Transformerless AC DC Power Supply

[Paradigm]

Dear All Expert,

I am currently working on an electronic project. My electronic project was ran by DC power supply of 5V and 3.3V using a maximum of 100mA.

Currently, the AC to DC power supply was using transformer stepping down from 230V to 5VDC + rectifier + Capacitor + voltage regulator.

The problem I am facing now was the form factor of my power supply due to the big transformer. I had a space constraint for the PCB design. I got only 65mm by 65mm by 30mm of space to include all my power supply and the electronic circuit.

I had search the internet for "Transformerless Power Supply" and found some design that had no transformer. To my surprise, the circuit is far more simple then I thought.

My Question now was. Can this circuit be used in my electronic circuit to power up my ICs chip? Is there any safety concern? By connecting the rectifier onto the L and N, does this means my rectifier will be big in size also? Will I blown up my circuit should there by any parts failure? Will the L and N into the PCB create any unwanted noise to the PCB thereby interferring with my electronic circuit with had a RF transceiver in it.

I am surprise that Iphone charger and some model of the phone charger can be as small as a plug. Any expert can share how can it be done? Taking safety into consideration.

I need a 230VAC to DC power supply of 5V/3.3V maximum current of 200mA that can be safely integrated onto my electronic PCB with L/N in.

Thanks a lot for any advise.

 

[keith1200rs]

You cannot use a transformerless power supply if there is any possibility of someone touching anything electrically connected to the circuit - you will kill them.

Tiny mobile phone chargers are small switched mode power supplies. Look at powerint.com for some designs and chips.

Keith

 

[BradtheRad]

Wise counsel in post #2.

To drop mains voltage using a capacitor, is a technique that makes sense with low-cost items such as led's.

Nevertheless it seems to be getting more popular, despite the risk of exposing high voltage to valuable devices.

It's up to you whether it's worth the risk.

If you need 200mA in a small space, you should start with smallish value components, then increase where necessary.

Example, the mains-dropping capacitor C1 may or may not need to be greater.

You can start out with a diode bridge at 1A per each, and bump up to higher current if you feel them getting hot.

The zener diode will protect your project against overvoltage in normal operation. 1W is okay for the zener's watt rating.

R2 is to limit inrush current at startup. My simulator shows R2 should be 2W in normal operation. If C1 or bridge diodes blow, then increase R2 resistance.

My simulator shows that smoothing capacitor C2 ought to be rated 200V rather than 100V (in case that were ever to be the only component attached to the diode bridge).

There's a chance you can put R3 on the left side of C2. Experiment to see which gives you less ripple. Or consider adding a second smoothing capacitor to the right of R3.

I am not responsible, etc.

 

[keith1200rs]

Just to clarify, my concern with the transformerless power supply is about killing people not devices!

Keith

 

[Paradigm]

Hi Keith, I refer to the website you stated and found the exact same small size iphone charger that I had on their design
The problem is after referring to the PDF file, they do use transformer also. But this transformer is a miniature transformer.
The PDF teach use how to wind the transformer ourselves to get the required spec. However, I cannot get the material need to wind the transformer.
Do you know of any other website or shop that sell the transformer that was needed by the project?

BTW, how big must my transformer be if I am using only 200mA of current 5VDC? How can I size my transformer correctly? Thanks a lot

 

[keith1200rs]

If you download the software from Power Integration web site it will do the design for you. Also, they have a transformer prototyping service where they will make a few transformers based on the design from their software (for a charge).

Keith

 

[Audiop]

Why is the 400VAC chosen for C1 in View attachment 68889?
Is it chosen to provide a conservative margin for voltage spikes above 230VAC in the absence of any other spike protection or is it something more specific?
Would an 275VAC X2 capacitor be sufficient in general usage up to 240VAC?

 

[mtwieg]

If you really want a transformerless power supply, then you need to make sure that it's acceptable to not have isolation. In most cases you do need isolation, so I wouldn't bet on it. Also, designs using zeners and voltage dropping capacitors like the one you're showing often require very large capacitors, so you may find it doesn't help much in terms of space. Also They suffer poor regulation and efficiency. Before you try such a method I strongly recommend you try simulating the circuit to see what I'm talking about.

For small offline converters, good transformer design and high frequency operation are critical. Also they often don't use actual control ICs, and are self-oscillating, which reduces the external component count. Finding the right transformer off the shelf is probably the biggest challenge for you.

 

[BradtheRad]

Why is the 400VAC chosen for C1 in View attachment 68889?
Is it chosen to provide a conservative margin for voltage spikes above 230VAC in the absence of any other spike protection or is it something more specific?
Would an 275VAC X2 capacitor be sufficient in general usage up to 240VAC?

Theoretically the capacitor can charge to 1.4 times nominal ACV (with a light load).

