Transformerless Power Supply 230VAC to 5VDC

[2X]

Hi,
do you know how "officially" is called the circuit on the picture? I would like to find information’s about how it works and how unsafe it is because it doesn't have transformer.

[tunk]

Capacitive dropper:
https://en.wikipedia.org/wiki/Capacitive_power_supply

Edit: There's also this recent EEVblog video: https://www.youtu.be/vJ4pW6LKJWU

[Benta]

It's extremely unsafe and should only be used on fully isolated devices that can not be touched by anyone in any way.
Apart from that, some of the component values are odd.

[Zero999]

The output voltage will be more like 4.5V, with any load, because D1 will drop 0.6V. C2 is huge and will take ages to charge up.

[Le_Bassiste]

Hi,
do you know how "officially" is called the circuit on the picture? I would like to find information’s about how it works and how unsafe it is because it doesn't have transformer.

don't know where you got that schematic from, but i find the PE symbol itself and its location in the diagram extremely disconcerting.

[2X]

Hi,
do you know how "officially" is called the circuit on the picture? I would like to find information’s about how it works and how unsafe it is because it doesn't have transformer.

don't know where you got that schematic from, but i find the PE symbol itself and its location in the diagram extremely disconcerting.

I have found from a fan control schematic. Why you say that "extremely disconcerting" about the PE symbol?

[Benta]

don't know where you got that schematic from, but i find the PE symbol itself and its location in the diagram extremely disconcerting.

That's not a PE symbol, it's just some ground symbol.
The only PE symbol that's standardized world wide is this (the circle is important):

[floobydust]

OP that schematic is garbage - unsafe in design as well as architecture. Wherever you got it from, shows a level of dumbness involving mains.
The fuse should come first, before C3 and be in the Line terminal, not neutral. Using Neutral as circuit common is simply wrong, 400V caps are insufficient for mains. The zener has a very short life. I could go on but burning down a house, electrocution, circuit failure are all part of the massive savings from not using a transformer.

[themadhippy]

and be in the Line terminal, not neutral. Using Neutral as circuit common is simply wrong

Whilst i agree ,throughout  a big part of the world  the the mains polarity's going to depend on what way you shove the plug in

[floobydust]

Assuming OP's project is plugged in, North America electrical codes requires a polarized line cord and fuse only in the Hot/Line. A fuse in Neutral gets a fail.
EE's need to repeat my mantra "the fuse comes first" for less fire.

I don't think beginners should have their mains-powered circuit designs up for discussion, other forums forbid threads
"Safety & Closing of Threads. While everyone is responsible for their own safety, the moderating staff, on a case by case basis, will close discussions if they get the impression that one of the members appears unwilling or unable to appreciate the risks in what they are attempting."

[Mechatrommer]

...I could go on but burning down a house, electrocution, circuit failure are all part of the massive savings from not using a transformer.

while i agree with your other points, capacitive dropper is accepted in engineering world as a cheap workaround when safety and economy are met... https://ww1.microchip.com/downloads/en/appnotes/00954a.pdf

[floobydust]

I'm not saying a propery engineered capacitive dropper is unsafe- it's the noobs trying to design one and not knowing the safety aspects.
That Microchip paper author with unknown creds and no actually putting the design through safety approvals, is misleading. It does not deal with the very high peak ripple currents which are hard on components and since these droppers are all about being cheap, the parts on mains age badly. Average power dissipation looks low but peaks are too high for the part. Just do a simple Spice sim showing the peak currents. Do a Spice sim powering it up when mains is at peak voltage. People wonder why the film cap/electrolytic fail in short time.
The part where you test by shorting out components is not done by all the speculators here. One can think it's safe but it's untested, making that rubbish confidence.
I did find one unsafe dropper in a smoke alarm. 120VAC, 100R 1W, 2.2uF 250V cap, grossly oversized because it needed power for the siren/relay which is usually not running.
https://www.eevblog.com/forum/blog/eevblog-1284-how-bad-product-design-kills-the-environment/msg2917546/#msg2917546
Another dropper, the zener open-circuited and electrolytic capacitor exploded, stunk up the house.

[Mechatrommer]

That Microchip paper author with unknown creds and no actually putting the design through safety approvals, is misleading.

https://patents.justia.com/inventor/reston-a-condit
https://www.linkedin.com/in/stan-d-souza-14942a5
how do i find your creds?

I did find one unsafe dropper in a smoke alarm. 120VAC, 100R 1W, 2.2uF 250V cap, grossly oversized because it needed power for the siren/relay which is usually not running.
https://www.eevblog.com/forum/blog/eevblog-1284-how-bad-product-design-kills-the-environment/msg2917546/#msg2917546

so is it designed by canadian noobs not knowing the safety aspects? who happened by luck to get ULC cert?

Another dropper, the zener open-circuited and electrolytic capacitor exploded, stunk up the house.

not burnt house right? but i agree, bad enclosure...

i can always find the same issues with transformered unit if i want to go on further...

[floobydust]

That Microchip paper author with unknown creds and no actually putting the design through safety approvals, is misleading.

https://patents.justia.com/inventor/reston-a-condit
https://www.linkedin.com/in/stan-d-souza-14942a5
how do i find your creds?

