Building a Balanced Power System

[DiminishingReturnsOfficer]

Due to some recent trouble with electrical interference and dirty power, I want to share what I've learned about building a Balanced Power System. The primary sources for information are eponorama Web Pages on Balanced Power and Eliminating Ground Loops, as well as 30 or more RFI's from Eric of Plitron's Engineering Department. I will also provide a link to an article from "Secrets of Home Theater and HiFi" on building the unit using the same transformer I use.

Let me begin with a beief explanation of Balanced Power: balanced power is identical to balanced interconnects. Much the same; typical 120V power with one "hot" leg and one grounded "neutral" leg can be called "Single Ended Power". Balanced power is only obtainable with a transformer. Any transformer; step up, step down, isolation (1:1) will provide galvanic isolation. This means that the electrical connection from the power line to the audio gear (or secondary of the transformer) is broken. In addition to breaking the electrical connection, it also provides some protection from surges. With a Balanced Power transformer, you have a grounded center tap. Instead of providing 120V between a "hot" and grounded conductor, the Balanced Tranny gives 2 legs of 60V to ground with 120V between legs. Any electrical signal seen as having the same polarity on both legs will be cancelled because the 2 legs are 180 degrees out of phase. This is known as "Common Mode Noise Rejection".

One possible drawback to using a transformer is as follows: a transformer has a characteristic known as inductance. Inductance opposes changes in current. This is not particularly desirable for power hungry amps that require large amounts of current instantly to provide for dynamic swings in music. These effects are mitigated to some degree by: 1) Power Supply capacitors in the power amp. help to supply this current during peak demand 2) Torroidal Transformers are much better at supplying current peaks than traditional units. The type of criticism often heard of power conditioners using transformers is that the resuling audio sounds "choked out"; a particularly good adjective since a "choke" is another term for a coil. You will have to decided for yourself whether a Balanced Power system makes sense for you based on you listening tastes and system constraints. I might point out that every residence that derives power from the grid has power that has already gone through untold transformers since it's origins in a generator. Just about all audio components have their own power supply transformers (good ones are normally torroids). All I'm saying, is don't fool yourself into thinking that you're eliminating inductance by avoiding balanced power, you're not.

I choose the Plitron 8575 2000VA for my project. This is the same one featured in the "Secrets of Home Theater and HiFi" article. Plitron makes high quality torroidal transformers for many highly regarded audio manufacturers; Bryston, Torrus Power, and BPT just to name a few. It's possible that a torroid of slightly better quality and better current response would be one of Equi=Tech's Model Q transformers. In my opinion, cost is prohibitive and the performance increase is slight. The 8575 weighs 27 pounds and costs a tad over $300 after shipping from Canada. 2000VA is plenty for most medium to large stereo systems. It could easily provide power for a system consisting of 500W Monoblocks and associated gear.

This was intended to be a "how to" article, but I think I'm getting to the upper limit in length for one posting (to hold people's interest anyway). I'll do another right after this that will get into a material list and step by step build instructions. As promised, here's the link to the "Secrets" article as well as the epanorama postings:http://www.epanorama.net/documents/groundloop/problem_solving.htm

http://www.hometheaterhifi.com/volume_10_3/feature-article-isolation-transformer-8-2003.html

 

[wgscott]

I predict this will be a highly-linked threat and a valuable resource.

 

Cost and (excess) energy consumption are two things that concern me (or inital costs and operating expenses).

 

[DiminishingReturnsOfficer]

240V power from the Utility is not balanced. The 2 legs are both 120V to ground, correct. If derived from 3 phase; like most services are, they are essentially 120 degrees out of phase. Your inverter has to work with Utility power, so I'm going to guess that it would have to be very similar in waveform. This was difficult for me to understand at first too because, my Physics schooling taught me that 2 waves 180 degrees out of phase should cancel eachother. This is true with sound waves to some extent. The crucial piece to a balanced power system is the ground reference (0 Volts). It is established at the center point of the transformer secondary winding which results in a phase voltage of 120V and a line voltage of 60V. If you connected one of these legs to ground (60V); instead of line to line(120V), you would have unbalanced power. 2 things are essential; 1) the ground reference, and 2) the phase angle (180 degrees).

As far as cost, I'm putting a material list together that will include prices. The transformer itself is the biggest cost ($330 shipped from Canada). A nice looking instrument case, supply cord, SPST switch, LED pilot light, audio grade receptacles, and power strip bring the price to over $600, but makes for a really nice unit; just as good, functionally, as the BPT or Torrus Power units selling for well over $2,000. In your case, you could do as barrows suggested and tap the Utility service ahead of the Grid Synchronizer and feed the Plitron (or other Balanced Torroid) with 120V (not 240V) from the grid. This way you'd be balancing fairly "clean" power to begin with. You could put the tranny in a wall mounted enclosure (example 12"x12"x8" Hoffman box). Bring a 20A line from the Utility power panel to feed it and then tie into the circuit that currently feeds you stereo (disconnecting it from it's present source, of course). Also it is not permissable by the NEC to make this connection in the service panel. This would eliminate most of the additional cost and put the tranny far enough away from the stereo so as not to be heard!

