Balanced 240V AC mains supply, simply does it

[Reid Malenfant]

I'm sure Ali will be interested in this one so i thought i'd get something up here now with a full test coming soon. However i trust my judgement having previously been involved in set up & testing of rather large 3 phase power supplys & battery chargers, 100KVA+ anyone? :lolsign:

I recently purchased two 3.3KVA site transformers, to be specific these ones here! (http://cgi.ebay.co.uk/110V-SITE-TRANSFORMER-3-3-KVA-NEW-BOXED-/190424455499?cmd=ViewItem&pt=UK_BOI_Industrial_Tools_Generators_ET&hash=item2c562e994b) I'll be honest & say that when i popped the top off i was cursing rather a lot as i didn't think i'd be able to do what i wanted, some thought led me to the conclusion that things couldn't be better as the transformer is "potted" up so will make less noise.

Here goes with the internal pictures...

http://i939.photobucket.com/albums/ad240/speakermark/test/Picture003-1.jpg
http://i939.photobucket.com/albums/ad240/speakermark/test/Picture002-1.jpg
http://i939.photobucket.com/albums/ad240/speakermark/test/Picture001-1.jpg

I'll attempt to make this simple so everyone can understand, if anyone has any questions then for gods sake ask as we are dealing with voltages & currents that can kill here :eek:

Unfortunately like everything in this world these days the 3.3KVA rating isn't continuous, there is a nice label on the side saying that continuous rating is 2/3 of the peak. So each transformer will do 2.2KVA continuously, please remember you'll need 2 of these things to get your balanced mains so in reality you'll end up with a 4.4KVA or if you like 4.4KW balanced mains supply. The peak will be 6.6KVA, & i'm sure it'll be able to supply this for more than a few minutes - how deaf do you want to be :doh:

Anyway, heres how to go about it... First up don't have the transformer plugged in when operating on it!!! To get to where we want to go isn't really hard at all. Firstly cut the two earth leads that dive down into the potting compound, i'd suggest you leave about 5cm length just in case you ever decide to use the thing for it's intended purpose :lol: That way a terminal block can be employed to reconnect as per standard.

Having cut the earths you'll have a 110/120V secondary with no earth reference. So find yourself a single block of terminal block & reconnect the earth from the mains that comes from the mains input plug to the earth that connects to the output sockets. Now get yourself a 30Amp terminal block... Disconnect the red & blue leads that come from the potted up transformer (there are 2 of each) & connect both reds into one part of the block & both blues into another terminal.

At this point in time you can toss a coin & decide which terminal you are going to connect the earth to, it doesn't matter, red or blue ;) However please take note of which one you connect it to! The reason being is that on the second transformer you do everything just the same, except you connect the earth to the wire colour that you didn't on transformer number one ;)

If this all appears to be a tad complicated don't worry, i'll be back tomorrow with pics of the completed article & a pretty scope trace & voltage readings to show you what is going on :lol:

Obviously both transformers will need to be mounted in a strong case to prevent little fingers doing what they do. I'd suggest some nice 18mm MDF. A couple of mains sockets on the side & you are pretty well sorted. Frankly i'll leave anyone who wants to tackle this to sort out the aesthetics themselves :cool:

Back tomorrow, cheers for now :) More details to follow...

[The Grand Wazoo]

Can I take the opportunity to express to all who may be interested in this that if you don't know EXACTLY what you're doing then you shouldn't even think about this.

I struggle with the wisdom of this even being allowed to remain visible to all and sundry. Remember, it's not just us regulars (who I think pretty much all have some common sense between our ears) who can read this.

[The Vinyl Adventure]

i wouldnt have the first clue what it even is, never mind trying to emulate it ...

[colinB]

Its a balanced mains innit:scratch:

[John]

Whilst I find Mark Article Fascinating I also agree with Chris Do not do this unless you seriously know what you doing. This could be lethal

[Dave Hewitt]

Hi
See Airlink Transformers web page regarding balanced mains supplies,seem cheap enough.
Dave.

