Balanced Isolation Power Supply

[Ronald]

Does someone have any experiences in Balanced Isolation Power Supply? Thanks a lot!

[planner]

I have 3 sets Sound Affairs Plixir Balanced AC. One for the analog equipments (3000 watts), one for digital equipments (500 watts) and one for music server (500 watts). Biggest benefit for me is no more ground loop issues. And it's super quiet even though I'm using tube amp and tube preamp. All my interconnects are unbalanced as well.

 

 

Sent from my iPhone using Tapatalk

[sandston]

I built my own balance power supply after doing quite a bit of research on the interweb. I had a ground loop issue that I was unable to isolate and the balanced power knocked it right out. Im sure there are things that can be done to improve it and I welcome comments but it solved my problems and I havn't notice any change/drop in dynamics, just a darker, quieter background. If you want to know how to cut out the aluminum for the outlets let me know and I can post a step by step.

Balanced Power Schematic-Model.pdf

[Speedskater]

A Neutral should never have a fuse.

But a linked circuit breaker is acceptable.

[sandston]

My amp tech, who also builds BPS's recommended fusing both. Don't remember why beyond protection of course. Can you elaborate on why the neutral in this application should not be fused.

[watercourse]

Looking good Sandston! Does the transformer itself hum or vibrate?

[One and a half]

My amp tech, who also builds BPS's recommended fusing both. Don't remember why beyond protection of course. Can you elaborate on why the neutral in this application should not be fused.

 

Sorry, rules are rules, the neutral conductor must not contain a fuse is regulation enforced by laws in at least 3 regions of the world off the top of my head.

You can break the neutral, provided the active or hot is broken at the same time as Speedskater mentioned, with a double pole circuit breaker.

For a toroid, recommend an EU classified type D breaker. Toroid can draw 10-15 times rated current for 1/2 cycle 25ms for 50Hz 30ms for 60Hz.

[YashN]

I built my own balance power supply after doing quite a bit of research on the interweb. I had a ground loop issue that I was unable to isolate and the balanced power knocked it right out. Im sure there are things that can be done to improve it and I welcome comments but it solved my problems and I havn't notice any change/drop in dynamics, just a darker, quieter background. If you want to know how to cut out the aluminum for the outlets let me know and I can post a step by step.

[ATTACH=CONFIG]28284[/ATTACH][ATTACH=CONFIG]28287[/ATTACH][ATTACH=CONFIG]28285[/ATTACH][ATTACH=CONFIG]28288[/ATTACH]

 

Very cool, thanks for sharing. Plitron is great.

[jabbr]

Equitech 2Q: 72 lbs and worth its weight

[One and a half]

I built my own balance power supply after doing quite a bit of research on the interweb. I had a ground loop issue that I was unable to isolate and the balanced power knocked it right out. Im sure there are things that can be done to improve it and I welcome comments but it solved my problems and I havn't notice any change/drop in dynamics, just a darker, quieter background.

 

The more I looked at the photo, the more there are items you need to address, the fixes are simple and won't take long.

 

Here's a list of improvements that make the power supply comply to safe practices:

 

1. Incorrect wire colour used for ground connections, should be green/yellow at the receptacles.

2. The ground cable from the IEC inlet is not connected to the frame of the enclosure. Should any item leak to ground, the upstream fuse will blow instead of livening up the chassis and sending you half way across the room.

3. The output of the transformer has two 'legs' both are hot, 60V to ground, there's no neutral.

4. If overcurrent protection is selected for the output, then there needs to be an overcurrent device in both legs, not just one.

5. The fuse in the neutral after the IEC inlet filter needs to be removed.

6. Since all the cables are the same colour, difficult to see if all the grounds end up at the same point physically. This is most important for noise suppression.

7. The output of the transformer must have a GFCI, 30mA rated.

8. The hinged lid must have a cabled ground connection, the continuity of the ground to a metal structure cannot be used by fixing screws only.

9. The wiring for the monoblocks need to be higher say 10AWG since amplifiers draw 25% duty cycle. In order to produce real watts, the amplifier will draw more current, overheating wires.