 

[wgscott]

Again, thank you! I keep learning a lot.

 

[Brand B]

I built one of these several years ago based on a thread at another forum using the same (I think) transformer when they were having a clearance on them.

 

A comment on the buzzing. Mine is dead silent EXCEPT when my family runs our toaster. It uses an SCR chopping dimmer to set heat level. When this is on, my plitron then starts making a pretty audible noise. So if yours is constantly making a similar noise, it may be a similar DC offset issue such my toaster creates. Might be worth turning off lights and unplugging other appliances to see if you can locate a culprit.

 

I went a little farther than just the basic isolation with mine, adding some RFI filters and other such to each of the outlets.

 

Mine pretty much serves the front end of my HT, including DVD deck, PS3, preamp, TV, DVR, apple TV, active XO, sub equalizer and the smaller of my amps. As you say, noise floor is dead zero, even with volume cranked (nothing playing) and ears to the tweeters.

 

Also picked up a while back a second smaller transformer good for about 10 amps from DIYCable that was intended for us in a hypex class D amp, but being plus/minus 60Volts will work as a small unit for the front end my 2 ch system. Right now it's just running through a tripplite rack mount filter, and the noise floor is no where near as good as my other system.

 

 

[DiminishingReturnsOfficer]

Item Part / Stock Number Price

 

2000VA Torroidal Plitron 8575-XO-02 $290.29

Hammond Enclosure 1426Q-B $45.80

Eaton 20A SPST switch 8492K1 $8.00

(3) ATL Audio Grade Receptacles $45.00 ea.

Power Strip PS1 *in this case it was free thanks to Lee from Cryo Parts / Locus Designs*

Square D Grounding Bar PK7GTA $7.00

Amp Cord Connector DB-3 $6.95

IDEC 22mm. Green LED APW299D $20.00

Hubbell Hosp. Grade 20 A Plug HBL8315CT $27.00

Cryo Parts 12/2 Bulk Wire $10.00/ft.

4"x1/2" mounting bolt (for horiz. mount) $3.00

1/2" nut, assorted washers $2.00

5/8" thick rubber mat (on hand)

10 AWG Teflon Insullated Wire from spool $1.99/ft.

14-10 AWG Ringed Crimp Terminals $3.50/box

Tyco Electronics Studded Split Bolt 2058729-1 $4.00

(4) chair leg caps (used for feet) $5.00

(4) 7/8" rubber chair leg caps (go over the metal caps to make great feet) *the flimsy rubber squares provided by Hammond are not sufficient for this purpose $0.75 ea.

(12) 8/32 machine screws $3.25

(12) nuts for the above $1.50

 

Odds Ends and Accessories

- wire nuts, electrical tape (these make a better parallel connection than terminal strips, lugs, solder, ect.)

-heat shrink

- I like to use 3/8" copper braid over power cables for additional shielding (land shield only at plug end)

- I also used Kevlar cord covering

- you can get a nice rubber switch boot for the toggle switch that really dresses it up IMO (Amp N1030A for around $6)

- the LED is by no means necessary, in fact, the above post makes me wonder if it doesn't contribute to noise. It looks nice (green), however, and does double duty as a night light!

- if you want to mount the power strip to the enclosure, you'll need a 3/4" minimum chase nipple and locknut.

- this is just a list of what I used. I don't mean to imply that you have to use the same brand or style that I use, but good luck trying to find a fairly inexpensive instrument enclosure, other than the Hammond 1426Q that will accommodate the 8575.

 

 

 

Tools

 

You'll need one of each type of screwdriver, a 8/32 tap, an adjustable wrench and 1/2" and 5/16" nutdriver, a drill, a heat gun or torch for shrink, a "uni-bit" (stepped drill bit), and wire cutters, strippers, and crimpers. I managed to tear my whole bag apart doing this, but these are the basic tools I used most often.

 

In addition to the 8575, Plitron also makes other balanced power torroids (bigger and smaller). They can also custom build almost any size tranny. Note: if you get a different tranny than the 8575; make sure it is a Balanced Power troansformer with a 120V secondary. Medical Isolation Transformers are not exactly what you want for audio. Make sure that you get one with a CENTER TAPPED 120V SECONDARY. Very Important: if you get a smaller trnasformer than what is listed here, you should think about getting a fuse or in line circuit breaker for it. A small transformer could overheat on a 15 or 20 amp circuit unless you're 100% sure of your power requirements and you don't plan to expand (ever).

 

[DiminishingReturnsOfficer]

You know, come to think about it, I use a LED pilot light on my Balanced Power Unit. Obviously, LED's are diodes much like the SCR on your toaster. Fortunately, I wired the switch to allow me to bypass the LED. It is connected to the primary side. All I have to do is throw a switch to test this (well, that and keep my fingers crossed that the breaker does't trip). By the way, have you had the breaker tripping problem with your unit? I was thinking if the 8575 was a subtractive polarity transformer, I could also place a second LED on the secondary side. Using both Forward Current LED's, this way the LED on the primary would be canceled out by the one on the secondary. Anyway, I've spent more time talking about it than its worth when I can just flip the switch to find out for sure. I'll be sure and post my findings. Thanks Brand B, and I'm glad you like your Unit. What type of enclosure did you choose?

