[paba]

Hi all,

I'm tempted to implement a balance power setup and I'm wondering if 250 VA size units are enough in this case. The next size up is 500VA.
today all my equipment is on the same wall outlet which I guess sits on either a 15 or 20 amp circuit and breaker on my panel.

one 250 VA unit for the tuner (probably too much)
one 250 VA unit for the CD player (probably too much but i want to keep isolated from the rest right?)
one 250 Va unit for left mono 8W 300B SET (class A) tube amp (maybe a bit short)
one 250 Va unit for right mono 8W 300B SET (class A) tube amp (maybe a bit short)

One guy told me I can get by with one 250VA for my two sources and one 500 for both mono blocs and to stop obsessing.

Thanks
paba

[Occam]

Paba,

I might be a good idea to get a handle on exactly how much power your system is going to be consuming, so that you can size your transformer(s) correctly.
If you're in the States, you can get a kill-a-watt, a handy dandy all purpose power consumption meter -
http://www.google.com/search?sourceid=navclient&aq=t&ie=UTF-8&rls=GGLR,GGLR:2006-06,GGLR:en&q=kill%2da%2dwatt+meter

and in GB -
http://www.maplin.co.uk/module.aspx?ModuleNo=38343&doy=13m11
I'd assume there is something similar in Europe, etc....

Your SET 300b monoblocs are going to consume somewhere between 80-175watts each, depending on their implementation, and it might be benificial to pin it down. There also going to consume the same power whether playing at a wisper or full belt.

You might well be able to power the whole shebang with a single 500va balancing transformer, and provide isolation between components and additional power conditioning (if desired or needed) with simple diy CMC based filters.

What specific balancing transformers are you considering?

Regards,
Paul

[paba]

Thanks !

I like that tool, I will look into getting one here in Canada.(I have greedy laser printer I want to measure)
I admit I am rather surprised that a single 500VA can do the 4 components. But very happy because that lowers the budget completely even if I add some extras as you suggest. And the weight on the shelf will be easier too.

I was looking at the Signal Transformer SU-1/4, SU-1/2 and the Powervolt IsoZtran PSU-250, PSU-500.

cheers
paba

[bluesky]

Hey!  Watt a great little device!  Trouble is I need to know if it works upside down being a resident of Australia! 

Anyone know if such a thing is sold here? 

Also I noted our native electronics magazine made a statement that using "a tubed preamp" (any sort, didn't matter apparently) would degrade the performance of its latest kit amp......hmmmm, I've always found it best to listen first and then decide on the synergy of different components.   

Bluesky

[Occam]

Paba,

You should note that the transformers you're looking at need to be downrated from their respective 500 & 250va ratings. For balanced power, you wire both primaries and secondaries in series, nominally 240vac. But as Jon Risch notes-
http://www.geocities.com/jonrisch/catch2.htm
you feed and extract 120vac to the series connected primaries and get out (after grounding the secondary's center tap to get 60-0-60vac (120vac split phase, technical balanced power).
Long story short, you're constrained by the current ratings of the winding (not voltage ratings), and the 500va transformer must be downrated to 250va, and the 250va to 125va.

Frankly, the benefit of these industrial isolation transformers configured as balancing transformers lie in the fact that they're often available surplus via eBay or local electrical shops (if you can make it to NYC, you can have a SU-7 for $150, good for a downrated use up to 3.75kva) Also, while shielded between primary and secondary, their secondary windings are not wound bi-filar, so are not as effective as those wound specifically for balancing use like those from Equitech, BPT, Plitron, etc....

As it appears you're planning to diy, you might consider a purpose built 'second' from Equitech that rated for a full 10amps, a son of 'Q' transformer, housed in a utilitarian Nema 1 steel case for $399 -
http://www.equitech.com/products/specials.html

or used....

FWIW,
Paul

[kyrill]

HI Occam

if you do parallel filtering on the transformer to lower RFI and EMI what is the total combined capacitance as a maximum? Most caps i see are .2 uF parallel but suppose you use 2 uF in paralell how many caps can be used?
thx for yr many informative posts

[paba]

Thanks Occam,

another good suggestion. In past projects I have under-estimated the complexity of the box/chassis and has proven to be at least as hard as some circuits for me. So a solution that already has a container is welcome.

