Convert 600VAC (?) to low voltage DC (5 to 12V)

Hey guys, good to be back. It's been way too long.

I'm dealing with two distinct situations, high-voltage AC and DC, but both with the same predicament - converting to low voltage DC.

Even though I have access to three-phase power in some situations, I don't have a neutral to generate the commonly used 230VAC, where I could just plug a household AC-DC adapter.

In the case of solar panels, I just have DC current with a great deal of voltage variation (0-780VDC).

The purpose is to power a small system from that. 5VDC 1A would suffice (Arduino, 16x2 LCD, wireless transceiver, ultrasound sensor, some LEDs).

Now, I know about the everyday DC-DC buck converters and regulators, but even in those cases I only know how to deal with up to 40VDC.

Some of you guys probably know how to do this - and probably I don't know the correct words to search for.

Hope you can help. Thanks in advance.

Both of these voltages are very dangerous and I would not be wanting to use them without good isolation.

Isolating the DC would require a switch mode with isolated output but getting one to work over that voltage range would be difficult.

A single phase 230/240v system has up to around 440v between phases. You may be able to get a transformer to bring this down to 250 or your working voltage.

Weedpharma

I’ll take one for the team and say what I’m sure many members will be thinking… leave this one to the pros.

600V DC solar is awesomely dangerous, far more so than AC mains. It contacts all your
muscles so you cannot let go, so death is pretty much guaranteed (solar panels output
several amps, you are talking electric chair conditions...)

It will also hold an arc of 1cm or more at full current, so conventional switches will not
guarantee to switch it off...

DO NOT DO THIS! Use a small separate solar panel instead.

Thanks all for your input.

I understand the risks and thank you for your concerns, but I'm analyzing this academically, I'd like to grasp the concept of this - how does one turn higher voltage into lower voltage in these special scenarios, especially with DC current.

I'm not looking for a "try this and see how it goes" :smiling_imp:
Of course high voltage is for the pros. I'm sorry for not mentioning all this when I started the topic.

In the case of the solar panels, one possible answer is: get a solar inverter and use its output.
Another possibility: see if you can rearrange the panels less in series and more in parallel, to obtain more current with less voltage.

My question was with regard to your knowledge and experience, essentially how this could be done in a number of ways...

In the case of three-phase AC (around 400V between any two phases), yes a two-phase transformer (?) may be a simple solution, thank you @weedpharma for the suggestion. I think you're referring to those simple ones, with a laminated core.

--- Rest assured guys, I'm not going to try any of this on my own.

Let's take for instance the case of VFDs. They are purely DC inside, correct? So they take in three-phase 400V AC and turn it into what? 600V DC? Even so, they have 10V and 24VDC outputs.

MarkT:
DO NOT DO THIS! Use a small separate solar panel instead.

This is also a workaround, of course. Thank you

For the AC, you can use the same scheme that some people use to steal power from your local power company. They stretch a wire parallel to the power company line and pick up the induced AC at each end of the wire. You can do the same by loosely winding an insulated wire around one of the high voltage AC conductors. This becomes a current transformer and will induce power into your wire based on the amount of power on the high voltage wire.

You will have to process the induced voltage on your wire and regulate it to give you your needed 5 volts. Vary the turns around the HV conductor till you get what you need.

This will not work for your DC options.

Paul

Paul_KD7HB:
(...) This becomes a current transformer and will induce power into your wire based on the amount of power on the high voltage wire.
(...) Vary the turns around the HV conductor till you get what you need.

That will depend on the actual power flowing through the high voltage wire, correct? Just like a hall effect current sensor?
Interesting idea. Although I can see why it wouldn't work all the time. If I'm correct, you'd need a significant load on the main wire at all times. This suggests using a backup battery as the main source for the low power application, and charging it through that "stealing" process. Not sure how much current I'd be able to draw on a local installation. That's something simple enough to experiment with on the three-phase scenario. With rectifying diodes and a DC-DC buck-boost converter maybe.

If I were to wind a wire around each phase, to gather power from all three, would it be best to implement a three-phase rectifier instead of connecting them in series? (since the current waves have different phasing, they would interfere with each other...?)

footswitch:
That will depend on the actual power flowing through the high voltage wire, correct? Just like a hall effect current sensor?
Interesting idea. Although I can see why it wouldn't work all the time. If I'm correct, you'd need a significant load on the main wire at all times. This suggests using a backup battery as the main source for the low power application, and charging it through that "stealing" process. Not sure how much current I'd be able to draw on a local installation. That's something simple enough to experiment with on the three-phase scenario. With rectifying diodes and a DC-DC buck-boost converter maybe.

If I were to wind a wire around each phase, to gather power from all three, would it be best to implement a three-phase rectifier instead of connecting them in series? (since the current waves have different phasing, they would interfere with each other...?)

You understand exactly what I am suggesting. I guess I would see what you can get from a single phase before adding to the complexity. Probably be easier to just wind more wire around one feed.

Put a 1,000 ohm resistor across your winding and watch it for awhile with a VOM to see what you have to work with.

Paul

Make sure you do this on YOUR side of the power meter! :wink: :wink:

Lol
Discussing ways of actually stealing electricity is off-topic, don’t you agree?

I’ll read up and experiment with current induction over the weekend.
Don’t remember the math of this, just that it’s a pain to work with (low efficiency).

Regarding the solar panels
I’ve been reading about DC/DC step-down power converters. And it seems that rearranging the panels would be the way to go.
The “low” power converters (<150W) that I found supported up to 320V DC.

I’ll be talking with a specialist later this week, and anything new I’ll report here.

Thanks

footswitch:
I'll read up and experiment with current induction over the weekend.

You ought to try get out more :wink:

JimboZA:
You ought to try get out more :wink:

You may have a point there!

News on the solar panels.
It seems like MarkT wins the internets, since he suggested a dedicated panel for the electronics.
Most solar inverter technologies use Maximum Power Point Tracking (MPPT), or similar, which takes as much juice as the panels can give. They're constantly trying to get more juice.
Connecting anything else to the panels, would interfere with the MPPT, and several things can happen, from low performance to failures/self shutdown. I believe the only thing to not cause this would be a purely resistive load.

A dedicated 50W solar panel + MPPT + charging regulator, should cost about 80 euros, maybe less. Plus a standard UPS battery, 12V 7Ah, costs about 15 euros. So for under 100 euros one may design such a solution (where it's expected to have a constant 8W load, which is 5W + power conversion losses). This isn't a precise calculation, it's just to give the general picture.