 

[Raym87]

If you are in the UK you could always have a look here...

http://www.airlinktransformers.com/balanced-power-supplies.asp

 

 

[One and a half]

I would add:

 

- High voltage insulation tester

- Cat 4 multimeter

- Overcurrent protection

- Ground Fault protection

- Wire nuts tend to loosen the joints when subjected to vibration, or even on off cycling. I would avoid them.

 

The toroidal transformer will always draw a lot more inrush current due to its construction. I got around the problem of inrush using a standard EI transformer 1:1 to act as a choke in front of the toroids. That transformer is also shielded, so common mode noise is gone. The input circuit breaker is a standard 10A (for 240Vac) gl curve, and it holds in every time for a 2kVA transformer.

 

But if you have poorly designed loads, you can setup interference all over again enough to pollute your new conditioner. As another poster suggested, it's best to isolate each load from each other if you can, but still provide a large transformer to allow for peaks for power amps and having the advantage of a balanced supply.

 

The only transformers I know of that buzz loudly even at no load are when they are supplied with rich harmonic content in combination with poor construction techniques. Potted toroids are about the quietest.

One trick of making US made EI transformers quiet is to load them so that become hot, then switch off the power and madly tighten every screw in the laminations before the transformer starts to cool. I haven't needed to do that with Europe made transformers however.

 

[steve2701]

You may want to double check if you consider 2kva is enough for big power amps. My personal thoughts are that it may well not be to allow for instant very high peak demand on transients. It is fine for ancillary stuff that has a steady demand, but I would be wary of a small one like that on an amp. I use seperste balanced ones on my pre and cd, but the amp will be fed by one next year when I can afford for it to be built. It will be slightly larger than 2 kva for what I have in mind. For us in the uk we cannot wire the mains as balanced, it is against the regs. The traffo need sockets on them to plug the kit into, and then they usually hum instead! I would love to have balanced mains but have to do

It slightly differently. It does definitely work for me, really lowering the noise floor. Please forgive any spelling mistakes, I'm sat on a riverbank

In the middle of nowhere and my fingers are cold and this keypad is tiny! I hope yours works out well for you.

 

[barrows]

of points.

 

1. Do not worry about noise from LEDs, these are extremely low noise devices. In fact, they are used as low noise voltage references in some of the finest audio gear in the world specifically because their noise level is so low.

2. If I was concerned about noise, I would not use twisted wires/wire nuts for connections. Any place electricity has the chance to micro arc it will, and any arcing produces noise (ever hear an arc welder, big version of the same idea). Terminal blocks, with crimped and soldered spade connectors on the wiring will producer a better connection and lower noise. Even better would be thick copper buss bars, drilled and tapped for spade connectors (crimped and soldered to the wiring)

3. If you have big power amps, make sure your transformer is overated for the peak current demands that these amps need, otherwise you may be starving them for current on musical peaks. A 2000 VA transformer described here will be adequate for systems up to about 800 watts peak load (you need headroom) but if you want to retain the dynamic potential of your system, you may want to consider a much bigger tranny running on a 20 amp circuit (and an inrush limiter circuit to allow you to switch it on without problems).

4. Note, that even the best transformer will increase distortion on the line, the only thing that can fix harmonic distortion is an AC regenerator.

5. If your transformer is making noise either it is not high enough quality for the job (unlikely with Plitron) or you have a significant DC offset (this is likely to change, be loud sometimes, quiet others, as the load on your AC line changes).

 

[Brand B]

I actually sort of fabricated my own enclosure. Where I work we have full shops of various sorts, in addition to material remnants and such. And where I live it gets up to 115° in the summer, so the spray on enamel spray paint basically becomes a baked on powder coat simply by using black and leaving it the driveway for about 1/2 an hour.

 

My big power amps don't plug into the balanced unit, as I would almost need one for each in my HT. My mains run off a Crown macrotech 2400, then a second one of those for the woofers, and then my subs are running off a macrotech 3600.

 

So basically everything else, including the surrounds amp, runs of the plitron. But still, that is enough to take all the noise out of my system.

 

My transformer is just mounted with a rubber bat, in the normal horizontal position, with a blot through the middle and a washer with another piece of rubber up top. I just went and double checked by putting my ear right on the case, and there is no residual buzzing I can detect (no one is toasting a the moment). However, going back and reading the thread when I built it, it did hum somewhat when I first made it, as did others using the same plitron. However, not the loud buzzing I get with the toaster. So I think Barrows has it right about the DC offset. That is certainly the case with our toaster, and it also occurs when dimmers are engaged. Note that these are on completely different circuits from my HT, so this comes through the panel even on a dedicated circuit. I would guess there's something in your house causing it, hence my suggestion to start eliminating lights and appliances to track it down.

