Is it possible? To Isolate a transformerless topology?

[ali6x944]

I was working resently on a transformerless supply i metioned in earlyer posts and some said it was not safe to use becaues it is not isolated from the earth refrenced ac mains, so i want to know if i could isolate it somehow?
Help Please

[Brumby]

The way to isolate two sides of any circuit is to have the energy transferred by converting it into a form that allows complete separation.

Transformers do this by converting electrical energy into magnetic energy with a separate circuit that shares the magnetic field which then converts back to electricity.

Opto couplers do something similar using light, but the energy transferred is miniscule and useless for a power supply.

The only other "practical" option that comes to mind is a motor/generator setup where the intermediate form is mechanical energy - but that makes transformers a more attractive choice.

[bitshift]

From what I understand (I could very well be wrong), you can basically isolate circuits in three ways using either:

• Magnetics
• Capacitance
• Light

Light and capacitance are not too good at transferring large amounts of power so they are typically used to isolate things like fast changing digital signals. In these cases, both sides of the isolation barrier have their own power source.

[Brumby]

I would not consider capacitance to be an effective isolation mechanism.  It's inability to handle instantaneous voltage changes without consequences rules that out in my book.

Such a problem does not exist with magnetic, optical or mechanical mechanisms.

[bitshift]

I'm not too certain on the uses of the capacitive isolators but you can buy them http://www.ti.com/lsds/ti/isolation/digital-isolators-overview.page

[rs20]

I would not consider capacitance to be an effective isolation mechanism.  It's inability to handle instantaneous voltage changes without consequences rules that out in my book.

I wouldn't be so black-and-white about it -- any isolation solution will have some capacitance between primary and secondary, and indeed even magnetic isolators have Y-class safety caps bridging the isolation between primary and secondary. Why not actually transfer power over these caps?

Put another way, if the "consequence" of an instantaneous voltage change is that the resulting current gets immediately sunk into the mains earth connection; well then that hardly seems like a serious concern. Like I said, most magnetic isolators have these Y-class caps already.

With a capacitance chosen to not trip the RCD in your house, and arbitrarily high frequencies, you can theoretically transfer an unlimited amount of power. Of course in practice, you'd only be able to get away with very small power levels, but there are plenty of circuits these days that'll happily run on milliwatts, so it might even be an worthwhile solution in those cases.

[Brumby]

The key word there is "digital".

On the basis of high speed digital signals, a low value capacitor can effectively transfer signal information which would have a significant high frequency component while providing negligible transfer of low frequency (eg mains) and DC (leakage) - as is stray capacitance around a transformer.

The sole premise of this being able to claim any sort of isolation is the extremely low amount of energy that is actually able to cross from one side to the other - which is in direct contradiction to the requirements of a power supply of any significant VA capability.

[Jeroen3]

Is it possible to Isolate a transformerless topology?

Yes. With batteries.

[Kalvin]

There are few solutions using piezo-technology, for example:

http://www.izm.fraunhofer.de/en/abteilungen/rf---smart-sensor-systems/key_research_areas/piezoelectric-power-supplies.html

https://translate.google.com/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.izm.fraunhofer.de%2Fde%2Fabteilungen%2Frf---smart-sensor-systems%2FForschungsschwerpunkte%2FPTSMPS%2FESPLIKA.html&edit-text=

I am not sure about the price or availability. Google finds also few other application notes using piezo-technology for galvanic isolation.

[orolo]

You can go all Tesla on your circuit and try EM coupling, via two coupled resonantors. Which is the basic idea behind a crystal radio. This would be a (somewhat) more efficient form of light coupling. I wonder if thermoelectric coupling could be viable in some exotic application.

[Zero999]

Capacitance isolation is fine. I've done it before to make a solid state relay. I built an oscillator with some logic gates and coupled the power to the gates of the MOSFETs via some capacitors and rectifier diodes.

I used the circuit on the right with a lower value of R1 and a 1MHz oscillator for the control signal. Safety wasn't an issue so I didn't need to worry about Y-rated capacitors. Using the old 2N7000 (this was a low power/signal switching application) I could switch the load at over 20kHz.

[Brumby]

I think there is a particular aspect of the design that is being overlooked by some...

