Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Data centers are transitioning from AC to DC

spectrum.ieee.org

Post reply on HN
Print view View original

391–400 of 432 posts

Re: Data centers are transitioning from AC to DC

#391

Post by reportingsjr »

t0mas88 wrote: post #300
riedel wrote: post #292

Earlier quoted context omitted.

An EV fast charger can do 1000V, so with a bit of logic that sounds doable.

EV chargers take a different approach. There is no power on the connector while you're plugging it in. It then locks in place before the contactor closes and power is delivered. Unplugging is the same, power is removed before the plug is unlocked for release.

This is effectively how hot swapping is done for may modern systems. It seems counter intuitive to call it hot plugging, but it’s what you’ll find if you look up “hot swap controller”.

Re: Data centers are transitioning from AC to DC

#392

Post by ssl-3 »

MathMonkeyMan wrote: post #42
bluGill wrote: post #16

Earlier quoted context omitted.

48vdc was common in phone exchanges. They filled the basement with lead-acid batteries and to could run without the grid for a couple weeks. In turn the phone was 99.999% reliable for decades.

Yeah I always heard that the phone lines carried their own power, and in Florida the phones did keep working when the power went out, but I never knew why. So the grid was always charging up the lead acid batteries, and the phone lines were always draining them? Or was there some kind of power switching going on where when the grid was available the batteries would just get "topped off" occasionally and were only dra…

Neither.

The batteries, the grid/generator-supplied power supplies, and the telephone switch equipment are all connected in parallel -- as if the entire DC power infrastructure consists of only two wires, and everything involved with it connects only to those two wires.

1. In normal operation, the batteries are kept at a constant state of charge. The switches are powered from the same DC bus that keeps the batteries charged.

2. When the power grid goes down, the batteries slowly discharge and keep things running like nothing ever happened (for hours/days/weeks). There is no switchover for this; it's just the normal state, minus the ability to juice-up the batteries. (Remember: It's just one DC bus.)

3. When the grid comes back up (or the generators kick in), the batteries get recharged. There is no switchover for this either; nothing important even notices. (Still just one DC bus.)

4. If the grid stays up long enough, go to 1. Repeat as the external environment dictates. (And as you might guess, it's still one DC bus and there's also no switchover here. Things just continue to work.)

--

You can play with this at home with a capacitor (which loosely acts like a battery does), an LED+resistor combo (which acts as a load), and a small power supply that is appropriate for LED+resistor you've chosen (which acts as the AC-DC converting grid input).

Wire them all 3 parts up in parallel and the light comes on.

Disconnect the power supply, and the light stays on for a bit -- it successfully runs from power stored in the capacitor.

Reconnect the power supply, and the light comes on and the capacitor ("battery") recharges -- concurrently.

Improve staying power by adding more parallel capacitance. Reduce or eliminate it by reducing or eliminating capacitance. Goof around with it; it's fun. (Just don't wire the capacitor backwards. That's less fun.)

Re: Data centers are transitioning from AC to DC

#393

Post by rdtsc »

Animats wrote: post #144
rdtsc wrote: post #135

Earlier quoted context omitted.

It is a pretty clever design > When it is detected that the PDB starts to detach from the interface, the hot-swap controller quickly turns off the MOSFET to block the discharge path from Cin to the system. After the main power path is completely disconnected, the interface is physically detached, and no current flows at this time > For insertion, long pins (typically for ground and control signals) make contact first…

Yes, I read that. There had better be a backup system. MOSFETs tend to fail ON, and there's a megawatt going into each rack. Somehow this seems the wrong approach to AI.

I figure the latch-up failure mode is a known failure mode. It's not something totally unexpected.

> Somehow this seems the wrong approach to AI.

One a more ironic level, it's seems on par with the whole AI approach!

Re: Data centers are transitioning from AC to DC

#394

Post by mikewarot »

Ok, this is just plain nuts.

800 volts DC, at the megawatt power supply levels, implies fault impulses of more than a megajoule. Google tells me that's about 2 hand grenades worth of boom. That's an optimistic lower bound.

The resulting copper plasma cloud is a burn and inhalation hazard, along with the overpressure.

Let's say you get a 10 kiloamp fault current, this will then induce voltages everywhere you don't want it to go. If all the interconnects are fiber, that's really not a problem, but you have to have everything EMP shielded if you don't want boards popping randomly after such an event.

The "efficiency" of removing the extra power conversions also removes filtering and surge suppression. It's entirely possible that one power supply over-voltage takes out half of your racks. The MOSFETs used tend to fail closed instead of open, making failures far worse than a simple outage.

Very smart people are making very smart mistakes.

Re: Data centers are transitioning from AC to DC

#395

Post by MrBuddyCasino »

lostlogin wrote: post #305
fc417fc802 wrote: post #162

Earlier quoted context omitted.

