# Power Supply current limiting ???

Discussion in 'Power Electronics' started by Faraday, Jan 31, 2012.

7
0
Jan 31, 2012
Hi,? I'm extremely rusty with my electronics so please be patient. I'm trying to create a high v power supply for my friend ... the load of which can vary from 0.1 ohms to 4 ohms !

I've got the basic bridge rectifier and filter concept (see attachment) , but I'm struggling with getting the right values to produce a smooth voltage when under such a huge load.

Also I want to limit the DC output so that as the current starts to get up to 13 A it gradually shuts down the voltage and current as shown in the attachment.

The challenges are

Removing ripple with high DC voltage (320 V) and low load
ensuring the input AC current does not rise above 20A
The extremely low load (0.1 - 40 ohms)
The current limiting in such a way that the AC current is also limited I want the voltage and current to gradually fade

Any help or suggestions would be greatly appreciated.

#### Attached Files:

• ###### PowerProject.png
File size:
45.8 KB
Views:
333

7
0
Jan 31, 2012
I've been looking at thermistors so far, but they don't seem to do one up to 240Vrms AC and 20 A, Do thermistors just behave like resistors ? i.e. could a wire a bunch in parallel to reduce the current through each ?

Last edited: Jan 31, 2012
3. ### jackorocko

1,284
1
Apr 4, 2010
Reduce the ripple by increasing your capacitance of the capacitor bank. But be extremely careful, a capacitor bank charged to ~400V is gonna be capable of killing you and your friend in spectacular fashion, Pretty dangerous what you are attempting.

4. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,489
2,830
Jan 21, 2010
Firstly, directly connecting to the mains has got to be a bad (read DANGEROUS) idea.

I have to ask -- what are you powering?

You will need an active circuit (i.e. no thermistors) and one which involves switching (to minimise power losses). The ideal thing to do is to place what is essentially a fast electronic switch between the top of the capacitor and the choke. You then need to add another diode and capacitor, and then add further circuitry to switch the transistor on and off rapidly, and to alter this with the load current.

It's all very non-trivial considering the high voltage and current.

edit: this is in addition to what Jackorocko suggested.

7
0
Jan 31, 2012
Firstly thanks for the replies. My friend is in a remote area with no mains and no money. His shack is all 12 volt lighting which is controlled by some box of tricks that's designed to run from either a large battery bank or solar panels....trouble is he doesn't have the batteries , and when the sun goes down hes in the dark. So he's bought a petrol generator which actually outputs 110 Vrms (not 240 as per my drawing) and he wants to connect that to the box of tricks when the sun goes down.

This transistor idea sounds interesting ....so I want the collector tied to the top of the capacitor and the emitter to the choke....then output a square wave to the base which when the current is too high will only be high for 1% of the cycle but when the current is acceptable will be high for 99% of the cycle ??? is that the idea ?

6. ### duke37

5,364
771
Jan 9, 2011
Just install a 110V light and keep the systems separate. No chance then of blowing up the control box.

7
0
Jan 31, 2012
That'd be far too simple It would involve quite a bit of rewiring , fitting a consumer unit etc.... besides it's just no fun, not often I get the chance to tinker with my soldering iron so I'm determined to build this thing.

So as the load increases I want to detect when the current is approaching 13A at which point I need Vbe to drop and switch off the transistor....which would immediately cause Vbe to rise again and turn itself back on ?? Anyone got an example circuit for such a device ?

8. ### jackorocko

1,284
1
Apr 4, 2010
I think I would have invested in a battery bank instead of a fuel powered generator.

I can see the satisfaction in wanting to build something that is hard and complicated. But electricity shouldn't be hard and complicated, just more things to break or go wrong. As steve has already said with such high voltages and currents parts will not be cheap for someone with no money.

Also, why such high voltages? there is no way that the solar panels are putting out that much voltage to the control box. Likely less then 48V is my guess.

7
0
Jan 31, 2012
[QUOTE
Also, why such high voltages? there is no way that the solar panels are putting out that much voltage to the control box. Likely less then 48V is my guess.[/QUOTE]

The solar panels are putting out 200V 0.1 - 5A , and he's only got 4 of them.

10. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,489
2,830
Jan 21, 2010
Oh, I don't know... Mine put out 440V at around 4A on each string...

But I'd want to know more about the solar panel setup. Mine max out at 5kW and there's a *lot* of panels. Given the requirements here seem to be for similar power I wonder how the panels are used -- especially without batteries.

11. ### jackorocko

1,284
1
Apr 4, 2010
Don't solar panels work like wind generators in that the voltage is limited to the clamped voltage, like say a battery bank?

7
0
Jan 31, 2012
I'll tell you as much as I know, he has 4 panels that put out a fairly steady 50 v , but the current varies up to 5 amps .... so total 1 kW maximum. He has a box that I guess splits the power from the panels to the individual circuits and a few old car batteries. Now I'd have thought that'd be plenty but the car batteries are apparently no good as they aren't deep charge batteries. He can't afford to buy these deep charge batteries as they're extremely expensive so he's now bought an old 110 v generator and wants to hook that in to provide his night time power. What's missing is the interface between the jenny and his power box. The only specs I could get from him were the panel specs (50v 5amp max each) so I figured If I could create a circuit to make the jenny output look like the panels output he'd be in business.

If I rectify the output of the jenny theres 160 v DC .... just need to limit the current so it doesn't blow his box to pieces. ... didn't sound too difficult initially but I'm guessing by you guys reactions that it is actually a bit too difficult at this power level.

13. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,489
2,830
Jan 21, 2010
In your case I would tend to recommend a fuse or a circuit breaker.

If I were your friend, I would try to get some batteries to use with his system. They need not be deep cycle batteries as long as he uses them just to even out the bumps and dips in the load vs solar output.

Then I'd use the generator to charge the batteries and keep the night-time load low so that you're not discharging the batteries (or at least not over the amount they're being charged).

If you have batteries you can use a MPPT regulator on your solar panels and gain up to 20% more power from your system. You could also run the whole system at 12V or 24V which makes life easier for external charging etc.

7
0
Jan 31, 2012
Thanks for the advice Steve, I' ll build the circuit and tell him to use it to charge his battery array during the day.

Just out of curiousity, If his solar box has MPPT built into it, would it be safe to connect the circuit directly to his box, as from what I've read the MPPT circuitry would automatically adjust the load to acheive the best results ? I guess it just depends how far it can adjust the load, do they do vary the load by 1 or two ohms , or can they go up into kilohms, megohms ???

15. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,489
2,830
Jan 21, 2010
MPPT regulators typically adjust the current they draw from the solar panels such that the output power is at a maximum.

When charging batteries, the simple way is to operate as a buck switch-mode regulator from a set of panels with a higher nominal voltage than the battery. The regulator adjusts its duty cycle to achieve maximum output current (since voltage is almost constant).

In reality, the charger may also monitor the battery voltage so that the batteries are charged correctly.

The load (the batteries) has the impedance of the batteries, and that is pretty low (well under an ohm -- typically measured in milliohms).

The point of having batteries (even ones which are not in the best condition) is to be able to provide power to the load when the solar output is low and to absorb it when it is high. Without batteries you're limited to a load that is less than the current (and always changing) generation capacity. Either that or you use load shedding to keep it that way.