POWER GAMES:Servo Stabilizer,CVT or Online UPS?

[ajay124]

From bryston website, the power consumption in bold for different modes, please crosscheck if i selected the correct mode
Bryston Limited - Music For A Generation

Power Consumption
& Heat Load At Idle 170 Watts
Max. Heat Dissipation 580 Btu/Hr.
2 channels @ 300W @ 8 ohms 1280 Watts Max. Heat Dissipation 8 ohms 2320 Btu/Hr.
2 channels @ 500W @ 4 ohms 2100W Max. Heat Dissipation 4 ohms 3750 Btu/Hr
Bridged @ 900W @ 8 ohms 2040W
Max. Heat Dissipation 8 ohms 3890 Btu/Hr

Does that mean I would need the 2000W Bhurji CVT mentioned in the brochure attached to my earlier post or something bigger or smaller?Considering that my speaker's need a max of 250 W at 4 ohms,what would the peak power consumption of the 4 B SST be in my set up?

 

[ajay124]

online ups is the best bet

servo stabilisers the correction is done mechanically( a moving contact connects to an exposed windings of a transformer) so it is too slow to prevent any fast risetime surge at the output

CVT has very good regulation and it wont let any spikes through but there is various issues like stability with dynamic load, noise, it has a overcurrent threshold with sharp knee,(see the pic), so i am not sure of its use for audio applications

In your first couple of posts you mentioned that a CVT has stability issues with dynamic loads.That if the load is not fixed and stable there would be occilation's.Can you elaborate on that?

 

[borg]

In your first couple of posts you mentioned that a CVT has stability issues with dynamic loads.That if the load is not fixed and stable there would be occilation's.Can you elaborate on that?

The heart of a CVT is ferroresonant transformer and a highvoltage capacitor, the capacitor and the leakage inductance of transformer(with a near square loop hysterisis core) forms an LC resonant network , any LC network is prone to oscillations(poles and zeroes etc)

While evaluating the performance of CVT's stability(other parameters are like line load regulation, and overcurrent threshold which i posted earlier) is an important criteria,I would test it by giving it a dynamic laod like say 0-150%-0, then i can see it breaking into oscillations, and once unstable typically it takes a certain % of load to get back to stable mode

unstable means there wont be any control on output voltage, i had some waveforms of this condition from a Design verification done 5 years ago, cant find it now

 

[arj]

If a cdp and pre consumes roughly 50W each,I would like to confirm what the 4B SST rated at 300W at 8 ohms and 500W at 4 ohms feasts on.The tech specs of my Vienna Acoustic speaker's recommend an amp with power rating of between 40W-250W presumably at 4 ohms.Given these specs would the 4B SST ever be operating anywhere near it's peak power consumption?
I am bad at all this stuff and would appreciate a little light on what may appear to be very naive questions

-dont go by output power, what you have to worry about is Input Power. Typical Class AB amplifier work in the 60-85% efficiency hence Input power will be higher.
-Another issue is Dynamic (instantaneous Power). this Peak power could be as much as Double the usual constant power rating (this is what gives the Dynamism and if your power source cannot supply it ,the sound will go dull (the usual problem with many power conditioners)

speaker power is deceptive since the power consumption is very dynamic across the frequency as well as the time axis . from what i tried to find out a couple of years ago, speaker efficiency (in the rage of .5-10%) as well as the SPL required at the point of seating can give an indication..but is too theoretical..

So i would guess double the max power rating of the amp should be enough for Max..min power would be much more dicey. usually most Class AB amps are biased upto 10% of output on class A 10% its rated output so 30W could be class A.. sine class A amps have an efficiency of around 20%.that means around 150W-200W inputs minimum ? (not sure of the Bryston...)

so you get around 200W min consumption

 

[borg]

In your first couple of posts you mentioned that a CVT has stability issues with dynamic loads.That if the load is not fixed and stable there would be occilation's.Can you elaborate on that?

I suggest you contact the manufacturer of the amp to find out the peak current demand and then confirm with the manufactuer of CVT to see that they match, if they dont have a readymade solution, guys like bhurji can easily give you customised solution, give them the precise info

powerconsumption alone is not the answer here because of the O.C characteristics of CVT

 

[ajay124]

@arj
"Another issue is dynamic (instantaneous power).This peak power could be as much as double the usual constant power rating (this is what gives the dynamism and if your power source cannot supply it,the sound will go dull, the usual problem with many power conditioners)"

Which is why Bryston does not recommend any kind of power conditioning.In places where the power supply is relatively stable a direct connection to the wall outlet is probably the way to go.Unfortunately in India one cannot take that stability for granted.The voltage in Chandigarh is relativel stable but does play tricks when it's raining hard and continuously.
From the 'dynamic' viewpoint,power conditioning may reduce rather than improve performance.Especially in a powerful solid state set up.
The positives from power conditioning seem to be
Surge protection
Regulated voltage with 1/2 % variation
Pure sine wave output
Isolation from noise present in the power supply
Back up in case of power outage
It has to be an individual decision whether we must have all these features or can be happy with a few of them.
Since I fixed the loose connection,the power supply has been satisfactory and the sound quality has improved dramatically.I feel that removing even the Belkin surge protector/extension board and plugging directly into three individual power sockets may improve the dynamics.Out of the positives listed above,if for a moment I sideline 'protection',then the only must have feature for me is isolation.A locally manufactured isolation transformer would probably cost between 2K-3K and a branded one 6K-9K.
I have no idea what 'isolation' will actually acheive,but I have this notion (may be right maybe wrong) that it may yeild some benefits in the area where I want them.Specifically,clean up a little muddiness in the mid range and add a little more 'air' and 'space' to the top end which seems a little subdued and closed in.
Like a clumsy ham fisted amateur,I am trying to break on through to the other side of 'hi-fidelity' .I am sure many of the paths I take out of ignorance,may be strewn with traps and pitfalls,but I am definitely getting closer to what I have been looking for.
Last night I played my eternal favorite,Schubert's Unfinshed Symphony No.8,and for the first time ever,after having heard it hundred's of times,I gave the reproduction of this music,on a system owned by me,a two thumbs up

