Very high current power supply (2012) | Hacker News Reader

Very high current power supply (2012)

danyk.cz

70 pointsmindcrime45 comments

A while ago there was a discussion here (https://news.ycombinator.com/item?id=20293041) about right-angle bends in PCB traces and why we avoid them. One of the suggested reasons was that they distribute current poorly and can cause failures under high current. This video (https://youtu.be/weJdXaGRJGY) from the article is a great illustration of this in action, giving a literal heat map of current density.

I recommend looking into so called Super Capacitors or Ultra Capacitors. They are widely used where a quick and high initial load (power)is required.

AFAIK super capacitors are used for backup power systems and electric car batteries. I think I hear that Tesla battery packs have super capacitors to protect those Litium cells from getting damaged during short high load needs.

I am currently planning a small hobby project where I want to build a small autonomous sensor station using ESP32 and LORA, powered with solar cells using super capacitors as energy buffer charged through a buck boost converter with simple MPPT functionality.

A protection circuit in front of you appliance is mandatory since super capacitors can unleash high loads for a short period of time. I use here a second buck-boost converter which outputs constantly 3.3 volts for the ESP32 board and sensors.

My favorite use would be rail guns. Cuz they're awesome. :)

https://www.youtube.com/watch?v=fKEaDhDTciQ

Some are funnier than they are awesome:

https://www.youtube.com/watch?v=vAs9EHtKfVc

I've worked at a much smaller scale then those (12" length), but yes capacitors are very fun.

I used a screwdriver taped to a meter stick to complete the circuit : )

My interest in all this started largely from a (currently stalled out) initiative to build a coil gun. Cuz they're also awesome. :-)

Having needed something like this recently, I would recommend utilising 48V-12V convertors designed for servers. Telecoms equipment has long been 48V (actually -48V) but some datacenters are moving that way too for internal distribution.

There are some automotive designs for 48-12V conversation, but these are not really available as modules as automotive is large volume and high integration. Using discrete components gets expensive fast at these power levels, a Renesas 2KW reference design I saw needed a 4 layer, 6oz PCB which would have cost ~$500 to prototype in China.

The Vicor BCM series can do 48V->12V 150A out in a very nice self contained module, with excellent EMI shielding, and takes cables directly, although definitely needs a heatsink! http://www.vicorpower.com/products?productType=cfg&productKe...

What did you need so many amps for? My imagination falls short of melting things (like OP did).

Switchgear testing, for HighVoltage (~110KV++) or MediumVoltage (~10KV to 110) you usually will test only the secondary side, for LowVoltage (~400V -10KV)you can test the Breaker with a primary source, these Breakers range from 100A to 5000A.

But then you would need IGBT's not MOSFET converters.

Powering some 12V motors for a robot from a 48V tether supply.

Quite easy to get into hundreds of watts with relatively small brushless motors, the type that would normally be used in hobby RC, powered from LiPos. The Vicor parts have a good amount of protection circuitry as well, useful when shorted wires can start glowing very quickly, and RC lipo packs have no protection at all.

Galvanoplasty, electrolysis, and many other electricity-dependent chemical reactions, strong electromagnets and small wire-length actuators (like rail guns).

Designing CPU power supplies that are efficient and are not just a great lot of buck converters in different phases (VRMs) is very hard.

On modern CPU that are all running at at around 1V, you get 1 amp = 1 watt

Switching element must survive that, and you have to be super responsive to transients, and you have to have super low ripple... I do not envy guys who make VRMs for GPUs and 200w server chips

Blade Chassis power supplies are cheap used, very efficient, and will do nearly 200A at 12V from 240V input.

Search HP 500242-001.

I'll second that, I believe people use them for RC chargers sometimes. I used the HP DPS 600PB for trying to drive a number of peltiers. I think mine only did less than 50A, will look into the model you mentioned.

I assume none of the blade PSUs also output 5VDC?

If you can pick up a DS850-5 it'll give you 12v @ 70A and 5v @ 4A. It'll have noisy fans that run constantly, though.

Most applications of such a power supply (like blade servers) would have a bunch of secondary SMPSes generating lower voltages from the 12v rail.