[Alex_UK]

I agree Chris/Hamish etc - when I saw this thread there was a "whooshing" sound - I just hope it was the sound of the idea going straight over my head and not someone reading this thread going up in smoke... PLEASE OBSERVE ALL WARNINGS, PEOPLE!

[anthonyTD]

hi all,
a true isolating transformer has a seperate primary and secondary, in this case 240v primary, 110v or 55+55 secondary this type of transformer is used mainly for plant tools on site for added safety. what mark is describing when done properly will work, unfortunetly i too strongly recommend that unless you are absolutely sure you understand what your doing and the consequences if you dont then i would leave well alone. however i understand that this thread will be of interest to those who are capable and competent enough to investigate the advantages that may be apparent folowing its correct implimentation.
A...

[Reid Malenfant]

Indeed, i'm sure i might have made it known not to attempt this unless you feel confident in what you are doing & that you can hopefully read, & follow instructions.

These will be comprehensive enough & in reality you'll need at least a mulimeter that can read 240V AC, ohms could be handy to, to check things are correct before you even think of powering any of this up....

If you have a multimeter you'll be able to do this, if you don't then i wouldn't attempt what i will do! You need to test certain things simply because you can't guarantee everything is the same all of the time. I'll explain this better when i post again :) When i'm a little more sane... 2355 2356


Like tomorrow :) Well hopefully...

[Barry]

Hi all,
a true isolating transformer has a separate primary and secondary, in this case 240v primary, 110v or 55+55 secondary this type of transformer is used mainly for plant tools on site for added safety. What Mark is describing when done properly will work, unfortunatly I too strongly recommend that unless you are absolutely sure you understand what your doing and the consequences if you don't then I would leave well alone. However I understand that this thread will be of interest to those who are capable and competent enough to investigate the advantages that may be apparent folowing its correct implimentation.
A...

+1

To confirm Ali's (Tait) comment, both of the secondary lines must be fused. And, strictly speaking, both being phase lines, ought to be coloured brown.

[Dave Hewitt]

Hi
Under fault conditions will both fuses blow at the same time? The commercial units seem to rely on an overload trip on the primary of the transformer.
Dave.

[Reid Malenfant]

To confirm Ali's (Tait) comment, both of the secondary lines must be fused. And, strictly speaking, both being phase lines, ought to be coloured brown.
Agreed, however the fuses will only protect the transformer secondaries, there is a much better way of incorporating vastly more protection :)

Hi
Under fault conditions will both fuses blow at the same time? The commercial units seem to rely on an overload trip on the primary of the transformer.
Dave.
Not necessarily, though in reality it doesn't matter ;) As long as one blows then there is no circuit fullstop.

Your comments about the circuit breakers are correct, & this can be used to an advantage. I was mulling it over as frankly i'd hate to have the second transformer still kicking out 120V or there abouts if one had shut down. Then i thought about what would be in the circuit & came up with a rather perfect solution. You see if one transformer shuts down by the circuit breaker opening there is still a circuit for the other transformer to put 120V though. This is simply through the secondary of the shut down transformer & back to the working secondary. I don't like this myself so i'll devise a simple system to shut the thing down fully.

What i propose is a simple circuit with a relay that is held open when both secondaries are powered (240V accross both) & drops out with only one working (120V). When this closes it'll deliberately create an earth fault & along with a 30ma 25A RCD will shut the whole thing down. Much like the test button on the RCD itself ;)

The 25A section of the RCD will protect the whole thing from any overload, the 30ma section will protect against either phase output from any earth faults in any equipment connected to the unit itself. The above mentioned relay circuit (which will need a push to break switch inline with the relay contacts to start the whole thing up!) will shut it down if one secondary goes offline due to a primary circuit breaker opening or one of the secondary fuses blowing...


Oh it's all good fun when things like this get you thinking :smoking:

[Snoopdog]

Fascinating stuff and despite all the caveats about safety etc., I applaud Mark's post, the knowledgable contents of his contributions to this forum and the quality of the accompanying 'artwork.'

[Barry]

I used the term 'fuse' as a general term for circuit protection. Agree with Mark, that under a fault condition one side of the secondary could still be connected and supplying 120v.

