment and it needed an active filter to prevent fan commutation currents to exceed input current ripple requirements. The only way out is to use large electrolytic capacitors. The problem is pretty much equivalent to needing PFC for AC powered equipment. And then you'll need rush-in current limiting (some kind of MOSFET contraption), galvanic isolation and reverse polarity protection (which is easy to implement using a bridge rectifier). The bottom line is: an AC or DC grid power supply will be equally complicated and expensive.

As a side node: the main reason telecom equipment is powered from 48V DC (more like 56V to 60V in practical cases) is that back in the old days telecom exchanges had a bunch of huge lead acid batteries which served as a backup power supply in case of a mains power outage.

[Marco]

Sorry, but you are very wrong about this. A DC grid needs to stay DC which means you'll need to filter large current ripples. This is a requirement for DC powered telecom equipment for example. A couple of years ago I oversaw the design of a piece of telecom equipment and it needed an active filter to prevent fan commutation currents to exceed input current ripple requirements.

For a 4 pole fan at 1000 RPM you need to bridge 250 microseconds, for PFC at 50 Hz AC you're bridging 10 milliseconds ... the problem is nearly two order of magnitudes different and that's less than the difference in volumetric capacity between electrolytic and film capacitors.

Now it's a mechanical component so do whatever is cheapest, but for locations where reliability is higher priority (led lighting, socket integrated converters, whatever) the order of magnitude difference in the timescales which need to be bridged do allow you to abandon electrolytics where you otherwise could not.

[nctnico]

Sorry, but you are very wrong about this. A DC grid needs to stay DC which means you'll need to filter large current ripples. This is a requirement for DC powered telecom equipment for example. A couple of years ago I oversaw the design of a piece of telecom equipment and it needed an active filter to prevent fan commutation currents to exceed input current ripple requirements.

For a 4 pole fan at 1000 RPM you need to bridge 250 microseconds, for PFC at 50 Hz AC you're bridging 10 milliseconds ... the problem is two order of magnitudes different and that's less than the difference in volumetric capacity between electrolytic and film capacitors.

Nope. As an example: the active filter I wrote about needed 1000uf electrolytic capacitors. The problems are way more complex than you think.

[Connecteur]

No one is calling for the end of AC current, of course.  That would be ridiculous.  Just the use of AC for power distribution.  It's only a question after all, nothing more.

[vk6zgo]

The tube filaments all drew the same current, with different voltages, so they could be wired in series and run directly from 120 volts AC or DC.

Televisions in the UK all used to be constructed this way. As a child I used to scavenge old equipment for power transformers, and I soon learned that TV sets did not have transformers inside them, even though they had many valves (vacuum tubes). They also had a huge wire wound power resistor to use a voltage divider. That thing was mounted in free air not touching anything because it got very hot in use.

TV came along in 1956 in Oz, & almost universally, used power transformers (the only exception I can remember was the Admiral 17" "portable".
It wasn't really portable, just a mains operated one with a handle on the top!
I could never fathom why the UK went for transformerless construction, as any saving in cost or weight with not having a transformer would be more than offset with the huge wirewound resistors used ( weren't they series droppers because all the valve filaments added together didn't reach 240v ? ) & all the extra insulation required.
Ok, I understand that there were DC mains still in some places, but surely they could have been built as specials.

[soldar]

As a side node: the main reason telecom equipment is powered from 48V DC (more like 56V to 60V in practical cases) is that back in the old days telecom exchanges had a bunch of huge lead acid batteries which served as a backup power supply in case of a mains power outage.

No, POTS actually works with DC, it needs DC, the subscriber loop is DC. It uses AC only for the ringer while it rings.

[nctnico]

As a side node: the main reason telecom equipment is powered from 48V DC (more like 56V to 60V in practical cases) is that back in the old days telecom exchanges had a bunch of huge lead acid batteries which served as a backup power supply in case of a mains power outage.

No, POTS actually works with DC, it needs DC, the subscriber loop is DC. It uses AC only for the ringer while it rings.

