Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Transformerless power supply with only Live wire

Status
Not open for further replies.
Hi again,


Here are the changes that will help the problems.

1. Short out the fuse where it is now, move it to be in series with the N input line terminal. Make it a 500ma unit if you can but no larger than 1 amp.
2. Change R560 to 330 ohms, at least 2 watts for now. We can probably get away with a 1 watt here.
3. Remove C2, the 394k cap.
4. Make sure the bridge rectifier is rated for at least 1 amp.
5. Disconnect everything from the bridge rect (+) output, except the C5 cap (+) terminal.
6. Keep C5 capacitor (+) terminal connected to bridge rect (+) output, but that's the only thing to connect the the bridge (+) output.
7. Connect a 100 ohm resistor between the capacitor C5 (+) terminal and the zener cathode.
8. Connect the rest of the circuit to the zener cathode (like the 555 etc.).
9. Redraw the circuit and post here.
10. Test.

The above changes will help it work better and also make it safer.

Hi again, after couple of months... I've decided to continue the thread as people are asking me in private messages how the things are going with this.

First of all, apologies to MrAl for the delay... I'll not go into details why, but I'll just mention I didn't stop working on this project when I had some free time.
I still have problems.
Second, I did tested the suggestions from MrAl, also did other tests, as a result - no sufficient power for my circuit and still too much heat generated from the current limiting resistor (and the one which protects the zenner).

The general issues are 2: heat produced (especially when bulb off) and not enough current to power the whole circuit - when the bulb is ON.

So therefore I've spent a lot of time reading arround and found out MrAl is right (actually I knew it before, but I wanted to understand better).

Now I am sure I need to go with the approach to detect the zero crossing and to delay the moment the triac is fired. Hopefully this delay right after the zero crossing would give me some power to feed my curcuit, without dimming the bulb (exactly as MrAl was pointing several times already before). The idea is to have as much as brightness as possible.

So I am now close to put on my breadboard the following:

/Important for new readers here: The whole thing should work without Neutral line! I have only the Live wire and the other one is the wire connected to the bulb, so the whole circuit will be in series with the bulb all the time!/

- H11AA1 will be used (not sure if it will operate properly as the bulb is connected in series) to detect the zero crossing (with 22k resistors infront, 1W at least) and it will trigger a CMOS 555 in monostable mode;
- The 555 will produce a time delay (not sure how much time... maybe 2 ms or so);
- After the delay the 555 pin 3 will go low and thus MOC3052 (random one) will fire the TRAIC gate;
- The Triac will turn off at the end of the sine wave; As it will be again zero crossing, the 555 will not let the MOC driver to turn on the triac until the next delay cycle finish.

Furthermore, I've managed to utilize my own RF receiver on 315Mhz (with the respective decoder IC), and got rid of the relay as MrAl was advising me before. That should save some power. So no relay anymore in this circuit! :) The output signal from the receiver drives a flip-flop CD4013, which is also driven by the manual button switch (debounced with schmitt trigger accordingly). The issue now is that threre is not enough power for the receiver when the bulb is ON. If I power this part of the circuit with battery, everything works as I want...(I don't want battery in this project).

I am going post diagram and results soon when time to do it.
Again apologies for the delayed posts but I am doing all this as a hobby only in the available time.
 
Last edited:
Hi again Ibob,


First, nice to hear from you again and nice to hear about this project again too. I had wondered how it was going myself.

This is one of those circuits that has to be done pretty close to perfect to get it right as there are a couple of things that have to be balanced. The power pulse width has to be long enough, resistance in line with the power flow has to be small, load cant be to heavy, stuff like that.

However, if you ever decide you've had enough of trial and error with the power stealing method (where the triac stays off for a short time in order to store some energy even during the normal 'on' time) then you can always resort to a small current transformer. That would supply some energy when the bulb is on and would not require any super special timing network. The load still cant be too heavy though.

Good luck with it either way.
 
MrAl,
Thank you.

I will keep this idea with the current transformer frozen for now.

How about using two power MOSFETS instead of the triac? That might save some current as the Mosfet is powered with voltage not current? What do you think investigating in this direction?
 
Hi,

I think you should stick with the triac. It should be possible. I could check other dimmers if i can get one somewhere, but im sure they use a triac. I'd be very surprised if they didnt.
 
I was thinking can I use the positive AC waves for firing the TRIAC (no phase delay) at Q I & Q II, and the negative AC waves QIII and QIV to obtain the power for the circuit?
Anyway I am not searching dimming features here... ?
 