A relay fitted in series with both sides of the secondary and operated by an over-current sensing circuit is the best solution.

Again I would advise extreme caution: 120V can be lethal - just remember that the next point of safety might be the RCB in your CU (if you have these fitted) and these usually require an inbalance of 10mA to operate. And they can take 25ms to operate (two and a half cycles at 50Hz). 10mA might still be lethal!

Regards

[Reid Malenfant]

<snip> I used the term 'fuse' as a general term for circuit protection. Agree with Mark, that under a fault condition one side of the secondary could still be connected and supplying 120v.

A relay fitted in series with both sides of the secondary and operated by an over-current sensing circuit is the best solution.
I'm giving this some thought, well i'm having all sorts of thoughts actually ;) You have given me a rather interesting idea that would in effect be better than a relay on the output. How does output shut off in less than 1 micro second (0.000001 seconds or less) sound in the event of a fault? :scratch: :)

I'm working on it 2357

Before anyone says that's not possible i can tell you it's very possible indeed :eyebrows:

[Puffin]

Mark. What are the pictures I can't see in your posts?

[Reid Malenfant]

Mark. What are the pictures I can't see in your posts?
You can't? That's odd, well i'll be sticking some more up inside of half an hour so feel free to take a look & let me know if you still can't see them. They are hosted on Photobucket, i'm not sure if that's the reason but i think everyone else can see them :scratch:

[Joe]

I can see 'em. Not that I'll be building whatever it is!

[Reid Malenfant]

I can see 'em. Not that I'll be building whatever it is!
:lol: Actually that's quite possibly a good idea from what i have just discovered :rolleyes:


Well some muppet has possibly messed this up & the muppet isn't me ;) It was mean't to be balanced mains, well i guess it would have been if both transformers actually produced the same voltage! As it is i'm not sure yet if the slight imbalance will force an earth current to flow, this will have to be tested before things proceed any further.

Honestly, never get a man to do a machines job! There is more than likely a turn or two extra on one of the mains primaries or one is a turn or two short, either way it's not balanced perfectly :steam:

Ok, here we are with both transformers powered up - looks ok at 235.8V.

http://i939.photobucket.com/albums/ad240/speakermark/test/235V-1.jpg

Here we have the (admittedly temporary) internal wiring. As you can see the transformers have different ends of the secondaries earthed. If you earth the same ends you'll just get about 118V from live to earth, though if you measured accross from both lives (and they were balanced precisely) you'd read 0V.

Just a few words here,this is why i mentioned you'd need a multimeter to test everything - because of muppets! You have to satisfy yourself that both (in the case of these transformers) red leads are actually connected to the same end of the secondary/s & are "in phase" before connecting them together. It's simple enough to measure between the two output sockets from red lead to red lead, i found 0V in both cases. Someone else might find otherwise - muppet strikes again!

http://i939.photobucket.com/albums/ad240/speakermark/test/TX2-1.jpg

http://i939.photobucket.com/albums/ad240/speakermark/test/TX1-1.jpg

http://i939.photobucket.com/albums/ad240/speakermark/test/internalwiring-1.jpg

This is the bit that you just don't even contemplate & only discover it when it's already rather too late in the day :rolleyes: All may not be lost yet though as i have a cunning plan. Unfortunately it's not something hardly anyone here is likely to be able to copy because it involves some seriously in depth electronics :eek:

Voltage output of each transformer 2359

http://i939.photobucket.com/albums/ad240/speakermark/test/TX2ononly-1.jpg

http://i939.photobucket.com/albums/ad240/speakermark/test/TX1ononly-1.jpg


Here incredibly simplified is the basic idea behind not only getting balanced mains but also virtually eliminating any distortion. This would take care of that slight inbalance & get mains distortion down to below 0.1%. It would involve two low voltage but high current power amplifiers & some circuitry to generate an ultra low distortion sign wave at the exact frequency & in phase with the incoming mains. If the output of either of the Live outputs is distorted each amplifier would add a signal to correct the error. The same or similar to the PS Audio Power Plant Premier except with a 6.6KW output :lolsign:

http://i939.photobucket.com/albums/ad240/speakermark/test/mainsregeneration.jpg


Before going mad though i'll stick a 1KW load on the transformers & see if any earth current flows ;)

[HighFidelityGuy]

This is fascinating stuff! :popcorn:
Thanks Mark.

