[Johnny]

I have a 5 Kva GE xfmr available to me which I have contemplated trying out in a balanced mains configuration. (I plan to diy some JR line filters as well.)

I have a question about the xfmr though. The name plate specs are 240/480v primary and 120/240v secondary. At first, I had thought to run a 20A 240v line to the unit and tap 120v off of the secondary. However, this requires parallel connection of the secondaries, whereas the 240v connection permits the grounded center tap required for a balance mains. As an experiment, I have connected 120v to the primaries, and with the "240v" center taped configuration on the secondaries I get the desired 120v balanced output. I then attached a "dummy" load consisting of a 7 amp hot plate and everything seems to be running fine.

My question is, is there anything wrong with this configuration, and how does it effect the specs? I assume the rated load might differ in this configuration. Any other concerns or potential problems?
Many thanks,
John

[Occam]

Sounds like a plan. From your description you have a 4 windings stepdown transformer, 8 connections in total, with 2 240v primary windings and 2 120v secondary windings. Is this what you have? If so, if you wire the 2 240v primaries in paralell, feed them 120 volts, and wire the secondaries in series ground the center, and you get balanced 120v out. Which is what I believe you described. The caveats are the usual that 120v AC can KILL you, which I assume you know, and that this configuration will require downrating that 5kVA rating. Have you checked you secondary's output to see that you get 60VAC from ground (assuming you've grounded that center tap on the secondary) to each end of the secondary?
This is covered specifically in Jon Risch's website. Look at Unit #2's configuration. I think this is what you're doing-
http://www.geocities.com/jonrisch/catch2.htm

PS - put the JR filter before the balancing transformer (per JR's reccomendation) and drop its inductor section.

[Johnny]

Hello Occam,
Yes, you've described the unit to a t.
Yes, I've been careful.
I grounded the center tap. I brought everything up on a variac in stages. I brought the unit up and checked no load voltage 60-0-60. Brought the 7 amp load up while watching load on amp probe.

I wasn't sure of the Kva rating change- up or down, or how to calculate?
Anyway, I assume with 5 Kva to start with I'll be ok. Main load is a 55w AKSA with 2x160 Va toroids- 2.6 amps? As Hugh always says in rush current is HIGH- 6 amp fuses have never blown. The rest is preamp, Creek analog tuner, cdp, dc turntable motor supply, phono preamp supply...

I missed JR recommendation for filter location on the supply side. I would have thought on the component side. Also elimination of inductors, leaving Caps only? Where is this? I've been pouring over the info on his website and missed this...

Oh, I was going to put safty devices (15a P&B breakers) on both 60v load side legs. Is that the way to go?

Thanks!
Yo da Man!
John

Sounds like a plan. From your description you have a 4 windings stepdown transformer, 8 connections in total, with 2 240v primary windings and 2 120v secondary windings. Is this what you have? If so, if you wire the 2 240v primaries in paralell, feed them 120 volts, and wire the secondaries in series ground the center, and you get balanced 120v out. Which is what I believe you described. The caveats are the usual that 120v AC can KILL you, which I assume you know, and that this configuration will require do ...

[JoshK]

John,

I believe what Occan means if you are using your step-down transformer in lieu of JR's inductor in the filter.

[Johnny]

Ohhh

John,

I believe what Occan means if you are using your step-down transformer in lieu of JR's inductor in the filter.

[Occam]

Johnny - The rating change would be downwards. For simplicity of calculations and a gross analysis, lets assume the rating if 4,800va rather than 5kva. This implies that each primary winding it rated for 10amps, and because of the stepdown function of the secondary, they're rated at 20 amps, each. At 10A/primary winding, we get that VA rating by multiplying the input voltage of 480v by 10amps for both primaries wired in series, or 20amps @ 240volts with both primaries wired in parallel. Either way that calculates the same 4,800VA. But you'll be feeding the amp with 120V subject to that same current constraint of 20amps for the primaries wired in parallel. 120v x 20a = 2400va, i.e., we cut the rating in half.

The rational for advising against the JR filter after the balancing transformer was an overstatement. Rather, I'd advise against any capacitors 'to ground', as in the JR filter. As its rather difficult to gat any 2 capacitors to have exactly the same charateristics, in that 'Y' configuration, it can lead to unbalanced leakage currents, which is what you've so mightily endeavored to banish, upstream. Add in a cap accross the 'line(s)' of the transformer secondary, an 'X' cap, would not cause this potential problem. You'll often find X caps at the ouput of balancing transformer, and both X and Y caps at their input., though the common mode behavior of the transformer should ideally make the Y caps superfluous. A JR filter, including the inductors, after the transformer, serving a specific low draw component, but without the Y caps  would serve to isolate components, fed by the transformer, from each other.
http://db.audioasylum.com/cgi/m.mpl?forum=tweaks&n=85922&highlight=balanced+Jon+Risch&r=&session=
http://db.audioasylum.com/cgi/m.mpl?forum=tweaks&n=104079&highlight=balanced+Jon+Risch&r=&session=

If you want to use a JR filter prior to the balacing transformer, including the inductors, let us know how it works. Folks have found that inductors in the path of a high current draw component can constrain dynamics..... but you can always remove the inductors.
YMMV

[Ulas]

Oh, I was going to put safty devices (15a P&B breakers) on both 60v load side legs. Is that the way to go?