So for 240 incoming AC, the capacitor charge can reach +335 or -335 V.

 

[nagabhargavamrita]

there will not be much problem but i tried a circuit without zener which helped tht can change the voltage accordingly by placing a capacitor but the problem occured was if u place less voltage then the designed oneafter some time the capacitor gets saturated which leads to the damage of the device

 

[Audiop]

Theoretically the capacitor can charge to 1.4 times nominal ACV (with a light load).

So for 240 incoming AC, the capacitor charge can reach +335 or -335 V.

Belated thanks BradtheRad!

I understand that Vpeak = 1.4 x Vrms. By the same token, I thought a 400VAC capacitor is rated at 400VRMS, that is, 560Vpeak.

I would have thought that a 275VAC capacitor (which is 388Vpeak) would be sufficient for a 230VAC supply (which is 324Vpeak). In fact I thought that 275VAC X2 capacitors are specifically designed to be connected across up to 240VAC line to neutral.

Please help if I am misunderstanding something!

 

[BradtheRad]

I understand that Vpeak = 1.4 x Vrms. By the same token, I thought a 400VAC capacitor is rated at 400VRMS, that is, 560Vpeak.

I would have thought that a 275VAC capacitor (which is 388Vpeak) would be sufficient for a 230VAC supply (which is 324Vpeak). In fact I thought that 275VAC X2 capacitors are specifically designed to be connected across up to 240VAC line to neutral.

Please help if I am misunderstanding something!

Yes a capacitor will often have WVDC printed on it. 'Working volts' is an average figure, as your math expresses.

And if you wish you can use a capacitor rated for the exact working level you expect it will be exposed to. Nothing says it won't handle it in normal use.

However with a light load, the volt level can soar to the incoming peak. This is 1.4 x working level.

Furthermore things can change over time. A capacitor can deteriorate. House power can surge high every so often. And these days even a new capacitor can have an overly high 'claimed' rating if it was manufactured in a certain far East country.

For these reasons it is common advice to use a rating of twice what you anticipate will be the maximum applied volt level.

 

[dselec]

ןf u insist killing your self and/or others least reduce the voltage using half wave rectification..
This way u may be half dead:!::twisted::roll:

https://ww1.microchip.com/downloads/en/appnotes/00954A.pdf
CHECKOUT PAGE 10

 

[samEEEf]

As these power supply doesn't have isolation, there is a chance of burning the load circuit if short circuit fault occurs in any component (capacitor or zener). This is dangerous for the load circuit. If there is MCU and other sensitive components those will definitely burnt out with the full input line voltage.

 

[Orson Cart]

... at least reduce the voltage using half wave rectification..

Just remember most developed countries have a limit of 5mA average DC draw from the mains supply (so the big supply transformers don't edge towards saturation with lots of crazy half wave rectified loads attached)

In any case you need a back diode to complete the charge/discharge cycle of the cap in series with the mains.

Regards, Orson.

 

[tpetar]

I avoid to use transformeless psu, usually I use small switcher or small quality iron transformer removed from new cheap adapter.


See several examples:







:wink:

 

[samEEEf]

For the topology shown in Microchip's App note is very simple and cost effective for low power transformerless power supply. One can ask - why I need such power supply? I can use SMPS or small iron core transformer instead.
But my point is if I need a power supply for a small MCU circuit required only about 50-100mA at 5v DC does SMPS or small iron core transformer really required or efficient?

 

[tpetar]

For the topology shown in Microchip's App note is very simple and cost effective for low power transformerless power supply. One can ask - why I need such power supply? I can use SMPS or small iron core transformer instead.
But my point is if I need a power supply for a small MCU circuit required only about 50-100mA at 5v DC does SMPS or small iron core transformer really required or efficient?

Nothing can be comparable with risk to get life in dangerous situation, as electroshock from mains power. For me device should be safe for users, also to be safe for unattended working on long time. I dont say that transformeless is bad and that is wrong. Transformerless can be good in some light variants, you mention one, for me I rather use safe and stable power supply voltage from earlier mentioned ways, I will keep that principe in future projects also.



I suggest reading for topics about mains noise problems, filtering, galvanic isolation, purpose of transformers,...

:wink:

- - - Updated - - -

I dont even want to imagine what will happen in moments of lightning strikes. :-?

 

[samEEEf]

What do you suggest for such small amount of power requirement (50-100mA current from 5v dc)?

 

[tpetar]

What do you suggest for such small amount of power requirement (50-100mA current from 5v dc)?

Depends from case inside space and case dimensions and from design of device. Can be small switcher, if you dont want to make yours, buy cheap mobile adapter 5V remove switcher PCB and put additional 100nF to lower spikes, and go on. If there is space use small iron transformer.