[Barry]

The two (half) secondary winding voltages of 117.2 and 119.0V, being in antiphase will simply add to give you 238.2v

I assume you are worried about any common-mode noise only cancelling out to 1.5% (= (119 -117.2)/119). Is this the distortion to which you refer? Have to confess I don't understand the power supply arrangements for the two amps in your circuit

Apropos, your protection circuit operating within 1us. I do believe that such a circuit can be built, but feel that because it is so fast, it would be tripping all of the time. The RFI and noise will be of the order of 100kHz to 1MHz and I'm sure dropout and over-voltage spikes can be of this order.

Following this thread with interest.

[Reid Malenfant]

This is fascinating stuff! :popcorn:
Thanks Mark.
If only life was easy, a lot of people wouldn't be interested but thanks for the interest & the post! Take it from me that a lot will look & crawl under a rock & dismiss it all ;)

The two (half) secondary winding voltages of 117.2 and 119.0V, being in antiphase will simply add to give you 238.2v

I assume you are worried about any common-mode noise only cancelling out to 1.5% (= (119 -117.2)/119). Is this the distortion to which you refer? Have to confess I don't understand the power supply arrangements for the two amps in your circuit

Apropos, your protection circuit operating within 1us. I do believe that such a circuit can be built, but feel that because it is so fast, it would be tripping all of the time. The RFI and noise will be of the order of 100kHz to 1MHz and I'm sure dropout and over-voltage spikes can be of this order.

Following this thread with interest.
Hi Barry. I'm worried about the lack of pure balance forcing a current to run through the earth.

The circuit diagram was incredibly basic. If you understand about electronics you might understand this so you know where i'm going. I'll use a fast op amp to generate a square wave driven from the mains so in phase. I'll then filter this with something like a 24Db octave filter at a lower frequency (say 20Hz), this will result in a low distortion sign wave. I need to set the output voltage of this sign wave, so rather than using a compressor i'll create a circuit that alters the power supply to the square wave generator & keeps the AC waveform at a constant voltage. Phase shifting isn't a problem & i'll most likely need this after the filter.

The output from this will be in phase with the AC input to the transformers & so all i need to do is compare the clean AC generated signal to the output from the transformers. The amps will correct anything up to a certain frequency & that depends on the transformers in all honesty. It's a simple feedback network & will need to be SOT (set on test). I see no reason why up to 500Hz or beyond can't be corrected so we get a pure AC waveform rather than the 5%+ distortion cobblers that we are used to having supplied to our equipment. I might need to use 500W+ to do this but it's the little things that add up to make the big picture :)

You know that late night stuff when nearly everything is switched off & the whole system seems to open up? Well that's what i want all the time :lolsign:

[Barry]

If only life was easy, a lot of people wouldn't be interested but thanks for the interest & the post! Take it from me that a lot will look & crawl under a rock & dismiss it all ;)

Hi Barry. I'm worried about the lack of pure balance forcing a current to run through the earth. Why should it? Only if it was feeding another centre-tapped transformer earthed at the centre point, would any inbalance current flow. Just as it does in an un-balanced Wheastone bridge.

The circuit diagram was incredibly basic. If you understand about electronics you might understand this so you know where I'm going. I'll use a fast op amp to generate a square wave driven from the mains so in phase. I'll then filter this with something like a 24dB octave filter at a lower frequency (say 20Hz), this will result in a low distortion sine wave. I need to set the output voltage of this sine wave, so rather than using a compressor I'll create a circuit that alters the power supply to the square wave generator & keeps the AC waveform at a constant voltage. Phase shifting isn't a problem & I'll most likely need this after the filter.