Rather than breakers on the load side, most balanced power units use a GFCI. With balanced power, the leg that used to be neutral (0VAC) now has 60VAC on it. A fault on that leg could present a serious hazard without tripping a 15A breaker.

[Occam]

Sorry, I missed the new posts as I was composing the previous response. As Josh indicated, the balacing transformer itself is largely serving the power conditioning function, not just balancing. A well implemented balancing transformer should'nt require the benefits (and potential limitations) of  inductors on the input side. On the output side, the JR filter, sans 'Y' caps, could provide additional isolation between components.

ALLWAYS fuse/breaker a component, including this balancing transformer, where the wall power meets the component. In the event that windings of a transformer  shorts to the chassis, you want the fuse/breaker  to disconnect  it from the wall! I speak as one who has been stupid enough not to do so, and getting blown accross a room (assuming you survive) is something you only want to experience ONCE!

Fusing both 60V AC outputs is your call.... that voltage is substantially LESS leathal than 120V AC, which is one of the reasons that many construction site transformers are balanced.
[EDIT - The import of Ulas's comments on Balanced/GCFI issues was not understood by me when posted. The facilitatior of the LabCircle has given me a stern lecture.]

[Johnny]

Thanks Occam, Josh, Ulas et al. for your patient and thorough resposes. You are a great resource. I have a much better handle on the filter capacitors and inductors now. I thought breakers on BOTH legs of the BP would be a practical minimum safty device, but certainly GFCIs would be more effective. Any recommendations for specific devices?

One of my concerns about going to a balanced mains was the lack of "stiffness" in the face of transient power demands of the power amp (As expressed in "AC and You - a New Era, POINT 1. Transformer based isolation devices are basically useless as AC line conditioners. ", by MGD http://www.boundforsound.com/tweak.htm
(Not that I have great confidence in his credentials. He states that "transformers create DC in our components"- well not exactly!) Still, I have direct  experience with voltage drop in isolation transformers trying to start vacuum pump motors...

But with a free 5Kva unit at hand I thought I might give it a try. Should I be concerned that it must be derated to ~2.5Kva in this application? JR recommends a minimum of 2-3 Kva.

[Audiopro]

I've been following this thread. I have a similar situation with a transformer but haven't yet tried it. Can't wait to hear how yours turns out.
One note on GFCI's. In some parts of the U.S. GFCI's are required by law on balanced power transformers. So it might be advisable to install at least one. GFCI's will protect any downstream receptacles, so only one is required.
 I've also noticed an interesting phenomem with GFCI's as well. Some equipment these days are made with built in line filters on the IEC's. The leakage rating on these inlet filters varies but can be as much as 0.5ma line to ground. This is enough to trip a GFCI. This is not a problem, just something to be aware of.
 I had this happen to a customer of mine, and we ended up having to replace the filter in his preamp. But not before lugging his 60 lb toroid into the shop to check it out.

[Occam]

Ulas - Many thanks for pointing out these safety issues for balanced power transformers. They are very real. I did a google search to expand upon your comments -
http://www.google.com/search?q=gfci+balanced&hl=en&lr=&start=0&sa=N
and found a very good thread -
http://www.htguide.com/forum/archive/index.php4/t-8452
I've come to think of DIY audio as Nature's humorous experiment in Darwinism. But this obviously only works on a macro level, as I, in spite of myself, have managed to reproduce....

Johnny - My experience with isolation/balancing transformers is limited to powering source components, and I've not used big honking iron.
But you do presently have the resouces to characterize your transformer under varying loads, the hotplate and voltmeter. Measure the input, wall voltage. Measure the output voltage(accross the 2 phases), once with no load attached, and again with your 800w hotplate attached(after the hotplate is hot and stabilized). From the latter 2 measurements you can calculate the 'regulation factor' of the transformer, which is = 100x(noLoadVoltage-RatedLoadVoltage)/(RatedLoadVoltage)
RatedLoad is a somewhat arbitray number. In the case of your distribution transformer, its probably the RMS load the would cause a 30-50oC. rise over ambient temperature. For audio use, its probably more like the load that which is below that the subjective perfomance begins to deteriorate.
You can, within reason, linearly interpolate/extrapolate your own arbitrary RatedLoad regulation.
Hopefully, you'll  find that the regulation for your honking big iron is less than 5%,as would be expected given its size.  You should also find that the winding of the transformers are compensated for its own internal losses, by comparing input and output voltages.

But all the calculations in the world aren't going tell you how its going to subjectively sound, though the calculations can sometimes explain why it subjectively sucks.