cheers
paba

[Occam]

kyrill,

Short answer yes. Long answer yes, its an incredibly complex problem.
Specifically, when a cap is on the output side, it forms, at minimum, a 2nd order low pass filter, which is exactly what you want a powerconditioner to do.  The transformer is source of the inductance. While the 'cutoff' frequency is determined solely by inductace and capacitance, the shape of the filter, its 'Q' is determined by those characteristics, and the load impedance characteristics as well. This is the same situation you deal with in passive crossovers in loudspeakers as well as their enclosure tuning. If you have a high 'Q' alignment of your enclosure or crossover, you have peaking. In a power conditioner, this means you can have amplification of noise rather then attenuation, if the 'Q' > 1.  For a given capacitance inductance product, which determines pole frequency, as the ratio tilts towards the capacitive, the Q increases. One should also note that the inductances, common mode and normal (leakage) can be very different is different transformer constructions, EI vs toroidal, for example.
The problem becomes even more complex, (as well as providing opportunities for optimizing conditioning for specific components/loads) when one adds additional series inductances and parallel capacitances for additional attenuation and isolation between components.
Long story short, without proper instrumentation and measurement techniques, you can obviously overdo it.

Caps 'across the line', at the input of a balancing transformer are a different situation. As the impedance of the source, your mains power, is typically very low, at first cut, one simply looks at it as a frequency dependant impedance in parallel with the load which is inevitably also frequency dependant. (But given the wild variation in what some folks are putting across the line, I'd measure that I've not formed an undamped resonant circuit with the balancing transformer)

Regarding the 'optimal' choice(s) for multiple parallel caps -
http://ultracad.com/mentor/esr%20and%20bypass%20caps.pdf

FWIW

[kyrill]

LOL Occam

well done 

so? how many capacitance in uF TOTAL for parallel caps ( hot to neutral or actually -V to +V in a balanced transformer) ) to be on the safe side is your best educated guess for best cleaning AC ( in a broadband matter) at the secondary site of an EI balanced transformer?

you spoke of <5uF but is that too less?

 aa

[Occam]

kyrill,

Sadly, its not something I can answer without doing actual measurements with a signal generator, various loads, and something to function as a spectrum analyzer. Whenever you have a 2nd order, or higher filter, which is what you have with a series inductance, the transformer, and parallel caps, you have a transfer curve with a 'Q' associated with it. Speaking in generalities, I've typically seen capacitances between 0.2-1.0uf across the phases on the output of toroidal transformer. On an industrial EI core transformer, I've seen them as high as 5uf.

Regards,
Paul

[paba]

Hi Occam,

got my Killawatt unit today from ebay. New one, but looks like it came from a weird batch because when plugged in, the display is upsidedown... lucky for me I can read Australian Works fine.

So far I can see my AC is at 117Volts and 59.9 Hz. When I get a chance I measure the audio gear and report back.

I wonder if it is fast enough to catch the current surges at power up most devices have. Otherwise I should just double the VA values of steady state to be safe right.

Thanks
paba

[Occam]

Paba,

Did you have to pay extra for that Aussie version?  One of the frustrating things about internet purchases is that its a royal pain to return something that is somewhat askew and one inevitably thinks 'it works' and I just deal with the (albeit) minor flaws.

I was looking at the Signal Transformer SU-1/4, SU-1/2 and the Powervolt IsoZtran PSU-250, PSU-500.

If you still intend to use these EI core transformers, if you can find them surplus at low cost, you need Signal DU or Powervolt PDU, not the SU or PSU versions respectively, which you'd wire per the Jon Risch instructions. The (P)SU versions are step up transformers. If you can't obtain these at very low cost, I'd recommend you get a purpose built transformer like the Equitech.

Also, unless your components actually have their chassis connected to ground, you're not really receiving the benefit of balancing, reactive leakage current cancellation, but rather, just those benefits of an isolation transformer. Many Chinese sourced (and other) components have 3 prong inlets that do not actually connect the mains ground to the chassis. While this is a rather large violation of most electrical codes, it does not mean that all such components actually comply. Its easy to verify that you actually have a grounded chassis with a multimeter checking continuity (resistance) between the ground on the IEC inlet and chassis. Some rather good bespoke components from Cat and Ayre also do this.

The downsizing of the Signal DU and Powervolt PDU transformers is not required because of turn on surges, which are transitory. I doubt that you'd be able to measure these events on the Kill-a-watt. The downsizing of these re-purposed transformers in necessary because you'd be wiring both the primary and secondaries in series for 240 vac but feeding and outputting 120vac, but the current maximum remains the same per winding. Half the voltage times the same current yields half the wattage/va rating.

Regards,
Paul

[kyrill]

hi Paul

i have 230 volts here in holland  so i bring primary and secondary in series and earth tho the middle tap of the secondaries?
i am thinking about those Signal DU ( but can use SU too?)
thx
kyrill

[Occam]

Hi Kyrill,

Yes, your description of the wiring is correct. If one were to use the SU type transformers, with dual 120v primaries and dual 240v secondaries, the primaries would be connected in series for 240vac and the secondaries in parallel, foregoing the balancing.