 

As for the on/off breaker tripping, I don't shut my unit off, as the DVR records things at odd hours, and I have an active XO in the rack that the builder advised me ought not be power cycled too often for longevity. So I am wasting a bit of electricity as the unit is not 100% efficient. I have not had it ever trip a breaker on the occasions where it is turned back on, i.e. after power outages, when I unplugged everything for lightning recently, or when I shut it off due to extended absences like vacation.

 

[Brand B]

Oh, and I verified the plitron was an 8575 same as yours, so behavior should be the same.

 

[DiminishingReturnsOfficer]

Thanks to Brand B for the input. I've contacted Plitron about the problem and expect a response from them tomorrow. They're always great about getting back to me in a timely fashion. Strangely, ever since I cycled my unit from the position where the receptacles are hot and the LED is on, to the position where the LED is off and the receptacels are hot, and back; it has been much quieter. I can still hear it. If you're saying you put your ear to the case and didn't hear anything, I've still got a problem. I've never heard the term "DC offset" used with regard to mains level power, only for line level and post amplification phenomena. I'm not saying its incorrect, and I think I understand what barrows means; that there is a DC component to the AC waveform on the primary. This DC, depending on polarity, decreases the potential on either the positive or negative half of the sine wave. It will be canceled due to common mode rejection on the secondary side, but perhaps the noise is limited to the primary. Of course, the type of "noise" I'm talking about is a mechanical buzzing; not anything I hear on the system. My stereo is on its own dedicated 20A circuit, and the Balanced Power Unit is the only thing plugged into the circuit, I have my TV, Marantz SACD player, Bel Canto DAC 3, YBA 2 Delta preamp, and Conrad Johnson power amp connected to it. Even with every other circuit in the house turned off at the panel, I still get the buzzing. I think you said that your toaster is on a different circuit, so I went ahead and tried it. The problem is, I live in a large appartment building, so my neighbor could have the same toaster you have, along with dimmed flourescent lighting, 3 lazer jet printers, a Unix server, the first CD player ever made, and 15 cell phones charging, and it effects me. Having wired appartments in the past, I'm assuming that each adjacent unit is on a different phase; A,B,C; 1,2,3 and so on. This really doesn't matter because, more than likely, the noise generated by these non-linear loads is coming through on the neutral which is common to every 120V circuit in the building. Being derived from 3 phase power (208Y 120), the single phase services are prone to harmonic interference every 120 electrical degrees. This may figure into it too.

As far as the wire nuts, I now have 2 knowledgeable people advising against them. Point taken, I am aware of micro-arching as a potential problem particularly in high voltage applications, but I'm aware it happens at household voltages as well. I've taken apart old splices and seen evidence of this for myself. In fact, I got a nasty shock once because the wire nut on a 277V shared neutral was so brittle, it disintegrated when I touched it. The thing is that I don't have room inside my enclosure for terminal strips; certainly not for a bus bar! We're talking about #12 AWG wire, so using a 32 Ton press isn't an option. The insullated crimp lugs (Stay Cons) don't crimp down any tighter or provide any more mechanical pressure than a properly installed wire nut. We'll just have to agree to disagree on this one. That being said, I do certainly appreciate your input, and hope that this thread will be beneficial to anyone considering building a Balanced Power System.

 

[One and a half]

on an active discussion which reveals more as the thread goes on.

Hand held crimpers are OK for 10 guage wire, 30tonne crimpers not really necessary, besides being difficult to maneuver into a small size box.

Looking at DC offset, news to me on this term, industrially for instrumentation, there's never a zero, it looks like on the AC waveform, there's no distinct zero volts, it's always above by a small amount. On an audio waveform, it's disaster, however we haven't go that far yet.

 

When I talked of harmonics causing buzzing on transformers,that's what they do. Harmonics come from mainly switch mode power supplies, rectifier circuits (as in amps) and inverter circuits. The only real way to determine if harmonics is a problem is to measure them. If you can get hold of a Fluke 43B, it can tell you what harmonics are on your system, and will also give you a snapshot of your raw AC waveform. You can also use that meter to verify the balanced supply contains no noise.

 

In all sincerity, by the time you design a filter to remove harmonics, the utility company will change things and you are back to sqaure one. Harmonic filters come in all shapes and sizes, some of them are very clever (http://www.mirusinternational.com/) but for audio where we seek purity of the supply, the most likely cleanest source of supply would be a UPS (without the battery), with a toroid on the output, and individual isolation transformers for each connected device that contains a switchmode, especially a computer power supply!

 

You would have to watch a few things, such as inrush, but the UPS should go into bypass while the transformer is energised, and then run in normal inverter mode when the inrush is over.

 

Commerically there a a few conditioners for audio, but they are not cheap, but DIY far more rewarding and enjoyable.