10. The cabling for the remainder of the system is 12AWG? Can't tell.

 

What's the intended function of the 0.47uF caps, spike suppression?

 

The transformer used is a toroid from the standard range from Plitron with regulation that could be as high as 10%. For Power conditioning, Plitron recommend their NBT range of toroids to suppress high frequencies. Bifilar wound balanced transformers are also available from Plitron that have a tighter tolerance between the two secondary windings. This improves the harmonics and assymetrical voltages rejection on the secondary. The problem then occurs that the noise has to end up somewhere, and finds itself on the primary windings, where the transformer can rattle. In this case, fit an isolation transformer or a choke to slow the current spikes and mitigate the noise.

The NBT and bifilar transformers are difficult to manufacture and are not simple = costly. They need to absorb energy, so the frames are consequently larger than a standard transformer to dissipate the heat.

Balanced Power Schematic-Model - REV 1.pdf

[Speedskater]

More of the above.

a] In a US type system, the output of the transformer must have a GFCI, 5mA rated. Only industrial circuits use high current GFCIs.

b] What's the intended function of the 0.47uF caps, spike suppression?

Any capacitors across the AC line need a power line 'Y' rating.

Any caps to the Safety Ground need an 'X' rating.

c] If there are no electrical components mounted to the lid, skip the jumper wire.

d] Oversize wires are always good but most 20 Amp circuits only need 12AWG.

With the exception of Class A amps, continuous current of power amps when playing music or movies is much less than their rated current.

[YashN]

Equitech 2Q: 72 lbs and worth its weight

 

jabbr and sandston, have you tried putting your balanced iso power supply on a seismic isolation platform - at least a triple ball-and-cup arrangement?

[jabbr]

jabbr and sandston, have you tried putting your balanced iso power supply on a seismic isolation platform - at least a triple ball-and-cup arrangement?

 

I have it on the Diament isolator, marble over balls & cups over plywood over bicycle tire.

 

I also have a few Topaz isolators that I use in various places. They are very solid.

[sandston]

Thanks for all the feedback. The saftey suggestions are all very much appreciated. To answer a few of the questions here.

 

My transformer does not hum or vibrate. I have read that some people have had this issue however this may have been with lesser quality toroids or caused by the amount or presence of DC offset in your mains. I explored the idea of adding a DC blocker but since my unit is quiet I left it out as there is some debate over the use of these in this application as I recall.

 

PLease keep in mind my technical knowledge of the hows and whys of balanced power works is very limited. I had a problem (ground loop noise) I searched for a solution which lead to using balanced power as a solution. I read, I copied, I tested carefully, I implemented, it worked.

 

Beyond this the capacitors as I understand it block noise being introduced to the mains from the power supplies in each individual component. One of the main articles i referenced I used to put together my BPS can be found here.

 

DIY Balanced AC Power - Home Theater Forum and Systems - HomeTheaterShack.com

 

The author states the caps are ," used as differential noise filters". This seems to be a common application of the capacitors in BPS be it DIY or commercially produced.

 

As far as the X and Y capacitors go this appears to be brought up all the time. Here is a quote from another thread I referenced. (Full Thread here DIY Balanced AC Power Conditioner)

 

"I use 600V rated Auricaps. X rated caps are normally used in applications across the AC mains. There has been debate about the use of Auricaps. The manufacturer, Audience lists AC line use in their application notes. I have been using them for the past 5 years without incident."

 

Perhaps its being a little cavalier on my part to trust these sources so please do this at your own risk. Looking at pictures found online manufacturers of BPS such as Equitech, Balanced Power Technologies, Transcendent Audio and Blue Circle Audio all use identical topology with capacitors as filters.

[sandston]

Yes the neutral line is mislabeled on my schematic. All grounds terminate at the same location. I read on several forums and was told by "my tech" not to ground to chassis. The outlets are isolated ground so no part of the circuit is in contact with the chassis itself.