I know that but the function of the batteries remains the same: backup during a power outage. During normal operation, the system is supplied from mains using a transformer + rectifier.

[Xena E]

The tube filaments all drew the same current, with different voltages, so they could be wired in series and run directly from 120 volts AC or DC.

Televisions in the UK all used to be constructed this way. As a child I used to scavenge old equipment for power transformers, and I soon learned that TV sets did not have transformers inside them, even though they had many valves (vacuum tubes). They also had a huge wire wound power resistor to use a voltage divider. That thing was mounted in free air not touching anything because it got very hot in use.

TV came along in 1956 in Oz, & almost universally, used power transformers (the only exception I can remember was the Admiral 17" "portable".
It wasn't really portable, just a mains operated one with a handle on the top!
I could never fathom why the UK went for transformerless construction, as any saving in cost or weight with not having a transformer would be more than offset with the huge wirewound resistors used ( weren't they series droppers because all the valve filaments added together didn't reach 240v ? ) & all the extra insulation required.
Ok, I understand that there were DC mains still in some places, but surely they could have been built as specials.

This was a technical history study subject of mine, it wasn't to do with the TV sets themselves per se, it was a study on the effects of asymmetric loading of the power grid, but it did throw up some interesting historical development details on the TV sets themselves.

Not all the UK TV production was transformerless, there were notable exceptions. (Example: 1957 PYE PTV, which were based on a studio monitor, apparently excellent sets, it needed good isolation as it had a metal cabinet!)

Early, just pre and post 2nd world war CRT sets had transformers as it was the prevailing design choice at the time and often had mains derived CRT final anode HT... absolutely lethal.

Later solid state colour sets nearly all had some form of switching supplies/phase angle thyristor regulation or similar, often non isolating. Exceptions were a few quirky hybrid sets that were very power consumptive, notable was the 1960s  Philips G6 chassis, that in 22 and 26" screen varieties consumed 400W: they also used shunt stabilisation of the EHT which was a radiation hazard to any foolhardy engineers that ran them without the sweep/line stage screening box in place.

The late fifties to early seventies monochrome sets did mostly  use transformerless techniques and achieved the necessary dropping of power for the tube/valve heaters with series resistive droppers in combination with series diodes to limit dissipation.

A few sets even used lossless capacitive droppers for the heater chains.

A whole range of tubes/valves were produced in Europe specifically for use in these sets and were designated 'P' range by the European manufacturers, these had heaters that were  for use in  300mA series strings, had suitably increased heater cathode insulation,  and were characterised for the relatively low plate/anode voltages provided by the ½ wave mains rectification.

The reasons the system was adopted was purely cost, it being very much cheaper to make ceramic resistive droppers than transformers, they were no where near as expensive as transformers to produce, and factors lighter, though that was coincidental considering the weight of the glass CRT. Virtually none of these sets would run on DC mains, (diode droppers don't work on DC, the tube/valve heaters would be overrun  for starters).

Not all of the sets sold in Aus were transformer PS types, notably European designed imports often weren't. I don't think there was ever any regulation introduced that they should be isolated supplies, it was just a sensible risk adversity on the part of Australian manufacturers, perhaps only of benefit to service techs.

Of course, in operation in customers homes, the extra safety of a transformer PS set was then arguably not applicable.

In the case of a resistive dropper if it is overloaded it will just go open circuit, a lot of sets in-fact had droppers that had fuseable spring links. Almost any old TV tech., will tell you that replacement dropper resistors were the bread and butter of their living.

Transformers when overloaded can of course catch fire, particularly the ones fitted in veteran sets, where the windings were often potted in wax or bitumen.

My personal choice would perhaps be the transformer set being the safest overall but I don't think such a study was ever done, in service reliability being a completely separate matter entirely.

Regards,
X

[rsjsouza]

The late fifties to early seventies monochrome sets did mostly  use transformerless techniques and achieved the necessary dropping of power for the tube/valve heaters with series resistive droppers in combination with series diodes to limit dissipation.

A few sets even used lossless capacitive droppers for the heater chains.