 

http://www.accuphase.com/model/ps-1210.html

 

http://www.psaudio.com/ps/products/detail/power-plant-premier_copy?cat=power

 

 

[DiminishingReturnsOfficer]

Just in case anyone wants to build this unit, and goes as far as to aquire the Plitron 8575-XO-02, let me give a brief summary on how to connect the leads and put all your materials together into a nice finished unit. If you choose to use a captive power cord (as I did), its advisable to make it at least 6' long. Another alternative is to use a 20A IEC inlet with a pre made power cord. Something I recongnized early on in this project is that it is not easy to make near perfect (industrial quality) holes in any shape other than round. IEC inlets as well as common duplex receptacles use semi-rectangular or oval(ish) holes. If you don't want to send your chassis out to a machine shop, and you can't do spot on work with a Dremel, you're somewhat limited. I hope that somebody can correct me on this, but I was not able to find any knock outs for these common shapes. If such a thing were available, surely Greenlee would make them. Well, they don't, and I don't know of any one else who does. If you spent a good bit of scratch on your chassis, insist on using IEC's and/or chassis mounted receptacles, and don't have complete confidence in your skill with a Dremel tool you might want to take it to a machine shop. You can use single round receptacles, but you'll use twice as many. I decided, instead, to use a cord grip entry connector for my power cord. I mounted an audio grade power strip housing directly to the back of the unit and chase nippled through with the feeds for the receptacles. I think the toggle switch took a 1/2" hole with 2 #6 mounting holes (one above and one below), and the LED took a 22mm.. One thing that I strongly recommend doing is getting a studded split bolt; drill a hole just big enough for the stud. You'll want the stud inside the chassis. Then, bring a grounding jumper to it on the inside and terminate it with a ring Stay Con.. This way the split bolt is on the outside making a very easy termination point for external grounding jumpers. If you do choose to star ground your equipment, this is where you want to do it.

You'll want to use some pretty substantial feet on the bottom because the tranny weighs almost 30 pounds. Maybe you've got an old amp that you could swipe the feet from, or you could use the method I discribed in the Material List. You could also elect not to attach feet directly to the unit and sit it on Audioquest Q feet or HRS Nimbus feet with spacers. This unit will be a critical part of your system, so you want to dampen it properly, and (presumably) have it look nice too. It doesn't have to look like a $20K amplifier, but you don't want it to look like a Middle School Science Project either.

OK, now for the important part; the transformer connections. This only applies to the Plitron 8575-XO-02, and the colors are subject to change so it helps if you can read a schematic. First, connect the primary; #1) Black, goes to the 120V hot, #2) White, goes to the 120V Neutral. Then, the secondary; take wires #3,5,6 green, yellow and blue respectively and connect them together. Finally, the red #4 wire goes to the hot side of your receptacles (brass screw), and the #7 grey wire goes to the neutral side (chrome screw). Remember, you no longer have a grounded neutral! Just a word on the internal wiring. Most commercial Balanced Power Units bring the secondary leads directly to the first receptacle, and then "daisy chain" the remaining ones. I've even seen solid silver bus rods used for this purpose. This goes against common practice in the commercial and industrial electrical field. When dealing with more than one receptacle inside a single enclosure, you would have a common splice point and individual wires going out to each side of the receptacles (wire nut and "pig tails"). To my way of thinking, this has 2 distinct advantages: any downstream receptacle is not dependant on the one before it; if the circuit continues on from this point the load side will not suffer from a problem with the line side receptacles, and you are able to remove a receptacle for maintenance purposes without de-energizing other receptacles.

Be sure to bring the unit on line with nothing connected, and don't be alarmed if it trips the breaker the first (couple of) time(s).

 

[DiminishingReturnsOfficer]

Here is a link to a .pdf for a study conducted by Vanderveen at Plitron on mechanical buzzing of transformers. He tests 4 trannys; a standard Plitron toroid, a LoNo toroid from Plitron, a standard (non toroidal) transformer, and a toroid of unknown manufacture.

http://www.dalitech.com/Resources/Measuring%20Acoustic%20Noise%20Emitted%20by%20Power%20Transformers.pdf

A brief summary is as follows: there are 2 causes of buzzing in trnasformers; 1) loose coils, and 2) loose core laminatoins. This is not surprising as these are the 2 basic structures present in any transformer. Factors that contribute to mechanical noise are: overvoltage, and DC contamination. The latter was defined at 250mV and was seen to be the major cause of buzzing in all of the transformers tested. All the trannys tested were 500VA (although the study says 500kVA > typo(?)) Naturally, the Plitron LoNo design was the best performing transformer in group. There is a lengthy explaination of the methods and measures as well as graphs and charts corresponding to the results. It is an interesting article, but definitely not for the general public.

 

As mentioned earlier, I am experiencing a loud (by comparison) buzzing eminating from my Plitron 8575-VO-02. I contacted the company regarding this matter, and eventually got a response saying that I would not likely be able to return my toroid as the buzzing condition has deteriorated the quality of the transformer. Eric did not answer my primary question which was: is the buzzing I'm experiencing (assuming the 250mV DC level in the test) typical of a properly functioning transformer? He recommended a LoNo toriod of similar size 8770-X0-01 (2400VA) and noted that it would be larger and more expensive. I would consider stepping up to this toroid (even though it would mean reconstructing my entire unit) if Plitron were willing to give me credit for my 8575-XO-02. If not, I'm not sure I want to invest in another one of their products. Perhaps Equi=Tech would be better at standing behind their product.