Hi,

Not sure what you mean there.
 
Hi,

Not sure what you mean there.

You're right, there is no sense in this.
If the traic is ON all the time, even during the positive wave only, the whole power supply will be shortened at this time.

next step... I will post the circuit diagram here and the video how currenty the circuit operates. It actually works, but bulb flickering, dimmed too much, and the receiver doesnt want to turn the bulb off very often (small current i guess). + too much heat produced by the current dropping R and the ones before the H11AA1 ZCD IC. And also the AC capacitor of the power supply gives a very anoing buzz noise. BTW I need to implement also a snubber filter. I am just not sure about the L value - maybe 150-200 uH? and the capacitor.. what kind of cap, value, etc. Maybe because there is no filter the bulb is flickering...

So much efforts in order to have a bulb ON/OFF with a receiver. I cannot belive it. Also I cannot belive there is not even one circuit on the internet for a power supply being connected in series with the load (the bulb).


I think I might have some errors though in my circuitry because I cannot belive some H11AA1, CMOS 555, flip-flop CD4013, schmitt trigger CD40106, receiver encouder PT2272-M4 and the receiver module itself + opto MOC3052 and the traic require so much power. For example what is the main driver not to overload the supply? high values of what? low values of what? is there are general rules? Also, maybe the LEDs are consuming too much? I have couple of LEDs for this and that. Maybe I am very close and some resistor is missing or in exceed and therefore I cannot do it.

Actually what is the generic reason for the heat? The circuit drawing higher power than the supply can provide? Or it can be something else too.

Those were my thoughts above. No need to answer (since no diagram posted yet) but obviously I am starting to be despairing a bit.
 
MrAl, (and others if still some readers left :)
I'd appreciate if you check out this component:
U2102B - Multifunction Timer IC - ATMEL Corporation

https://pdf1.alldatasheet.com/datasheet-pdf/view/256745/ATMEL/U2102B.html

Inside the document there are explanations that this IC can be used in "two wire" systems.
I suppose I can utilize it with some help, as long as you say it might work.
I know that if we decide to try it I will need to forget the TRAIC and go with MOSFET or IGBT switch, i.e. "trailing edge", or in other words to cope with reverse phase control.

BTW, the dimmer in my bedroom (cost about 30EUR) is perfect! It operates with two wires, it has a manual button and it operates with infrared commands from any TV remote control (i can program which button to trigger it). If the remote button is pressed - it changes ON/OFF, if you hold it - it starts dimming. The same if you press/hold the manual button on the dimmer. I've opened it... it has two MOSFETs or IGBTs (I am not sure exactly) and some PIC as a heart. And it is totaly silent! I mean totaly. Also it can run the bulb from maybe 0 to almost 100% with very small steps. I can post some images if someone is interested.

Again to underline here (for new readers) - I am not after dimming functions here. I just want to build properly the power supply without neutral wire, in order to have enough power for the RF receiver and to just turn ON/OFF my bulb on the ceiling. Obviously that ON would mean (90% for instance, as I would need some power for the remaining elements in the circuit). So the dimming will be set by soldering on 95% lets say, in order to have some phase delay for the power supply.

Here additional useful details about how this IC is used and a picture of such "trailing edge" dimmer (two wires can be seen on the photo) and also shown the measured waveforms and explanation:
(scroll somewhere in the middle of the document):
**broken link removed**
 
Last edited:
Hi,

That chip looks like it might work, and they seem to suggest that it will work in a two wire system, as long as there is a minimum phase turn on point in a reverse phase topology. That means again we have to wait to turn on even though we want full bulb power. That's not any different than before though, where we had to do that the same way.

So it's up to you if you want to try this chip out or not, i dont know what the cost is offhand. There's also the chance of burning it out during testing so you have to have at least two of them to start with, preferably more.

So you really want to try a 'reverse phase' topology? Again, that's up to you.

BTW did you price any small current transformers? What's the verdict there so far?
 
Hi,

BTW did you price any small current transformers? What's the verdict there so far?

I am experiencing problems finding one. I even don't know how to search for them. The only thing I can guess is to type in 'current transformer' in Google... but nothing. Also I need a compact version. If these have some IDs like the ICs.... some datasheets or something that would help.

About the type of phase control - i really don't care, also it is the same if TRIAC or MOSFET or IGBT.... as long as i find a solution that works... actually the problem i have (and you know this even before I realized it) is the power supply.. once I manage to derive enough power, then doesn't matter if i do it 'leading edge' or 'trailing edge' :)

Hey MrAl, many thanks again!