Refer to this:
N-mosfet as an external switch for mc34063

hello everybody,
I'm working in a switching transformer less power supply based on the great mc34063 ic.
I want a fare current  capability and a duel output voltage supply that could reach 0v, so i will put it like this:
220/110v ac,60hz-------> -15v -0- +15v @ 3A, because I'm using a switching supply we would have some voltage ripple,i could deal with 100mv pk-pk .
 
after some work i got these measurements -raw rectified dc-:
@ 220v AC, i got 306.6v dc and a max current of around 325mA,
@ 110v AC, i got 152.2v dc or so, and a max current of around 175mA,
so the first problem is the big voltage drop that should be present to run the mc34063 and the second problem is the the output current capability, the third is efficiency related problems. 
oh! by the way, I'm using a capacitor based transformer less design, I already dealt  with inrush current problems, so u don't need to worry about that.
please help!
thanks .

The OP was after a +/-15V split supply at 3A.  He is after power, not signalling.

Unless I've missed something.............

[ali6x944]

it is power, I'm talking about power here

[Brumby]

That's what I thought.

I still find a capacitive solution that provides the power required with a degree of real isolation somewhat elusive.

[macboy]

That's what I thought.

I still find a capacitive solution that provides the power required with a degree of real isolation somewhat elusive.

If you only consider that the capacitor are connected from primary to secondary side (for lack of better terms) then you are stuck. The capacitors will transfer energy, but they also AC-couple one side to the other. This means that this solution is only viable if the potential difference between sides is a fixed DC voltage. Any AC potential difference is transferred directly through the capacitor. Big power = big caps = low Z AC coupling.

However, consider a solution where a capacitor is first connected to a DC source on the primary side, then disconnected and connected to a reservoir capacitor on the secondary side. Repeat frequently, and you can transfer a significant amount of energy without having a capacitor directly connected between sides. This is called a charge pump, and while it is commonly used to boost and/or invert voltages (e.g. MAX232 RS-232 line driver uses them to derive +/-10 V rails from a 5 V supply), it could in theory also be used to isolate.

[rs20]

If you only consider that the capacitor are connected from primary to secondary side (for lack of better terms) then you are stuck. The capacitors will transfer energy, but they also AC-couple one side to the other. This means that this solution is only viable if the potential difference between sides is a fixed DC voltage. Any AC potential difference is transferred directly through the capacitor. Big power = big caps = low Z AC coupling.

...which you then just shunt away by shorting the ground of the secondary to mains Earth. Problem solved (provided your primary-to-secondary C's aren't too large.)

[Brumby]

However, consider a solution where a capacitor is first connected to a DC source on the primary side, then disconnected and connected to a reservoir capacitor on the secondary side. Repeat frequently, and you can transfer a significant amount of energy without having a capacitor directly connected between sides. This is called a charge pump, and while it is commonly used to boost and/or invert voltages (e.g. MAX232 RS-232 line driver uses them to derive +/-10 V rails from a 5 V supply), it could in theory also be used to isolate.

I see what you're saying and it sounds quite reasonable.  You could do this mechanically with relays - but that's not a particularly practical solution.  I'd be interested in the effectiveness of the isolation for a solution that is.

[mikerj]

That's what I thought.

I still find a capacitive solution that provides the power required with a degree of real isolation somewhat elusive.

That's because it simply doesn't exist.  A capacitive dropper does not isolate you from the mains because the output of the capacitor is still referenced to ground (i.e. the ground you are standing on).  For any significant power transfer you need a transformer to provide isolation (ignoring solutions like motor/generator sets which tend to be impractical).

[Zero999]

That's what I thought.

I still find a capacitive solution that provides the power required with a degree of real isolation somewhat elusive.

That's because it simply doesn't exist.  A capacitive dropper does not isolate you from the mains because the output of the capacitor is still referenced to ground (i.e. the ground you are standing on).  For any significant power transfer you need a transformer to provide isolation (ignoring solutions like motor/generator sets which tend to be impractical).

A transformer also connects the secondary to ground, via the inter-winding capacitance.

[Brumby]

Yes, but that's like talking about ants in a conversation on elephants.

[mikerj]

That's what I thought.

I still find a capacitive solution that provides the power required with a degree of real isolation somewhat elusive.

That's because it simply doesn't exist.  A capacitive dropper does not isolate you from the mains because the output of the capacitor is still referenced to ground (i.e. the ground you are standing on).  For any significant power transfer you need a transformer to provide isolation (ignoring solutions like motor/generator sets which tend to be impractical).

A transformer also connects the secondary to ground, via the inter-winding capacitance.

Pretty much irrelevant in the case of low frequency power transformers from a safety perspective.  If you were grounded, would you be happy to touch the output of a capacitive dropper type supply?  How about the output of a low voltage transformer?

[Zero999]

That's what I thought.