I did respond to Eric's e-mail saying that he failed to answer my question. In all fairness, he has been very good about answering my questions in a timely manner. I hope to hear back from him about the buzzing and I'll be sure to let you know how it turns out. I'm also going to take the Balanced Power unit to a few different locations and see if the buzzing is present on an A/C line with less DC contamination. This is the only way for me to isolate the cause of the buzzing. If the Plitron buzzes loudly at 2 or 3 different locations, I can only assume that it is defective. If it doesn't, I'm completely willing to conclude that the problem is my mains power.

 

[DiminishingReturnsOfficer]

This is not a dirt cheap undertaking, and I want everyone to know right up front that the transformer will do 2 things that you may not like: #1) it will buzz loudly even with no load, and #2) it will trip your breaker, possibly the main, if you turn it on at the exact moment of a peak waveform with regard to Zero Crossing. The buzzing can be mitigated by placing it in another room; perhaps at the service panel. I have mine sitting on the same stand as my power amp. It sits on a 5/8" rubber mat, mounted with both rubber and plastic washers, additionally; I put big rubber feet on the enclosure, and "floated" the transformer. It still makes enough noise to be heard from the listening position with the volume all the way down. Ironically, it provides strikingly black backgrounds, but you can hear the buzzing during periods of silence. Of course, you can generally hear a refrigerator, HVAC system, or cars driving by too. You get used to it and, to me, it is still well worth keeping in the system.

As far as tripping the breaker, this happens because of the extremely large inrush current required to set up the magnetic field in a torriodal transformer. Instantaneous inrush current can exceed 500 amps (this comes from Eric at Plitron). Its all dependant on the A/C waveform's position with regard to Zero Crossing. If you plug it in or throw the switch (recommended) when the wave is at exactly zero on the vertical axis, theoretically, no current flows. If you send power to the 8575 at one of the peak values (positive or negative), an extremely large current will flow in the circuit. Depending on the type of overcurrent protection your home has, it can prove difficult to get the transformer online. Once its up and running, you won't have any problems keeping it going. Inrush is a one time deal. Its also important to note that this current does not flow in the secondary. I confirmed this with Chris specifically out of concern for my equipment. If the idea of 500 amps still scares you, as it did me, you may want to disconnect your gear until you start the Balanced Power System. In the event of a power failure, unless you use a 3 wire control relay, the Balanced Power Unit will start automatically. This has happened to me, as well as an instance where my wife turned the circuit breaker off and on again. Neither time was any damage sustained to my gear or Power Unit. Other possible concerns are that you probably don't want this unit on the same circuit as any critical loads. If this isn't enough to scare you off and you still would like to build a very good power conditioner, I'll start with a complete list of materials and approximate prices in the next entry...

 

[DiminishingReturnsOfficer]

Thanks to One and Half for his input. I didn't include any kind of testing equipment, and that may have been an oversight. You will certainly want to test your connections before energizing, but something that some people may not realize is that if you go to test your transformer connections for shorts (by setting a multi meter to read Ohms for continuity) you are going to read a dead short from line to neutral on your primary and line to line to ground or your secondary. I would advise using the multi meter to verify proper secondary volage once the tranny is energized. As far as what you're calling a "high voltage insullation tester", these instruments, known as Meggers, are fairly expensive and not something, even a Journeyman electircian such as myself, has on hand. The tranny should have been tesed for leakage at the factory. Also, I believe that wire nuts are as effective a means for connecting wires as you'll find. There are boxes in factories containing wire nut connections made back in the 50's that still provide useful service today. My reason for mentioning factories is that you'll not find a more vibration replete environment. The only choice that I would think preferable would be taking the secondary leads of the transformer directly to the receptacles, and terminating the primary leads on an IEC inlet. Naturally, no connection is always preferable to any splice, but I'd much rather have a wire nut on a connection than a terminal strip, crimp lug, or even solder. I've seen 50 pound high bay lighting fixtures hanging by their leads with only the wire nuts supporting them. Of course, I don't like to see a wire nut with more than 6 wires under it. Even though they make them big enough to accommodate this, they are rated for fewer large AWG wires; not many, many smaller wires. The only place where you'll need to connect more than 2 or 3 wires together with this project is the ground. For this, I recommend using a ground bar screwed directly into the chassis. Naturally, do what suits you, but I don't think you'd be sacrificing any performance by using wire nuts. Take the cover, off any receptace in your house, what do you see?

I agree completely with One and Half about overcurrent protection, in the case of a smaller unit. If you use the Plitron 8575, it is rated for 2000VA which translates to 16 amps. The NEC requires that branch circuits not be loaded in excess of 80% of the circuit (breaker) rating. This comes to 16 amps. Unless you have branch power circuits in your house connected to a 30A or larger breaker (a code violation in itself), you should not need additional overcurrent protection for the 8575. It is absolutely imperative with a smaller tranny (I think I mentioned that).