This was also common in Brazil's sets of the time as well - with the difference that, having two mains voltages around the country (110/220V at the time), this was switched by simply dropping half of the 220V using a half wave rectifier. For 110V, a TV set could get by with a very minimal resistor, since the combination of a PL36 (horizontal - 25V) + a PY88 (damper - 30V)+ a combination of PCL82/PCL84/PCL85 (audio,vertical ~ 15V each) would already get you to 100V.

A whole range of tubes/valves were produced in Europe specifically for use in these sets and were designated 'P' range by the European manufacturers, these had heaters that were  for use in  300mA series strings, had suitably increased heater cathode insulation,  and were characterised for the relatively low plate/anode voltages provided by the ½ wave mains rectification.

I would imagine the "H" (150mA) and the "U" (100mA) series would be much more popular in Europe due to the higher supply voltage.

[SeanB]

Yes the non isolated chassis was common till the age of the VCR, because the TV set only had an antenna input, and a headphone jack, both of which are easy to isolate, using 2 1n 3kV capacitors for the antenna input, and a simple robust insulation audio transformer to drive the speaker, also having as bonus matching for the 16R speaker to the 3R output impedance of the transistor output stage, running off around 16V. Non isolated just meant plastic shafts for all the controls, and long shafts as well, to place them at least 5mm inside from the plastic or wood case and front, and a plastic rear cover with a warning label, and a detacheable mains cable that came off with the rear backing.

Then came the first generation of VCR's, and a demand arose for a way to connect audio and video into the set, helped by the adoption of SCART in the European area, which predates the EU, but where the assorted countries wanted to harmonise on TV standards, and have a common interconnection for all this consumer audio and video gear. So the first ones used a non isolated chassis, and had a small power supply that provided an isolated supply for this circuitry, with assorted forms of transfer across the isolation barrier, from photodiodes and phototransistors, to modulating it on a 20MHz or so carrier and using some 10pF capacitors to carry it, to using RF transformers to do the same.  Then you got sets later on that had a isolated control and input, and non isolated HV generation, deflection and such, like Salora, who used the LOPT as both power supply and isolation, with the primary drive and horizontal deflection running non isolated, but EHT, 200V CRT drive, heater and all other parts running off isolated windings on the LOPT, and then a low power SMPS doing the standby supply for the microcontroller and the jungle chip, which performed all the set functions in a single package, all controlled over I2C to set up hundreds of internal registers to determine bands covered, IF bandwidths, colour carriers and TV system along with audio types, all in one single package.

Till the final versions isolated all the TV functions, using a SMPS to provide a 5V supply for the microcontroller, and a second switchable one that provided all the other rails when commanded, so the TV set would be in standby till the SCART input commanded it on, or video was present, or the IR remote powered it on.

As to asymmetric loading, the first generation Phillips G11 used a thyristor power supply, and, as the UK has polarised sockets and plugs, and these came with a 2 wire lead with factory attached plug, they all were wired identically, so they were responsible for a good number of older distribution transformers catching fire, as the 11.6A pulse of current every second mains cycle all added up, in residential areas, to a massive DC current flow in the older transformers, which would then have the core saturate, and then blow fuses, or overheat and catch fire from massive primary side current as the inductance dropped to basically air cored.  Rewinds put in a distributed oil gap instead, stacking the laminations to give a small gap between them in the centre of the core, instead of a butt join, and a slight correction to the lamination stack and windings to compensate, plus bigger oil coolers.

[madires]

No one is calling for the end of AC current, of course.  That would be ridiculous.  Just the use of AC for power distribution.  It's only a question after all, nothing more.

From your post starting this thread:

So why do we need AC in our homes? We don't. A house could be supplied with DC power from the mains, and converted to any type of power we could require. Imagine a future where DC power receptacles are installed in a house, at a safe voltage, where electrocution is impossible, and converters simply transform the safe current into anything that is required.