I still find a capacitive solution that provides the power required with a degree of real isolation somewhat elusive.

That's because it simply doesn't exist.  A capacitive dropper does not isolate you from the mains because the output of the capacitor is still referenced to ground (i.e. the ground you are standing on).  For any significant power transfer you need a transformer to provide isolation (ignoring solutions like motor/generator sets which tend to be impractical).

A transformer also connects the secondary to ground, via the inter-winding capacitance.

Pretty much irrelevant in the case of low frequency power transformers from a safety perspective.  If you were grounded, would you be happy to touch the output of a capacitive dropper type supply?  How about the output of a low voltage transformer?

Either, providing it's properly designed. If the capacitive dropper supply uses low value Y1 capacitors for isolation, with proper clearances on the PCB, then it's no more dangerous than a transformer.

[Psi]

A transformerless psu followed buy a small cheap isolated dcdc works well but is of cause not transformerless any more. But still safer than a cheap mains to DC transformer module. Also, you can often run most things off the non isolated side and only use the dcdc for external interfaces that need power. That way you can use a lower power dcdc.

[bson]

I'm not too certain on the uses of the capacitive isolators but you can buy them http://www.ti.com/lsds/ti/isolation/digital-isolators-overview.page

Useful for high-speed isolation where the impedance of an inductance is too high and optics can't be switched fast enough.

[mikerj]

Either, providing it's properly designed. If the capacitive dropper supply uses low value Y1 capacitors for isolation, with proper clearances on the PCB, then it's no more dangerous than a transformer.

Wrong, capacitors do not provide isolation as already stated.  This should be entirely obvious since the voltage at the output of a capacitive supply is still referenced to ground.

As for the touching the output of such a supply, you are clearly either very trusting or can't see the inherent danger.  Just consider what happens if live and neutral are reversed with the typical resiator/capacitor/rectifier/zener design...suddenly the output is at full mains potential w.r.t ground.

[rs20]

That's what I thought.

I still find a capacitive solution that provides the power required with a degree of real isolation somewhat elusive.

That's because it simply doesn't exist.  A capacitive dropper does not isolate you from the mains because the output of the capacitor is still referenced to ground (i.e. the ground you are standing on). 

Strawman detected! You've just leapt from "any capacitative solution" to "the standard capacitative dropper". Anyone reasonable agrees that a standard capacitative dropper does not provide isolation.

I think the OP's question is; what about the idea of a (hypothetical) topology that did provide galvanic isolation from primary to secondary, using only capacitors. Two interesting/reasonable (albeit very low power/inefficient/expensive) proposals have already been offered:

• Push-pull through pair of Y caps to be rectified on the secondary side (which is safety earthed on the secondary side to discharge the AC current that would come through those Y caps, just like the noise suppression caps in a traditional magnetic converter.)
• Relays/solid state switches that charge up a large cap on the primary side, and then "hand it over" to the secondary side. Fascinating concept; achieving safety ratings on those Relays/solid state switches would be "challenging" though.

I agree that this is probably more academic than practical.

[Brumby]

I think the OP's question is; what about the idea of a (hypothetical) topology that did provide galvanic isolation from primary to secondary....

I believe the Op was actually building a transformerless supply and, in response to concerns posted, wanted to find out if it could be isolated.  A practical solution was the objective.

[ali6x944]

https://drive.google.com/file/d/0B5vW-k7HbsL4d1JZUVpFUVhpVkk/view?usp=sharing
that what i want to isolate

[Zero999]

Either, providing it's properly designed. If the capacitive dropper supply uses low value Y1 capacitors for isolation, with proper clearances on the PCB, then it's no more dangerous than a transformer.

Wrong, capacitors do not provide isolation as already stated.  This should be entirely obvious since the voltage at the output of a capacitive supply is still referenced to ground.

As for the touching the output of such a supply, you are clearly either very trusting or can't see the inherent danger.  Just consider what happens if live and neutral are reversed with the typical resiator/capacitor/rectifier/zener design...suddenly the output is at full mains potential w.r.t ground.

Wrong, Y1 capacitors are connected across the primary and secondary in most switched mode power supplies. They provide more than adequate isolation at 50 to 60Hz.

It's possible to use 1nF Y1 capacitors and a 1MHz oscillator to couple the energy from the primary to secondary side. The 1nF capacitors will have a high impedance at 50Hz (3.18M) and an impedance of just 160Ohm at 1MHz. They're rated to withstand 8kV spikes, designed to fail open circuit and give the same degree of isolation as a bulky transformer.