GFCI protection is a good idea, but it will only be effective for connected audio gear if installed on the secondary side of the transformer. If you're building a Balanced Power System as a separate unit, this would mean either installing a GFCI breaker inside the unit or using GFCI receptacles. I choose to use Audio Grade receptacles instead, and I'm not aware of any Audio Grade GFCI's. Let me, explain the reason for concern: With Balanced Power, you have 2 "hots" instead of a Phase Leg and Neutral. In the case of an incandescent light bulb, we would have an energized screw shell which could cause an electric shock if you were to touch the shell while changing the bulb. Also, if you have a very old 2 wire component (and I'm not aware of any specific components) that makes a connection between the neutral and the chassis, running balanced power would put 60VAC on the case (a potentially life threatening condition). If this is the "case" (excuse the pun), don't use balanced power. In fact, replace the component at once! There are other issues with single pole switches that have pilot lights lit when the unit is not powered up that might cause problems when connected to balanced power. I'm leaving it up to the end user to address those types of concerns. If you know your gear, and a little about electricity, you should be fine with this project.

 

[DiminishingReturnsOfficer]

I tried switching the Balanced Power unit to the position where the LED is not illuminated, and it did not have any impact on the buzzing. I must have realized this potential from the beginning because I installed a DPDT switch that allows me to bypass the LED should it be necessary. I've, now, gotten 2 responses saying that they don't have any problems with buzzing in their transformers. Do you hear a buzzing when you bring your power amp.s online? How do you have your transformers mounted? I know Plitron offers a vertical mount for the 8575. Mine buzzes even with me holding it in my hand. I've heard this complaint from many other sources, and assumed it to be a necessary part of the balanced power equation. I will be contacting Eric at Plitron to see if perhaps I have a defective torroid. In my experience, a transformer of any size buzzes; some more than others. When I walk into the main electrical room in a commercial or industrial facility, the 60Hz buzzing is so loud that you have to raise your voice to be heard. I've seen and installed many transformers, and can usually tell when one is buzzing too loudly. The fix for it is to "float" the transformer. They come from the factory with the coil assembly bolted to the chassis very tightly (for shipping). It is, generally, necessary to go through and loosen these bolts and torque them to manufacturer's specs. I have my torroid mounted with a 1/2" bolt that goes through the bottom of the chassis but not the top. I placed it on a 5/8" thick rubber mat, and used the factory supplied rubber washer beneath it as well. With the tranny energized, I tried tightening and loosening the bolt and listening for changes in buzzing. Originally, I had it just barely wrench tight (too tight to loosen with fingers), but I found that it quited down a bit when I tightened the bolt a bit. Since I don't have any sound level measuring deveice, I'm just going to have to say that it is audible from the listening position (8' away) you have to focus to hear it and its no more noise than a refrigerator makes in the next room. I'll let you know what Chris has to say about the noise.

I also wanted to address something I found on another forum about building Balanced Power Conditioners. Digital components, with their switching power supplies, can contaminate power to analog components. One possible way of dealing with this is to use a separate transformer for your digital components. They could both be housed in the same enclosure, and be fed from the same 120V circuit. You would need to first distinguish between digital and analog components. CD/DVD/Bluray players, DAC's, pre pros with digital signal processing, computers, and digital HD TV's are all digital. Turn tables, analog tape players, and power amplifiers not classified as digital are all analog. Normally, your amp(s) are going to be the biggest load, and would require the bigger transformer. Its important to note that your amp's Watt Rating is different from the Peak power it might be called upon to output. I would size the transformer according to the analog section's Peak demand, and size the digital transformer accoring to the manufacturer's rating on the rear panel on the unit; possibly allowing for expansion. Again, overcurrent protect any transformer capable of drawing more current that the rating of the circuit its on. I can't attest to how much of an improvement this will make, but it sounds good in theory. It should be noted that there are professional studios that use just one transformer for all their gear (digital and analog). You can also get transformers with more than one secondary circuit (these are used for multi channel amps).

 

[Brand B]

To clarify my exact situation further -

 

No noise from my speakers under any circumstances.

 

No mechanical noise audible through the metal enclosure I made under non-toasting conditions.

 

Very audible electric 50 Hz hum when toasting from the power unit itself, not from the spakers, clearly audible from 20 feet away.

 

My plitron is on its own dedicated 20 amp as well. Do not know if the toaster is on the same 120V leg of the 240V service to our house or not.

 

On enclosures, what I did for the outlets was cut rectangular holes in the metal back plate that the duplexes would fit through, then had an acrylic decorative plate with labeling and such laser cut to go over that. You might look for signage shops and such, or even look into shops that can waterjet or lasercut thin metal plate to order. You might be able to make a decent back plate for not too much ($40-50) if you can provide a CAD file for them to work from.