[Xena E]

A whole range of tubes/valves were produced in Europe specifically for use in these sets and were designated 'P' range by the European manufacturers, these had heaters that were  for use in  300mA series strings, had suitably increased heater cathode insulation,  and were characterised for the relatively low plate/anode voltages provided by the ½ wave mains rectification.

I would imagine the "H" (150mA) and the "U" (100mA) series would be much more popular in Europe due to the higher supply voltage.

In the UK "U" series (100mA) were predominant in radio sets with transformerless chassis late fifties onwards, "P" series were the most used ones in domestic TV sets, very often combined with dual use "E" range in the signals stages that were designed to have 300mA heaters and controlled warm up.

In an early valve/tube based transformerless TV the total voltage of the series heaters meant that the "P" range was the choice having perhaps 13 or more valves/tubes with the line output stage ones having a significant volt drop on the heaters, adding up to a total too much for other lower current series.

Transformer powered TV sets obviously used mostly "E" range. These were the minority in the UK.

[Marco]

Nope. As an example: the active filter I wrote about needed 1000uf electrolytic capacitors. The problems are way more complex than you think.

The ripple current on that capacitor has to be almost non existent, so it's a strange circuit ... if you get near rated ripple current at 250 us cycle time it would have to be a very big fan. Big enough that using film capacitors even at mF level is but a rounding error in the cost.

Here are some of the industry dreams of DC Microgrids, obviously a lot of this was investment bubble driven, but it does show a broad interest from industry at least. Also a nice picture of a high power lighting power supply compared to an AC one.
https://www.dc.systems/
https://currentos.foundation/

On a long enough timescale, the advantages of driving circuits while not needing to worry about the comparatively huge 10 msec gaps across single phases and PFC across those same gaps is going to create demand. Note that when I made the argument about reliability it was about single phase power at residential sockets, the needed capacitance for three phase is less but at the same time the required high voltage switches increases massively. So that doesn't help AC much.
Industry demand is going to lead to more DC when it becomes available at the distribution level. Not 100 year old plants, but new plants ... and new plants slowly replace old plants eventually. On a long enough timescale without other apocalyptical events you could see a slow transition to DC because of the combined forces of distribution and industry/microgrid demand. Unfortunately making it that far without an apocalypse looks doubtful.

PS. before someone brings up the sand filled fuse, consider the failure modes a thermal fuse protects against in a modern microcontroller controlled power supply. Resettability of a fuse is not relevant when failure was only possible because the control loop failed and the power supply is likely screwed any ways. The microcontroller can read a thermistor just fine.

[David Hess]

On a long enough timescale, the advantages of driving circuits while not needing to worry about the comparatively huge 10 msec gaps across single phases and PFC across those same gaps is going to create demand. Note that when I made the argument about reliability it was about single phase power at residential sockets, the needed capacitance for three phase is less but at the same time the required high voltage switches increases massively. So that doesn't help AC much.

Three phase does not need to be higher voltage, and obviates a large part of the input holdup capacitor size.  Active power factor correction removes the bulk input capacitance in all cases of course.

In the US part of the lack of 3-phase power to residential housing is political.  It helps to prevent people from installing machine tools in their garage and running a business.  We already run 3 wires to homes for split phase 240VAC.  If the neutral was used for 3-phase, then less metal would be required for the same power.

[Marco]

Active power factor correction

Not really an example for simplicity over the complexity of DC though.

The need for active PFC for high power inverter loads, or very large and load dependent LC filters, belies the supposed simplicity of AC. Once most loads become inverters, AC is not simple.

[IanB]

In the case of a resistive dropper if it is overloaded it will just go open circuit, a lot of sets in-fact had droppers that had fuseable spring links. Almost any old TV tech., will tell you that replacement dropper resistors were the bread and butter of their living.

Transformers when overloaded can of course catch fire, particularly the ones fitted in veteran sets, where the windings were often potted in wax or bitumen

The resistive droppers did get rather hot though. If you had a TV stored or switched off for a while, then a layer of dust would accumulate on the dropper resistor. If you then switched on the TV, you would get a very characteristic burning smell coming out the back of the set as the resistor heated up. Ah, nostalgia.