Try it. Your total RMS load is south of 100watts. Your Aksa 55, which is the  'peak' draw, isn't a monster. Its certainly not going to cause your transformer to overheat. Try its with different combinations of your components plugged into it.

[Johnny]

Right! I owe it to myself and to all of you to give this hulk a try.

Right now, I'm just trying to sort out the safety devices.
Will a standard gfci wall outlet do the job on balanced AC- wired up to ground, -60v on "neutral" and +60v on "hot"? (It looks like the Furman unit desrcibed in the links you sent uses a standard Home Despot type device.)

'course this opens up another can of worms. Can you get cryo'd, plated w your favorite precious metal, and otherwise tricked out gfci outlets?  

[Ulas]

Another thing to be aware of is ground loop-like problems with some equipment that comes with polarized two-prong AC plugs. The underlying assumption in the design of such equipment is that the neutral leg of the AC connection is the same ground.

I had a hum problem with my NAD tuner that appeared after upgrading my system. Part of the solution was discovering and removing a 1 MegOhm resistor inside the tuner that was connected between the tuner’s ground reference and the neutral leg of the AC. That fix reduced the level of the hum. Finally, I installed an Iso-Max on the antenna lead-in to isolate the tuner from the RF-amplifier mounted on the antenna. It may be the hum was always there but I didn’t notice it until it was acerbated by the balanced power and revealed by the new 110dB speakers!

Now, before installing any new gear, I measure the resistance between chassis/audio ground and each blade of the AC plug. If there is any continuity I don’t plug it into a balanced outlet.

[Johnny]

Ulas, thanks for the tip.

Occam, so I did the hot plate test- 120v noload, 116v with the 7 amp load, a 3.3% drop. The load is one third of the 2.5 Kva maximum load. Linear extrapolation yields about 10% regulation. A little disappointing, but linear assumption is probably not valid.

I suppose one could watch the load side AC wave form on a scope while playing loud bass heavy music through the system, and see if there are any noticeable fluctuations....

[Syrah]

Hi Occam,

I figured I'd resurrect this thread rather than start a new one on almost exactly the same topic.

I have a Signal Transformer SU-3 - see here - http://www.signaltransformer.com/content/power-supply-step-updown-power-isolation-transformer

My plan is to wire it up to my 120-0-120 balanced line and step it down to 60-0-60.  I see in the above posts, Johnny was going 120v in and 60-0-60 out.

I take it I should wire mine as follows:

Primary - one 120 on the 240 primary, the other 120 on the other 240 primary, neutral on the two 0s.
Secondaries - 60v out of 120v secondary, 60v out on the other 120v secondary, ground the two 0s.

Would this give me a 14 amp max or a 7 amp max?  The spec sheet says it is 7 amps when wired in series, but that assumes going from 120 to 240, not the other way around (if this makes any difference).

Also, how should I rig up the circuit breakers - on the primary side (120-0-120) or on the secondary side (60-0-60).  And should I put one breaker on each hot, and what value?

Sound wise I suspect the breakers are best on the 240 primary side.

Thanks!

[Occam]

Syrah,

You can't get there from here.... Those SU transformers are step-UP transformers-


You'd have to use the secondaries, wired in series, as the primary, and use the primary wired in series, with the center tap connected to safety ground as the secondary, to go from North American 240vac (120-0-120) to balanced (technical) 60-0-60vac. I've never done this with such a large transformer, and therefore can't advise you to do this. Nor do I know the resultant transformer regulation resulting from an arse backwards implementation.

I've never been particularly impressed with the results of using the Signal isolation products for power filtering, whether it be for balanced power or isolation. This is not to say that vendors like BPT, Furman, etc don't make excellent balancing transformer units, as they use purpose built transformers.

My advice would be to simply build a Felix(s) if you're amenable to diy.  Do a search for 'Felix' on Audiocircle and you'll find much discussion.

Regards,
Paul

[Syrah]

Thanks Paul!

The problem is that I have it now and it's too heavy to use as a paperweight.  In for a penny, in for a pound...

Signal describes it as a step up/ step down trannie, so I'm not too worried about using it in reverse.

If I do this, how should I configure the breakers?  One on each hot leg?  240v end or 120v end?

Cheers.

[Speedskater]

I couldn't find the technical info at the Signal Transformer page other than it's a big transformer. Is there another info page?
Like Occam wrote some transformers are wound to account for internal losses, so in reverse the secondary voltage may be off.
Each individual winding has a current limit regardless of voltage, so it will be a 1500 VA (not 3000 VA) transformer.
Need the tech page to answer the remaining questions.
Remember that NEC rules are not friendly to residential use of balanced circuits.

[Speedskater]

So I went back to the Signal Transformer site and found the tech info.
(didn't we cover this same subject a few months ago someplace?)
A four winding are identical so frontwards and backwards are the same.
The current limit on the 240V windings are 14A (so with both in series it's still 14A).

[Speedskater]

Occam pointed out to me, that I mis-read the tech sheet.

The correct current rating should be 7 Amps.