I might order 5 chips for the try (1$ or something per IC - from China of course), and together with something else in the shipment.
But it will take time, so I will continue investigating the previous options we've discussed while I receive them.


BTW, is using a IC bridge rectifier an advantage or disadvantage? How about using only 2 diodes?
 
Another fundamental (or maybe stupid) question came to my mind.
When the bulb is ON, the TRAIC is powered for most of the half wave, let's say the bulb is at 95% brightness and there is enough voltage for the power supply.
When the bulb is OFF, then what?
> the TRIAC is totay OFF? ... or
> the TRIAC is powered only for some small portion at the end of half wave so that the bulb is 0% bright (100% dimmed)?
If I choose the second option does it mean that most of the power will be handled by the bulb and thus there will be no heat on the current dropping resistor? Also would that be a problem - the TRAIC to operate non-stop? I am starting to think that when I solve the problem with the power supply, then the problem with the heat will still persist. The power should be handled somehow, when the bulb is off.
 
Hello,

What circuit are you talking about?
What dropping resistor?
 
Hi,

He must have gotten it to work or else he'd be back here :)
 
Hi Ibob,

Have you managed to make any progress with this. I am looking for a similar solution.

Well, it is really a long story.... bottom line... you cannot do it unless you utilize switching mode power supply.
There is just not enough power with the conventional methods I wanted to utilize and plus... there was a lot of heat produced and/or the intensity of the bulb was not satisfactory, etc.

Once I've hooked up SMPS module, supplying both 5 and 12V DC, I've managed to power all the rest.
So, I wanted to drop few lines here, but I was hoping to find another way. So far nothing... SO...... I found very cheap SMPS - only 2.5 EUR (here in BG), and there are also lots of them in aliexpress.
They are so light, cheap and effective that even if I build my own one, I will never achieve the same price, maybe not the same small size and for sure I will loose much more time.

So, unfortunately, I will stick to the switching power supply modules available and I will continue with my home automation projects. At least I've learnt a lot of things here and by diving into all this...

Big thanks goes to MrAl for all his time dedicated to this discussion.
 
Having read your posts and the replies, you appear to be unfamiliar with mains electrical safety - an amateur. Therefore you should definately stay away from mains power and find another project.
You may end up starting a fire or worst still, killing someone.
 
Having read your posts and the replies, you appear to be unfamiliar with mains electrical safety - an amateur. Therefore you should definately stay away from mains power and find another project.
You may end up starting a fire or worst still, killing someone.

Really? I didn't know that. I will immediately stop working with electricity and I will light candles in my house. /or the fire risk then will increase?/
God, they should have told me this when I was studying for electrical engineer....
Thanks for helping me out!

Whenever you see something posted by me, from now on, please don't bother to read/reply.
 
Hi All,

After almost 11 month of googling, I have finally found this forum that has something about what I am looking for.
I need to do a similar project.
Could you please help me with this?
This circuit requires GND and 3v3... Where can I get these from only the live wire?
I need to power up this circuit along with my microcontroller circuit with only live wire.
both circuits need 3v3~5v and <250 milliamp

Thanks a lot in advance
 

Attachments

  • ac_dimmer_schematic_v.2.PDF
    201.3 KB · Views: 456
Reading through the thread I think you'll find that this project was unsuccessful :(.
 
Reading through the thread I think you'll find that this project was unsuccessful :(.

I once again feel the URGE to inform about Transformerless Power Supplies. Namely they need TO BE UNDERSTOOD...and then they bust balls and kick ass left right and center...

Regulars here know my passion for them..if understood. Understanding is the thing. They are DANGEROUS if not understood..

And the most hilarious thing is.....used in lot's of places that those that condemn then do not even know about...

Give me one SMPS that is rated @ 220VAC....that can handle 5 Minutes of 400Vac.. (phases gone wrong)...

So funny to watch everything else pop....and the my little light does not even know about it. All in a Days work.

Albeit, a little 1W LED light...around only 70 to 80 Lumens, a little PP3 is the battery.

A team that handles everything thrown at it. Just thought I could once again share. So many Websites prohibit discussion about X2 Caps being used as Mains Droppers...what I will say from my own experience..do it right and you have a product that will never fail...no matter what happens:)

And that is all I have to say.

Thanks Guys,
tvtech
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top