The output from this will be in phase with the AC input to the transformers & so all I need to do is compare the clean AC generated signal to the output from the transformers. The amps will correct anything up to a certain frequency & that depends on the transformers in all honesty. It's a simple feedback network & will need to be SOT (set on test). I see no reason why up to 500Hz or beyond can't be corrected so we get a pure AC waveform rather than the 5%+ distortion cobblers that we are used to having supplied to our equipment. Can you please explain what you mean by distortion. Do you mean deviation from a pure 50Hz sine wave? I might need to use 500W+ to do this but it's the little things that add up to make the big picture :)

You know that late night stuff when nearly everything is switched off & the whole system seems to open up? Well that's what i want all the time :lolsign:

I can see what you are trying to do, and your circuit idea might work. However it does strike me that it's all starting to become very complicated. Why don't you build a crystal controlled oscillator to produce a clean 50Hz sine wave and use this as an input to a 500W power amplifier?

Regards

[Reid Malenfant]

I can see what you are trying to do, and your circuit idea might work. However it does strike me that it's all starting to become very complicated. Why don't you build a crystal controlled oscillator to produce a clean 50Hz sine wave and use this as an input to a 500W power amplifier?

Regards
Quite easy that one :) AC mains isn't exactly 50Hz, it deviates all over the place frequency wise but over a 24 hour period it must have completed the exact amount of sign waves that it would have if it was at exactly 50Hz.

Taking that as a given i couldn't just get an accurate 50Hz oscillator because it would very shortly go out of phase with the incoming mains. The circuit needn't be complicated at all, a fast op amp to generate a + & - square wave which switches at zero crossing & a dual op amp to make a low pass filter (24Db per octave). One more op amp to control the DC voltage feeding the square wave generator, i decided this would be the best way of regulating the AC output to keep it constant rather than a compressor which would add distortion. A few other bits & pieces & we are off. 500 or so watts & it'd be up & running.

If i was to do the same thing with an accurate 50Hz signal i'd need to regenerate the whole 6.6KW as the only way of doing it like the PS Audio PPP is to keep it in time with the mains ;)


Yes, when i say distortion i mean any deviation from a pure sign wave.

If what you say is correct & there isn't any earth current then there may well be no need what so ever to go to these lengths & balanced mains will probably be enough.

Cheers for your input Barry :)

[Reid Malenfant]

Well i have had a think, yet again :lol:

I decided i have a fairly easy way of getting them properly balanced to within a few hundred milli volts of each other. My secret weapon is a 60VA toroidal transformer that will have a new secondary wound on it. All i need is somewhere around 1.7V but at 29A which is within the 60VA rating.

Looks like i'll soon find out how many turns per volt i'll need & then i'll wind a load of 0.9mm enamelled copper wires in parallel to give me the correct current rating...

Back in a bit :)

[Ali Tait]

Interesting stuff Mark,just seen this thread.Have you tried the transformers as is and do you have any current flow to earth?

[Reid Malenfant]

Hi Ali, no i haven't but in reality i don't need to :) When Barry mentioned the Wheatstone bridge i had to cast my mind back a rather long way, he's correct, unless something is forcing an earth current to flow then there won't be one ;) The only thing that could do that is (if all is in order) an earth fault in the equipment that is connected to it.

Hopefully just about to win on ebay a 25Amp 30ma RCD :) I decided to add the 60VA transformer just to get everything balanced as best as possible, it costs me nothing but a little time to wind the secondary i need but it'll get the thing balanced better which will result in better interference cancellation. As the whole idea is to cancel out interference & it costs just time then imo every little helps.

These transformers have what looks like 5.6% regulation (at 3.3KVA) so the addition of a few volts will help out here again. From what i'm getting mains wise i can expect 240V AC at no load & 226.6V with a 6.6KVA load. Frankly i seriously doubt i'll ever use that much power, let alone continuously :lol: So i should be good for 235V AC in normal use (maybe more) & maybe a tad less when pushing the envelope (trying to go deaf).

This is going to turn into a full project now. Maybe at a later date i might go mad & add power amps to generate 6.6KVA of mains with less than 0.5% of distortion, but right now the balanced mains is the aim :)

Bests to you fella..

[Ali Tait]

Ok,can you try it as is and tell us how it sounds?