Ahckhually US residential is split single phase, not two phase. https://en.wikipedia.org/wiki/Split-phase_electric_power I think the answer to your question is that it mostly doesn't matter for personal mug size quantities of hot water and if it does matter to you there are readily available competing options such as dedicated taps for your kitchen sink. Perhaps the biggest reason is that a traditional kettle on any…

I use an induction stove on maximum to boil water as I get irritated at the time it takes to boil water with 240v. I’m surprised that American exceptionalism can tolerate half powered sockets.

The induction stove boost function is glorious. Using twice the power temporarily is such a timesaver!

Re: Data centers are transitioning from AC to DC

#396

Post by chris222 »

Animats wrote: post #118

800V to each rackmount unit, with hot plugging of rack units? That's scary. The usual setup at this voltage is that you throw a hulking big switch to cut the power, and that mechanically unlocks the cabinet. But that's not what these people have in mind. They want hot-plugging of individual rackmount units. GE has a paper about the power conversion design, but it doesn't mention the unit to rack electrical and mechan…

A high powered EV charger is 800 volts. Totally safe and outdoors in thousands of places.

Re: Data centers are transitioning from AC to DC

#397

Post by shiroiuma »

manwe150 wrote: post #93
jacquesm wrote: post #53

Earlier quoted context omitted.

Not going to happen. For the same reason that the US never converted to a higher domestic voltage even though there are many practical advantages. The transition from one system to another at the consumer level would be terrible, even if there would be some advantage (and I'm not sure the one you list is even valid, you'd get DC-DC converters instead because your consumers typically use a lower voltage than the house…

I suppose that still begs the question somewhat, since the US does have 240V (2 phase) already driving many appliances. Why hasn’t it ever become standard for luxury kitchens to have a European-style outlet for use with a European kettle? I know the US already has a different 240V plug shape, so it might have to be an unlicensed installation, but surely someone wanted hot tea faster and did that calculus before?

>Why hasn’t it ever become standard for luxury kitchens to have a European-style outlet for use with a European kettle? I know the US already has a different 240V plug shape, so it might have to be an unlicensed installation, but surely someone wanted hot tea faster and did that calculus before?

I've heard of Americans getting EU plugs for things like this, but it's extremely uncommon. The simple answer is: Americans just don't drink tea very much, and the ones who do aren't going to go to this much trouble and expense just to be able to boil water for their tea twice as quickly.

Re: Data centers are transitioning from AC to DC

#398

Post by fragmede »

shiroiuma wrote: post #397
manwe150 wrote: post #93

Earlier quoted context omitted.

I suppose that still begs the question somewhat, since the US does have 240V (2 phase) already driving many appliances. Why hasn’t it ever become standard for luxury kitchens to have a European-style outlet for use with a European kettle? I know the US already has a different 240V plug shape, so it might have to be an unlicensed installation, but surely someone wanted hot tea faster and did that calculus before?

>Why hasn’t it ever become standard for luxury kitchens to have a European-style outlet for use with a European kettle? I know the US already has a different 240V plug shape, so it might have to be an unlicensed installation, but surely someone wanted hot tea faster and did that calculus before? I've heard of Americans getting EU plugs for things like this, but it's extremely uncommon. The simple answer is: Americans…

If speed is the issue, Japanese water boilers keep water at or near boiling so there's less of a wait, and they operate on 110v.

Re: Data centers are transitioning from AC to DC

#399

Post by the_cosmos »

wat10000 wrote: post #288
HWR_14 wrote: post #241

Earlier quoted context omitted.

So, there is a true value for 0? Does that mean when you have electronics and use multiple dc-dc converters all the inputs and outputs share the same ground, it's not just the values for that pair of wires? And if I want to use a telephone on an incorrectly wired 48dc circuit, I could switch the positive and negative wires, as long as the circuit in the telephone is isolated and never touches ground? Thanks. Somehow…

There is a true zero potential. You can detect this because two charged objects with zero delta between them will still repel each other. I think a circuit should mostly care about the deltas, but when you’re talking about things like phone lines, the earth becomes part of your circuit. You can’t influence its potential (it’s almost exactly neutral because any charge imbalance gets removed by interaction with the int…

Voltage is _always_ relative; you only ever measure differences between two points. Voltmeters have two probes. Ground is just a convenient reference, not a universal zero. And while Earth can act as a reference or return path in some systems, it’s not perfectly neutral or fixed...it’s just large enough to approximate that in practice. (The interplanetary medium aka aether doesn't exist)

Objects don’t repel because they’re at the same potential. Electrostatic force comes from electric fields due to charge. If two objects truly have zero potential difference and no field between them, there’s no force.

You’re correct that circuits care about voltage differences. After all, all work requires a force gradient of some kind.

Post reply on HN