[ali6x944]

this is the circuit i want to isolate
https://drive.google.com/file/d/0B5vW-k7HbsL4d1JZUVpFUVhpVkk/view?usp=sharing

[ali6x944]

photos of the real deal:
https://drive.google.com/file/d/0B5vW-k7HbsL4QnYwUHdJWWM5YVU/view?usp=sharing
https://drive.google.com/file/d/0B5vW-k7HbsL4eldnV0trMnlQdVk/view?usp=sharing

[Zero999]

Why not simply attach the schematic?

No, that's not isolated at all. What's the load in real life?

[JacobPilsen]

First of all, there must be also some "C3" in neutral wire.
Capacity of C1 or "C3" could not be over 820µF, due to 30mA in ordinary RCD.

[rs20]

First of all, there must be also some "C3" in neutral wire.
Capacity of C1 or "C3" could not be over 820µF, due to 30mA in ordinary RCD.

Is the aim here to not trip an RCD, or to not kill a human? What happens when lightning strikes? Why would an RCD be tripped anyway, when there is no Earth connection in sight?

[Zero999]

First of all, there must be also some "C3" in neutral wire.
Capacity of C1 or "C3" could not be over 820µF, due to 30mA in ordinary RCD.

Don't you mean 820nF?

Anyway, that's not good enough. If this were to be used for isolation, the total capacitance between the primary and secondary should be under 5nF and it needs to be a Y1 rated capacitor or two Y2 capacitors in series.

[JacobPilsen]

Is the aim here to not trip an RCD, or to not kill a human?

Both.
RCD is definined such way.

What happens when lightning strikes?

Basicaly the same as with transformer tested for same voltage (5000V for Y2).

Don't you mean 820nF?

Yes, sorry.
(But it should be 415nF.)

If this were to be used for isolation, the total capacitance between the primary and secondary should be under 5nF and it needs to be a Y1 rated capacitor or two Y2 capacitors in series.

In this case we have only 360µA to use.

[Zero999]

If this were to be used for isolation, the total capacitance between the primary and secondary should be under 5nF and it needs to be a Y1 rated capacitor or two Y2 capacitors in series.

In this case we have only 360µA to use.

Yes, that's the sort of leakage current allowable across double/reinforced insulation. Only tiny Y1 capacitors and high value, high voltage resistors are allowed.

It's possible to get more useful current but not at 50 to 60Hz. The solution to this is to power an oscillator from the capacitive dropper, couple the energy to the secondary side via tiny Y1 capacitors and rectify it with fast Schottky diodes.

[ali6x944]

no you misunderstood me, what i was trying to say is i built this circuit and i want to isolate it, how?   

[Zero999]

no you misunderstood me, what i was trying to say is i built this circuit and i want to isolate it, how?

With an isolating transformer.

[rs20]

Isolation isn't just something you tack on as an afterthought. If you decide you want isolation, you should probably throw out your existing design and start from scratch with your requirements.

Having said that, you can just add an isolated DC-DC converter, although by the time you've arranged all that you'd be better of just buying a off-the-shelf wall wart.

[mikerj]

Either, providing it's properly designed. If the capacitive dropper supply uses low value Y1 capacitors for isolation, with proper clearances on the PCB, then it's no more dangerous than a transformer.

Wrong, capacitors do not provide isolation as already stated.  This should be entirely obvious since the voltage at the output of a capacitive supply is still referenced to ground.

As for the touching the output of such a supply, you are clearly either very trusting or can't see the inherent danger.  Just consider what happens if live and neutral are reversed with the typical resiator/capacitor/rectifier/zener design...suddenly the output is at full mains potential w.r.t ground.

Wrong, Y1 capacitors are connected across the primary and secondary in most switched mode power supplies. They provide more than adequate isolation at 50 to 60Hz.

The presence of Y1 caps across the primary and secondary creates a direct path from the AC main inlet to the DC output, ergo it's no longer truly isolated, albeit the maximum leakage current should be safe.  However, since the leakage current is limited to a very small value, it's not actually going to be very useful for powering anything (micropower circuits maybe).  Up the capacitor value to get more current and we are back to where we started.

Strawman detected! You've just leapt from "any capacitative solution" to "the standard capacitative dropper". Anyone reasonable agrees that a standard capacitative dropper does not provide isolation.

I think the OP's question is; what about the idea of a (hypothetical) topology that did provide galvanic isolation from primary to secondary, using only capacitors. Two interesting/reasonable (albeit very low power/inefficient/expensive) proposals have already been offered:

• Push-pull through pair of Y caps to be rectified on the secondary side (which is safety earthed on the secondary side to discharge the AC current that would come through those Y caps, just like the noise suppression caps in a traditional magnetic converter.)