As I understand it, in the Netherlands, mains 240vac single ended power is provided via outlets/plugs without mains safety grounds, and protection is provided by ACD/GFI breakers where mandated by your applicable electrical codes. Given that, I don't see any reason for going to split phase balanced/technical power. If the mains connection does not require a separate safety ground connected to the chassis, the issues of reactive leakage onto the chassis is moot, and power conditioning can be provided by less expensive/bulky methods.

I've tried repeatedly to power my Aksa power amps via a DU-3, a 3kva transformer configured as a balancing transformer, and have never been satisfied with the results, inevitably finding some diminution of dynamics. Similarly, Mark Maloof has given up on using his DU-7.5, 7.5kva transformers configured as either a balancing or isolation transformer. Admittedly, in a 240vac country where the current draws are halved, this might not be so. Nor have I been able to try a purpose built for balancing transformer like those from Equitech or BPT. Scott Faller is powering his lower current (and less variable) draw tube based system off a BPT balancing conditioner and been extremely pleased with the results.

I do have very positive results in powering my LifeForce and Soraya off a little Felix conditioner built into the middle of a powercord -

using a JW Miller 8120 2.4mh 17Amp CMC choke. Not only has this provided a lowering of noise, better soundstaging and increased resolution, but it has extended and tightened the bass response. As to the later, it might not be noise filtering, but rather a result of a change in power factor in the amp's linear powersupply.
I'm also in the process of replacing my Felicia balancing conditioners for my source components with Felixs based upon the JW Miller 8109 30mh 2.3A CMCs. The Felicia balancing conditioner is still, IMO, a wonderful, worthwhile conditioner, but its bulky, relied upon no longer available surplus transformers, and were very constrained as to what it could power.
No matter what I try, I've not been able to improve on the built in conditioning incorporated into my CAT preamp. 

In Europe, the Wurth WE-CMB CMCs are nominally equivalent to the Miller 8100 series CMCs, but I've no experience with them. And your higher mains voltage, lower current draw might dictate different specification CMCs for optimal implementation. YMMV.

FWIW,
Paul

[kyrill]

thx for yr reply

ahh i did not know you have a Soraya
have you put yr lisstening impressions somewhere? is it tube like?

Kyrill

[Occam]

off topic -

Kyrill - My views on the Soraya are largely echoed by those who've already posted their impressions. As Hugh is a close, personal friend, I don't think it appropriate for me to post what would sound like a gushing 'fanboy' review, which it would. Briefly, its not tubelike, I use my CAT preamp for that, but rather, it lacks those artefacts of hardening, grit and congestion that I associate with most ampsat that price level and power, hollow or sandstate, when driven hard, while retaining superb resolution at all levels. (See, I'm starting to gush...)
I'll bring my Soraya to Phil's NYRave in mid December, and hopefully, more objective subjectivists will be able to opine and offer comparisons to other amps present.
-Paul

[KBK]

The loss in dynamic contrasting from undersized transformers is notable, as you have noted.

If attempting to go balanced power with whole audio systems, in my experience, transformers rated at about 10 times (10x) the nominal power draw is necessary.

[Johnny]

I do have very positive results in powering my LifeForce and Soraya off a little Felix conditioner built into the middle of a powercord -

using a JW Miller 8120 2.4mh 17Amp CMC choke. Not only has this provided a lowering of noise, better soundstaging and increased resolution, but it has extended and tightened the bass response. As to the later, it might not be noise filtering, but rather a result of a change in power factor in the amp's linear powersupply.

Hi Occam,
Are yous still happy with this Felix implementation on your Aspen Amplifiers- that is the JWM 8120, 0.47, 0.1, and 0.01uf caps, or have you found improvements since the original post?
Many thanks,
John

[Occam]

Johnny,

I still am using that same basic Felix topology on my Soraya, and am still pleased as punch. Subsequent changes to my commercial versions have to do with the choice of specific models of suppression cap, not their values, as well as non invasive surge supression. As I now worship at the feet of Mamon, I do keep certain things proprietary.

All I can suggest is make that minimal investment in the components you mention, slap her together, and let your ears be the judge.
If you're planning on implementing you Felix with a receptacle output, and you don't have a favorite bespoke outlet(and your wall outlets as well), I heartily recommend the HBL5262 specification grade duplex outlet, available for $12 from -
http://www.stayonline.com/detail.aspx?ID=7088&gclid=CNjQotypoJgCFQquGgodTUfunQ
I've no idea as to why receptacles have their own subjective characteristics, but they do.

FWIW,
Paul

[Johnny]

Paul,
I didn't realize you had a commercial stake in this 
Thanks for sharing as much as you have in the Felix project and the tip for wall outlets as well.
I just sent an order off to DigiKey for the Felix parts.
All best,
John