I have 5 at home, exactly from a Blade Center.

Any idea what I can do with them?

This guy has a YT channel DiodeGoneWild. If you're into EE or dodgy stuff from eBay, do check him out. Recently he tore down a modern ATX power supply and did a very in-depth explanation.

Interestingly, working with low-voltage, high-current power supply has an uncommon danger: it won't kill you by giving an electric shock, but you may suffer from serious injury due to joule heating. You would create a massive electric cooker if there's any accidental short from power to ground, especially, don't wear any jewel or ring.

Bonus points if high frequencies are used, as they are in this one I bet. Your things are antennas, and the transformer if any is an electromagnet too.

You mean 12 volts at ~unlimited amps supply ran through my body may heat me up from the inside? Isn't the human body resistance too high for that, even if you go below the skin on both ends? At 10 ohm it would be just ~15 watts, hardly dangerous given that human bodies at rest produce ~100 watts of heat.

Edit: fixed numbers.

You can also get burned without current flowing through your body, e.g. by accidentally shorting the circuit through some sweat on your skin.

Uh right. But still - my skin sweat having less than 10 ohm resistance sounds like a stretch. It is more salty than blood though (because of evaporation).

If the droplet of sweat has a resistance of 10 Ohms, 1.2 Amps will be flowing at 12V. The droplet will be heated by 14 Watts of energy. Doesn't sound much. But it's enough to bring a droplet of sweat to the boiling point within 10 seconds.

Your sweat starting to boil of your skin doesn't sound like a very pleasant way to pass the time.

I was thinking on a longer distance, like between left and right hand, or between a hand and a leg (which I still think is near impossible to cause any harm with 12 volts). But surely it can happen on a distance of less than few cm, when the + is in a very close proximity to the - (touch here to get your fingertip heated!).

Doesn't need to be a terminal, touching a broken wire with sweaty hands would be enough to induce flash evaporation of fluids on your skin.

not with 12V. 12V batteries are pretty safe for human skin contact. If you put your wet tongue across a 12V battery it will hurt, but not more than that (it's a good way to test 9V batteries actually)

only on HN..

Well, if you put your tongue on it, you'll obviously have a lot more distance to cover and a lot more fluid to handle the current.

Batteries in general are fairly safe because their voltage breaks down at high loads. The chemistry can only react so fast internally and is controlled by various factors.

Contrast this with capacitors, where there is always a general safety warning when handling them in live circuits. They have no limit based on chemistry.

You don't want to test a capacitor with your tongue. A capacitor has no other current limit than it's internal resistance, with a modern sub10mOhm ISR a capacitor can, for the briefest of moments, discharge at up to 14 Kilowatts, quite the upgrade from the 14 measly watts that your spit will handle. A 100 Millifarad Low ISR Cap can sustain a kilowatt of output for almost 0.1 milliseconds, at which point you'll likely regret the choice of mishandling a cap. (It's only 14 Joules but these 14 Joules will be moving very fast)

When I was young I got a good burn on my skin pinching a tiny bit of my finger between wires hooked up to a 6v remote control car battery.

Smelled awful and had a brown patch a few mm wide on my finger for a while.

This happened more or less instantaneously.

When I was a kid, I stuck a 9V battery across my braces (top to bottom teeth). Just about knocked me out.

> You mean 12 volts at ~unlimited amps supply ran through my body may heat me up from the inside?

No. But you'll be injured if you short it with any metallic objects while touching it, like the wiring, the board, your probe, the ring on your finger, etc.

Electric submarines have pretty large batteries with very low-voltage cells but where a cell is big enough to deliver huge amounts of current. Completely safe to touch with bare hands, even wet hands, but if you drop a metal tool that shorts a cell, the tool may be vaporized and the resulting metal fumes are very dangerous.

Yep. Imagine a ground plane and a positive +12V pole able to supply infinite amps.

You can short circuit using your body and you'd probably feel a tingle (your body resistance is not high, but not too low either). But if the ring you're wearing on your finger is causing the short circuit, thousands of amps will flow through it, as it has an ultra-low resistance, and heat it up to temperatures that will quickly cause burns and whatnot.

