Transformerless Power Supply of 1A at 5Volts!

[Ioannis]

The topic of P.S. direcly from the mains is very dangerous and I do not recommed to do so.

But there are cases that this is unavoidable.

Current designs have the dis-advatage that are of low current.

The attached one is a clever design from Elektor and gives about 1A at 5Volts.

I repeat that all parts are live and should never be touched!

Ioannis

[rsocor01]

Ioannis,

Unless I am missing something here, K1 in your schematics seems to be a transformer. The datasheet for the LM2595 gives a maximum input voltage for this chip of 40V. If K1 is not a transformer, it looks like you would fry this chip with a power supply input of 230Vac.

Robert

[Ioannis]

Robert.

K1 is a connector!

As for the 2595, yes you are right on the max input voltage. BUT, look carefully at the schematic.

The 2595 will get a max of 40 volts and nothing more than this. Zener, T1 and T2 will make sure that it stays in safety region.

Thats why I stated that the circuit is clever.

Ioannis

[rsocor01]

Ioannis,

Yes, you are right about T1 and T2 keeping the voltage low on the LM2595 chip. However, I am still not convinced with this circuit. After looking at the datasheet for T2, it seems to me like the Drain and the Source inputs should be swapped for a proper grounding on LM2595. The current should be flowing away from the ground in LM2595, not towards it.

Also, the maximum VGS voltage on T2 is 20V, so if for some reason T1 fails T2 would also be fried. I don't know why Elektor designed this circuit without a transformer, and I agree with you that this is very dangerous.

Robert

[Ioannis]

Technically the circuit is correct. No need for pin swaping. It is OK as it is.

At first I know it is a bit confusing but trust me it is very clever trick.

The purpose of zener and the 2 transistors is to let the voltage raise up to say 40 volts on the capacitor and 2595 input. Then T1 is biased and switches off T2 so that the raising voltage cannot hurt the 2595.

This is repeated again on the next half period.

As for the dangerous this refers to the live voltages all over the circuit and the circuits connected to this. Other than this I consider it "safe" in a matter of speaking to use it.

Ioannis

[mtripoli]

The Microchip website has examples of running micros off the AC line... search the app notes...

[Ioannis]

Not at 1A though...

Ioannis


The Microchip website has examples of running micros off the AC line... search the app notes...

[Megahertz]

I know some places have extreme voltage drop. Will this circuit work if AC voltage drops to lets say 160V from 230V & also if it rises to Lets say 300V from 230 V? Thanks

[sayzer]

I know some places have extreme voltage drop. Will this circuit work if AC voltage drops to lets say 160V from 230V & also if it rises to Lets say 300V from 230 V? Thanks

In what place of earth can municipal power lines go from 230V to 300V?

I can understand 240 or even at max 260, but 300? Such a case would end up for compensation in court.

-------------------------

[Ioannis]

The circuit will work from say 10 Volts AC up to whatever the rated voltage of the componets used is, mainly the bridge.

So the answer is yes.

Ioannis

[Archangel]

In what place of earth can municipal power lines go from 230V to 300V?

I can understand 240 or even at max 260, but 300? Such a case would end up for compensation in court.

-------------------------

Hi Sayzer,
Probably depends upon how you measure it. If you use an a/c voltmeter then it is calibrated in RMS and gives accurate results. If he is using a D/C voltmeter and a diode, then he is seeing peak voltage - diode drop, and will see an inaccurately high value.

[Megahertz]

The country in question is India. When I was there, I actually saw the voltage drop upto 170V and during power faliure when the backup generators were fired up, voltage at few places seem to be as high as 280-300V and so I asked the question just to have the knowledge :)
Thanks

[Ioannis]

The circuit of Elektor will NOT work when input voltage is rising above 39 or so volts. So it will work perfectly well if input AC voltage is just up to 110. No need to change anything, or almost anything. May be just the R5 changed to a lower value, say 15 ohms.

The circuit is very clever and the SMPS is disabled by the pass transistor wheninput voltage is greater than 39 so volts. Beware it is not isolated.

Ioannis

[roopsvinay]

Did anyone try to build this circuit? Can this work at 230v AC

[Ioannis]

It is mainly for 230 VAC

Ioannis

[shottootto]

It is not AC but rectified AC to result in DC. Also note the Zener that drops the voltage.

[Ioannis]

Please review the circuit more carefully. The circuit works from zero up to about 39 volts. After that point it cuts off.

It works very good from 230 VAC.

Ioannis

[jmgelba]

Be interesting to see what happens from 90Vac to 304Vac.

[Demon]

Agreed, I'd be interested in 110VAC range.

Robert

Edit: Did you guys check Melanie's design?

http://www.picbasic.co.uk/forum/showthread.php?t=3265&p=17531#post17531

[Ioannis]

Be interesting to see what happens from 90Vac to 304Vac.

As I said it does not do anything above 39 or so volts. So where is the problem?

After the sine wave has reached the 38-39 volts the circuits is not powered any more from the power line.

Ioannis

[elipee]

Very interesting circuit, however, a think it is not going to work very well, look at T2 ( BUZ41A or IRF730) , both have a maximum 20V VGS, so when voltage get bigger than 20 , T2 gate get damage. Maybe it work for a time, but T2 VGS is out of specification.

[Ioannis]

It won't get that high for long time to destroy the Mosfet. Also the Zener is not a switch but it does start to operate at lower that rated voltage.

It does work with great success and efficiency.

If you are concerned about Vgs, there are other Mosfets (sure more expensive) to replace this one.

Ioannis

[Colin55]

"voltage at few places seem to be as high as 280-300V"

No so.

Do you realise the current taken by a globe (and lots of other devices) will increase considerably as the voltage increases. When the voltage rises by say 10%, the current will rise by much more than 10% and this will put a very heavy burden on the generator and in most cases the generator will not be able to supply the load. This is why the voltage will not rise as you have stated.
Globes and radiators are not linear devices, but if they were, a 1,000 watt radiator at 250v would take 1250 watts at 280v. This is 25% increase in wattage for 13% increase in voltage.
The figures would be much higher for 230v supply.