[Reid Malenfant]

Yes, give me a day or so though, i'm not as fit or have the time like i used to.

Everything in the room at the moment is powered from the same point which is a spur from the fusebox, so i'll swap it over & have a good listen & let you know. A few days or so should see differences & i'll report them ;)

[Barry]

Hi Mark,

I like the sound of the 60VA transformer. Can you sketch a circuit diagram for us so that we can see what you propose doing?

I don't think regulation is going to be much of problem; it's the mutual cancellation of common mode noise that is the more important benefit to be gained with a balanced mains supply.

Keep us informed. :)

Regards

[Reid Malenfant]

I like the sound of the 60VA transformer. Can you sketch a circuit diagram for us so that we can see what you propose doing?

I don't think regulation is going to be much of problem; it's the mutual cancellation of common mode noise that is the more important benefit to be gained with a balanced mains supply.

Not a problem Barry, i'll get a sort of circuit diagram up as soon as i can. I just wanted the two phases to be as equal as possible, not my fault that monkeys wound the transformers :lol: This is my little cure though & you'll see how shortly ;)

Bests to you :)

[Ali Tait]

Did you bolt the trannies together?

[Reid Malenfant]

:) No they'll be both going into an MDF box, my favourite medium as it's easy to work with. I'll drill it for cooling but frankly as the transformers are potted up they'll not need a lot in that department ;)

I'll simply turn the whole lot upside down & bolt the transformer enclosures to a base board & make some kind of enclosure around them...I'm sure you'll see pictures.

Merlin Gerin 25Amp 30mA RCD on it's way to. Not sure how i'll mount it but i'll suss it out once it gets here & i take a look at it ;)

[Ali Tait]

Room between the potting compound and the lid to squeeze it in?

[Reid Malenfant]

Room between the potting compound and the lid to squeeze it in?
The extra 60VA toroid? Yep not a problem :) I could fit a 200VA in the extra space :lolsign: Seeing as the lower voltage output transformer will in theory be supplying near as damn it the same current as the slightly higher voltage one all i need to do is connect it to the same Live & Neutral supply. I will however give it it's own fused supply ;)

VA ratings should balance things out, just like it should cancel interference :lol:

[Ali Tait]

I was thinking about the RCD.If there's room,cut a square in the lid to push it through (assuming it's the type that fits in a CU)

[Reid Malenfant]

I'll get back to you soon Ali ;) I'm off into Alice in wonderland territory, finding out how deep the rabbit hole goes......

Until tomorrow chap :) But there should be no problem fitting it in with an MDF box, mounting it might be a different matter..We'll see :lol:

All the best, Mark.

[Ali Tait]

Okay,let us know you get on.

[John]

Hi Mark
How this go did you finish the project if so has made a improvement

[Reid Malenfant]

Hi John, tbh i'm waiting for a friend to get some free time & come over & help me clean out the garage so i can use my table saw :(

As soon as i get access to it i'll be able to make a decent, safe housing for it as i'm not prepared to run anything on it without it being cased up & have all the protection installed.

I'd sort the garage out myself if i could but my back is in no way up to the task, one wrong move & i might literally be on my back for months :eek:


Don't worry, i'll return to the task in hand just as soon as i can :trust:

[Reid Malenfant]

Well i have decided i'm going to go the regenerator route while i'm about it :) Before i do though i'll lash the thing up & give it a listen for at least a few days so i can see what difference it makes & i'll report back. As long as i'm not going to be using loads of power (which i won't be) then it should be safe enough without a shed load of protection.

It may seem a little mad but in reality it won't cost me a lot to do this at all. I just purchased 2 x 300VA 0-12V 0-12V transformers for the princely sum of £20 approx & the only other things i'll need to buy are a couple of very high capacitance electrolytic capacitors. Looks like i'll be using boy racer car caps at about 1 Farad each as they are cheap enough :eyebrows:

I have all the heatsinking & electronic components that i'll ever need to knock it together so i might as well go for it in reality ;)

It should be fun whatever happens :lol:

[Ali Tait]

Good stuff,keep us posted.Wahey,3000th post!