If the device does not have infinite amps, then this is the case where normally the power supply's fuse would blow, and thus the ring would never heat up and nobody would get hurt.

To summarize: Low-voltage, high-current electricity is indirectly dangerous, whereas any high voltage electricity is inherently directly dangerous.

> Low-voltage, high-current electricity is indirectly dangerous, whereas any high voltage electricity is inherently directly dangerous.

Very high voltage with very low current is not inherently dangerous. Static electricity is this case. The total energy (voltage x current x time) passing through the body is the critical determinant of harm.

Try measuring your resistance with a multimeter. You'll find that it is way more than 10 Ohm, which is why 12V is not dangerous to humans.

Just do NOT stick the probes into your fingers, ie, don't break your skin. Depending on the multimeter this can be extremely dangerous.

https://darwinawards.com/darwin/darwin1999-50.html

Wow! You're not kidding...

> As my electrical safety instructor said, "The reason we now have to teach the electrical safety course to all electricians at least twice per year is because some joe was bright enough to be the one person in the world who could figure out how to kill himself with a 9V battery."

> Isn't the human body resistance too high for that

That must be true normally. But years ago I recall reading about someone who fixed an 12V underwater light in a pool. You would imagine that would be harmless, but he managed to create a circuit that ran across the pool. It was a salt water chlorinated pool, and after doing the job he dived in, and it killed him. I don't know how wide the pool was, but in any case the voltage across his body must have been less than 12V, and possibly a fraction of that.

These are very strangely drawn circuits, particularly the power supply that powers this power supply; in that one, the circuit seems to almost be drawn right to left, at least for the low-voltage portion (which is tapped off the high voltage portion if I read things right).

This strangeness is probably - in all likelyhood - due to regional differences (east vs west schematic "standards"), plus the fact that we are dealing with power supplies, which are always messy and strange for non-trivial implementations.

Also something to keep in mind here - for the 60 volt power supply that powers this high-current supply, the AC input is 230 VAC, likely at 50 Hz (ie - European AC power standard). The circuit as-is likely won't work or will work poorly, without being redesigned for the US standard wall-power (120 VAC, 60 Hz) - or even if you use the higher US standard (220-240 VAC, 60 Hz).

Just something to keep in mind - that basically the initial input stage to the first power supply will need to be altered (and perhaps that weird low-voltage supply tap, too); everything after that will probably be OK, since it is DC after that point (albeit 300+ volt DC - so be careful!).

Having created for my personal use low voltage/high amperage transformers myself to use for house repairs when need it and already knowing the danger of using too much voltage with hundreds of amperes (where the total power would rise well above 1.5kW) I wonder about this circuitry long term usability. I know from personal experience that melting metal using such a transformer is generating a lot of heat that goes back to your transformer isolation. And my transformer would be 1.75 V in secondary or as high I ever went was 2.25 volts. So reading this diagram that has hundreds of amperes at 12V seems highly unlikely. That would be over 2kW easily and using this more then 1 minute would definitely destroy the transformer at least. I use my transformers for maximum 5 minutes then I let them cool for 30 minutes on the cold concrete.

This appears to just be a 60v->12v isolated converter at ~2kW. I've never seen a push-pull topology used for this power level since it makes such inefficient use of the transformer and FETs (note it uses 4 parallel FETs for each branch of the primary). Looks fun though

Interesting design, though I think if isolation isn't needed, you would do much better with just a buck topology - half the fets, and no output diodes. Or a surplus supply as other comments mention.

Damn, he is doing some crazy stuff - take a look at the x-ray experiments: http://danyk.cz/rtg_en.html

Some of those videos seem a mite unsafe. What's the purpose of surrounding a white-hot piece of metal with highly flammable paper? It combusted on two occasions that we see. The he just removes the paper, and you realize that all that radiant heat is now hitting the worktop and circuitry...

> What's the purpose of surrounding a white-hot piece of metal with highly flammable paper?

He likely wanted a nice white background for his video. But, well, it backfired.

Reply on news.ycombinator.com