I did find one unsafe dropper in a smoke alarm. 120VAC, 100R 1W, 2.2uF 250V cap, grossly oversized because it needed power for the siren/relay which is usually not running.
https://www.eevblog.com/forum/blog/eevblog-1284-how-bad-product-design-kills-the-environment/msg2917546/#msg2917546

so is it designed by canadian noobs not knowing the safety aspects? who happened by luck to get ULC cert?

Another dropper, the zener open-circuited and electrolytic capacitor exploded, stunk up the house.

not burnt house right? but i agree, bad enclosure...

i can always find the same issues with transformered unit if i want to go on further...

The Microchip guy is on many patents but it doesn't mean he's smart or qualified for mains safety IMHO. First mistake is assuming dropper currents are sinusoidal, ooops they aren't if you've ever poked around and done your homework. So his math is no good.
I'm a P.Eng. for a few decades now and put several products through approvals with UL/CSA/Intertek, no need to advertise.
I dislike the mistakes made in forums and white papers. Examples, the input surge resistor being "fusible" or "flameproof" EE's are clueless about what specs the part needs to have and if it will prevent grandma's house from burning down. The film capacitor must be a special part for long life.  Mains transients are also an issue for the designs.

Common business is to import cheap chinese mains-powered devices (which use a capacitive dropper inside) like smoke alarms, bathroom fans, night lights etc. and get them approved and resell them for great profit.
UL217 smoke alarm standard is only concerned about them working (detect smoke, particle size etc.) regardless of electrical safety such as how hot the components are running 24/7 or if there is a fuse, or even if they false trigger. Latest Rev. 8 might deal with that lol smoke alarm makes it's own smoke and false triggers.... "that's fine". It was painful talking to them about it, but many safety standards have huge holes in them and importers dodge things using loopholes.

[Benta]

I have a very ambivalent attitude to supplier application notes.
It's about knowing the engineers behind them. There are real big shots that put out brilliant stuff. They're unfortunately scarce today.

I've worked in/with the semiconductor industry for 40+ years, and know how things run. Writing application notes is a pain, delegated to the newest angineer in the department. The exception being the absolutely newest, ground-breaking devices which will be written or co-written with the design engineers.

The rest are left to engineering interns during the summer vacation, where they can earn some points, and with luck a few dollars. The supervisors are on vacation.
Review of the papers? The supervisors have better things to do after the vacation.

If you think that a stamp of TI, ON, Analog, LT,  etc. on an application note means that it's "university quality" and the eternal truth: think again.

Trust your own engineering skills.

[floobydust]

This kind of crap is what makes me skeptical of half-assed white papers.
Same "big brained" author earlier Microchip TB008: "As mentioned earlier, the neutral should be connected to earth ground through a fuse. This would insure protection in case of improper wiring." 

edit: this already hit the fan a long time ago https://www.electronicspoint.com/forums/threads/microchips-tb008-reprise.114717/ yet OP posted the same idiot fuse/ground placement.
edit 2: Rod with Elliot Sound already sounded the horn https://sound-au.com/articles/power-supplies2.htm see section 7 Cheap Death ala Stan D'Souza.

"Rod,

We apprecaite your saftety concern and taking the time to contact us.

I've looked into the design quite carefully, and there is no issue with the design.  The purpose is to cover the accidental case of plugging in the plug backwards (unlikely since most plugs are no polarized), and to cover the common case of a miswired outlet of a swapped hot and neutral.  If these situations occurred with this design, the neutral would have power on it (due to the swap) and would be immediately grounded.

The result would be a rapidly blown fuse (the one connecting to neutral in the design) and thus provides a safety to prevent the neutral line from having hot on it.  The hot line would actually then be connected to neutral, thus the design would have no power, and no return, and no hot connection in this case.

I'd agree that since it is a high voltage design, a warning would be a good idea if only from a legal perspective.  However, this was not a design presented in some magazine to the public, but is an engineering document.  It is assumed that engineers would be viewing it and would take normal precautions when working with AC power circuits.  However, we should not make that assuption, so I will request a warning be added regarding working with AC power.

The design otherwise does not contain a flaw.

Also, if the power is connected correctly (neutral to neutral, etc), then the connection of neutral to earth should be harmless since they are supposed to be at the same potential.  If they are not, it suggests a wiring fault, or other issue.  I'd argue that perhaps a small resistor should be added to deal with ground being a few volts above and below neutral and not blow the fuse - this situation can happen when heavy loads are placed on the power lines, and the voltage drop it causes can cause a small difference in voltage between grounds of outlets on different circuits, or a difference in neutral and ground potentials.

If after this discussion you still feel the circuit presents a hazard, please detail how this is the case - where the current would flow, what conditions, etc.  It would also be helpful to indicate where this violates code.  I do believe that it is not permitted in house wiring, but there is nothing I am aware of that says it can not be done in a design.  Also note the lines are not directly connected, but are fused.  However, the worst case even for a non fused link, would be to trip the circuit breaker on the house, which would detect the excessive ground current (due to the live being connected to ground).

If the resolution provided does not solve your problem, you may respond back to the support team through the web interface at support.microchip.com.  Telephone support is also available Monday-Thursday between the hours of 8:00am and 4:00pm MST and on Friday between 10:00am and 4:00pm."