[One and a half]

Yes the neutral line is mislabeled on my schematic. All grounds terminate at the same location. I read on several forums and was told by "my tech" not to ground to chassis. The outlets are isolated ground so no part of the circuit is in contact with the chassis itself.

 

I would recommend your tech revisit the recommendation of the grounding of the chassis of the power supply. Unless the appliance is deemed double insulated, there's not a requirement for a ground connection. BUT. Double insulated appliances have no ground in their cable, and are two prong devices.

 

Your tech may have argued a case for an instrument or technical ground, which has merit for audio use. In this case, all the audio devices grounds are connected to one point and usually some distance between is connected to the protective ground as in page 46 of the Middle Atlantic paper.

 

 

Any metallic enclosure that contains electricity that's dangerous to people must be grounded. No exceptions, can't put it any plainer.

[Speedskater]

Your tech doesn't understand an isolated ground system.

[Speedskater]

jabbr and sandston, have you tried putting your balanced iso power supply on a seismic isolation platform - at least a triple ball-and-cup arrangement?

If the room is shacking, you have bigger thing to worry about than the AC power transformer.

[jabbr]

If the room is shacking, you have bigger thing to worry about than the AC power transformer.

 

All rooms have vibrations -- surprising amounts-- electronics are often very sensitive to vibration -- particularly those with clocks but also: http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4319130&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D4319130,

 

http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=7208876&url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel7%2F20%2F7330079%2F07208876.pdf%3Farnumber%3D7208876

[plissken]

All rooms have vibrations -- surprising amounts-- electronics are often very sensitive to vibration -- particularly those with clocks but also: IEEE Xplore Abstract - Vibration-Induced Phase Noise: It Isn't Just About the Oscillator,

 

IEEE Xplore Abstract - Vibration Reduction of Inductors Under Magnetostrictive and Maxwell Forces Excitation

 

ery often, actual, measured levels of vibration-induced phase noise in signal generation equipment are significantly higher than predicted

 

Good thing this is power generation and not signal generation.

[YashN]

If the room is shacking, you have bigger thing to worry about than the AC power transformer.

 

Doesn't have to be life-threatening shaking to have an effect on SQ - cf. presentation / charts by Townshend.

 

This said, I generally agree with you in the case of strong shaking...

[One and a half]

ery often, actual, measured levels of vibration-induced phase noise in signal generation equipment are significantly higher than predicted

 

Good thing this is power generation and not signal generation.

 

In keeping with vibration and reducing the effects of it on signals, how come, an active speaker has a lot of vibration but the amplifier is inside been shaken to hell.

Is this some compromise to the SQ or is the amp fully potted to avoid vibrations?

[jabbr]

ery often, actual, measured levels of vibration-induced phase noise in signal generation equipment are significantly higher than predicted

 

Good thing this is power generation and not signal generation.

 

Everything is signal generation

 

If I were limited in the equipment which I could vibration isolate, I would pick the DAC as 1), 2) and 3) but when I got my rollerballs we got a great price for a whole bunch of them and so I isolate all my equipment including servers and RAID arrays, hard drives are notoriously vibration sensitive as well. Also anything with a crystal is well piezoelectric...

[plissken]

Everything is signal generation

 

If I were limited in the equipment which I could vibration isolate, I would pick the DAC as 1), 2) and 3) but when I got my rollerballs we got a great price for a whole bunch of them and so I isolate all my equipment including servers and RAID arrays, hard drives are notoriously vibration sensitive as well. Also anything with a crystal is well piezoelectric...

 

 

 

Hard drives are vibration sensitive when placed, en-mass, with other hard drives in a very dense enclosure. These enclosures are designed to reduce the rotational vibrations that, as an array, will start producing.

[plissken]

In keeping with vibration and reducing the effects of it on signals, how come, an active speaker has a lot of vibration but the amplifier is inside been shaken to hell.

Is this some compromise to the SQ or is the amp fully potted to avoid vibrations?

 

Same goes for passive speakers because 99.9% of the time the X-over is in the same enclosure as the speakers.