 

[DiminishingReturnsOfficer]

My next step is to take my unit to a few friends', finacee's, and brother's house this weekend and see if the buzzing persists. If it does, I find it hard to believe that 5 different places would have the same DC contamination. I've already turned every breaker in the apartment off and still get the buzz. As I explained earlier, this just means that the theorized interference isn't coming from my unit. It could be coming from anywhere in the building.

 

I had a thought that may work, if I have to resort to spending more money to fix the problem: I'll get a slightly larger 2400VA toroid, but this one will be one of Plitron's LoNo NBT (Low Noise Narrow Bandwidth) models. I'll either have to get a bigger enclosure for both of them or make another one for the other one. The idea of the LoNo toroids is that they perform better with A/C that has an unwanted DC component to it. Since I would never need the full 4400VA, I would not connect the trannys in parallel. I would have them in a series cascade arrangement; in other words the 2400VA LoNo toroid would feed the 2000VA unit which would feed my digital front end and TV. My thinking is that, the LoNo tranny should be able to deal with the the dirty mains power and operate much more quietly. It's secondary would be free of DC, and therefore, could feed the 2000VA unit and it should not buzz. This would provide the ultimate in isolation for the digital components, and I'd still have plenty of power to run my analog.

If my 8575 (2000VA) unit still exhibits buzzing at the different locations, its going back. Hopefully, they'll be willing to give me credit towards the new one, but I'm not sure. Its in perfect condition, and only 3 months old.

 

[steve2701]

If I may just add an air of caution - please try and hear what each addition of kit does to the distribution, especially those that have a smps in them. I have spoken to a great many folks who know a lot more about this than I do, and they highly recommend a balanced sy stem, but run individual balanced units for each item.

If for instance you have a very noisy switch mode supply in the tv - it is being fed very clean power, which it then pollutes (thank you very much) and you are then directly linking that polluted power to every other component that is on the same distribution point, which is what you were trying to avoid in the first place.

I guess if you want the best noise rejection you may want to look at a seperate balanced line for 'analogue' supplies and smps supplies.

So with what you are saying above this will be a much better way of doing it, so long as each transformer is fed and kept very seperate.

Just my 2p worthh.

 

[DiminishingReturnsOfficer]

The buzzing occurs with the transformer at idle. The problem has nothing to do with equipment connected to the secondary. In fact, the buzzing is loudest with no load on the transformer. In addition to the link I provided to the study conducted by Plitron, I found another interesting article online that offers a possible solution to the problem by way of building a "DC blocker" circuit using diodes and capacitors. Here is a link for it: http://sound.westhost.com/articles/xfmr-dc.htm

The circuit has to be tuned to a rather small amount of current, and the article proposes using electrolytic capacitors to keep costs within reason. Again, we are talking about mechanical buzzing and not anything heard through the speakers. I recognize and have commented on the influence of noise from non-linear loads. In the case of balanced power, the degree to which components interfere with one another is reduced to just differential mode noise, as all common mode noise is cancelled. For anyone interested in building one of these units; I have recommended using a separate transformer for your digital and analog components. In fact, further isolation, if necessary, can be obtained by cascading transformers; First tranny feeds all analog components and a second transformer for the digital components. Be sure to check the secondary voltage of this final transformer as its possible for the voltage to creep up to values greater than what's recommended for safe, efficient operation of your componenets. The Plitron 8575-XO-02 has a secondary that is 6 volts higher than what is fed to the primary, cascading 2 of these would result in an operating voltage 12V higher than line voltage for your digital components. Plitron offers units with mulit-tapped dual primaries to compensate for over / under voltage. The NEC recommends that you say within the range of + or - 5% of the nominal voltage (120 for US). 126V, therefore, is the highest permissable on a 120V line.

 

 

[DiminishingReturnsOfficer]

For anyone concered, the way things stand presently with the folks at Plitron is as follows: I have verified my connections and checked to see that the buzzing is present at different locations; presumably with different levels of DC offset present on the mains. In each instance the toroid exhibits the same (or very near the same) buzzing. I contacted a gentleman is sales at Plitron regarding a return authorization for the transformer. He informed me that he would need to consult with Eric in Engineering before he could make the determinatoin. I made the additional request that I be able to return the 8575-X0-02 for a credit to be applied towards one of their LoNo models. It will be quite expensive to ship the transformer back to Canada and ship the replacement to me. If I'm going to incur this cost, I want a transformer that will not buzz. I certainly don't want the same unit shipped back to me as this would represent a $100 loss for the sake of an inconclusive bench test and my transformer getting a tour of the Great North. I always mark my returns in some inconspicuous manner before sending them back to ensure that I get a true replacement. Its kinda like putting a bit of White Out on your oil filter before sending it in for an oil change: It keeps people honest. I'm not insinuating that Plitron would be anything less; "Trust but Verify". So far, they have been very good in communicating with me, but they do seem rather reluctant to accept a return, and I seroiusly doubt that they'll issue a credit towards a more expensive model. Even making such a request makes it look like I'm just wanting a no cost upgrade, but that is not the case. As I stated earlier, what I want is a dead quiet toriod; much like the CA members who have contributed to this post claim theirs to be.