Still not isolated, but now you've added a ground connection to guarantee the output is referenced to ground.  That will get rid of those pesky leakage currents, but what would happen if you tried to e.g. float the output on top of another voltage which is ground referenced?  Bang.  If the output was isolated you could do this with impunity (obviously limited by the maximum isolation voltage in the supply).

• Relays/solid state switches that charge up a large cap on the primary side, and then "hand it over" to the secondary side. Fascinating concept; achieving safety ratings on those Relays/solid state switches would be "challenging" though.

Congratulations, you just (re)invented a flying capacitor converter   There are COTS off-line switcher ICs that do just this, but you are then at the mercy of a semiconductor device between you and 110/230VAC.  Can you absolutely guarantee that device will fail in a safe state?

I'm honestly quite surprised at how many people don't understand what galvanic isolation is.  As a simple thought experiment, take the output of your proposed supply and connect one or the other output terminals to ground referenced voltage (e.g. 240v live terminal).  If it's galvanically isolated then nothing exciting will happen and no significant additional current will flow.

[Jeroen3]

Any topic regarding galvanic isolation leads to a debate here. Maybe it needs a Fundamentals Friday?

[Zero999]

Either, providing it's properly designed. If the capacitive dropper supply uses low value Y1 capacitors for isolation, with proper clearances on the PCB, then it's no more dangerous than a transformer.

Wrong, capacitors do not provide isolation as already stated.  This should be entirely obvious since the voltage at the output of a capacitive supply is still referenced to ground.

As for the touching the output of such a supply, you are clearly either very trusting or can't see the inherent danger.  Just consider what happens if live and neutral are reversed with the typical resiator/capacitor/rectifier/zener design...suddenly the output is at full mains potential w.r.t ground.

Wrong, Y1 capacitors are connected across the primary and secondary in most switched mode power supplies. They provide more than adequate isolation at 50 to 60Hz.

The presence of Y1 caps across the primary and secondary creates a direct path from the AC main inlet to the DC output, ergo it's no longer truly isolated, albeit the maximum leakage current should be safe.

The situation is no different to a big ass isolation transformer with a large inter-winding capacitance.

  However, since the leakage current is limited to a very small value, it's not actually going to be very useful for powering anything (micropower circuits maybe).  Up the capacitor value to get more current and we are back to where we started.

Use a high frequency oscillator to couple the energy across the capacitive isolation barrier.

A transformerless psu followed buy a small cheap isolated dcdc works well but is of cause not transformerless any more. But still safer than a cheap mains to DC transformer module. Also, you can often run most things off the non isolated side and only use the dcdc for external interfaces that need power. That way you can use a lower power dcdc.

Be careful, some of those isolated DC:DC converters only have functional insulation between the primary and secondary and are not to be relied upon for protection against shock.

[ali6x944]

could i not use the bleeder resister? and include a 3A fuse and an MOV?

[Zero999]

could i not use the bleeder resister? and include a 3A fuse and an MOV?

Yes you could but a bleeder resistor, fuse and MOV won't provide isolation or any protection against shock.

What is the power supply for? Why not just buy a proper isolated power supply?

[Brumby]

I'm tending to agree with that.

It's like - "I've built a car and want to know how to make it safe - but I don't want to use seatbelts."  After all the discussions, debates, alternatives, pros and cons have been thrashed out - it would seem the simplest - and most straightforward - solution is to put in the seatbelts.

So, for this case, use a transformer.  If size is an issue, go for a switchmode solution.


(But I do like the idea floated earlier about Dave doing a blog about this subject.)

[ali6x944]

yes i liked it too

[ali6x944]

but i can not simply buy and use a proper isolation transformer for this project because space constraint, but i was going to use a switching solution  a -mc34063- because it is the simplest to use and i have a lot of it, but due to over voltage problems i could not use it.
so i searched in my junk box and i found this UC1846-EP and i was wondering if i could use it instead ...
any way thanks a lot everybody 

[Jeroen3]

What are the size contraints?
You can get a 50 Watt power supply in a real small box today. Apple's macbook adapter is 45Watts and is really small. And it has several transformers!

Do you want to create a 50 Watt power supply on a SD card?

[ali6x944]

they are 5.5cm length , 5cm width, and 2.2cm in height.
if we were talking about line transformers i can't use them, but for switching no problem to have them.
for the power out 45 watts @ 15v & 3A.