 

[hifiashok]

Last night I played my eternal favorite,Schubert's Unfinshed Symphony No.8,and for the first time ever,after having heard it hundred's of times,I gave the reproduction of this music,on a system owned by me,a two thumbs up

in the end, as they say here..... it's all about the music!

 

[arj]

Great to hear Ajay.. But you dont know the impact of what you are missing by giving good power to your source. the Impact of good clean power to a CDP /transport/dac has to be heard to be believed ! (at least in bangalore !)

 

[Asit]

In Kolkata too!
Regards.

 

[shershah]

i have a 500 VA CVT (Bhurji) and the legend on the label says 2.2 amp output.
it has a fuse of 4 amps. This CVT has been in constant use for 12 years now, but not for audio.

i have unearthed some info about CVT -
Constant Voltage Transformer Operation

hope this helps

 

[hifiashok]

A locally manufactured isolation transformer would probably cost between 2K-3K and a branded one 6K-9K.

if i am not wrong, all the possible (perceived) issues with a CVT will continue with an isolation tfr, but not vice-versa for the advantages. and, there is a question of quality too, when the equipment being protected is expensive (and hard to come by). copper today is very expensive, from a manufacturer's perspective. corners could be cut w/o it being apparent to u and/or me.

 

[borg]

^^
An isolation tfr cannot provide any voltage regulation that a CVT or online ups can, and nor can match the high level of isolation and the surge/spike arresting capability of a CVT, but there wont be any inherent currentlimit or dynamic load instability like cvt
due its(iso tfr) leakage inductance however much smaller than a ferro resonant tfr can limit the maximum current drawable

I dont think an isolation tfr as the only protection makes much sense,it has very limited capabilities as protection device, just use it as a complimentary device to some active protection(example, at the output of a transformerless UPS)

 

[ajay124]

^^
An isolation tfr cannot provide any voltage regulation that a CVT or online ups can, and nor can match the high level of isolation and the surge/spike arresting capability of a CVT, but there wont be any inherent currentlimit or dynamic load instability like cvt
due its(iso tfr) leakage inductance however much smaller than a ferro resonant tfr can limit the maximum current drawable

I dont think an isolation tfr as the only protection makes much sense,it has very limited capabilities as protection device, just use it as a complimentary device to some active protection(example, at the output of a transformerless UPS)

Does that mean it's a stand off between a CVT and an Online UPS,the former providing isolation but impacting dynamics,the latter better with dynamics but not providing isolation?

 

[borg]

earlier you said Another issue is dynamic (instantaneous power).This peak power could be as much as double the usual constant power rating (this is what gives the dynamism and if your power source cannot supply it,the sound will go dull, the usual problem with many power conditioners)"

We dont have a precise data on how much this is going to be, engineering is all about precise measurements and knowing exactly what to do

this will be an issue here, for online ups this is determined by a parameter called crest factor. i see that the models you had suggested has a crest factor(Crest factor - Wikipedia, the free encyclopedia) of 3:1, You need to buy one with highest crest factor to minimise this issue. Other points like smps polluting the supply will be there , i dont have any data to tell you how much impact this makes in audio effect,
so you can buy the ups with least e.m noise. i am aware of apc SURT's are cispr22 class A,which is not the best ,but it is difficult to reduce the noise to minimum with switching powersupplies

In CVT if you see my earlier curve, that was 15A rated ferro output dropping to zero V at 25A, if you can find the actual peak requirement from amp mfgr , then get a cvt designed to meet this peak demand, also get a minimum load connected always to avoid the stability issue i had explained earlier, you can get this parameter as well from the cvt manufacturer

 

[ajay124]

I have spoken at length to several Indian suppliers.Mr.Manjunath from Vertex Mr.Prashant from Krykard and Mr.Yogesh from Elecon have all been very patient and helpful.The Chandigarh dealers have been helpful but no one is willing to give a home demo.
I've sent a mail to Bryston for their recommendation between a UPS and a CVT and the power required.Hopefully I will get a clear,non ambiguous reply from them and can proceed from there.

 

[arj]

earlier you said Another issue is dynamic (instantaneous power).This peak power could be as much as double the usual constant power rating (this is what gives the dynamism and if your power source cannot supply it,the sound will go dull, the usual problem with many power conditioners)"

We dont have a precise data on how much this is going to be, engineering is all about precise measurements and knowing exactly what to do

this will be an issue here, for online ups this is determined by a parameter called crest factor. i see that the models you had suggested has a crest factor(Crest factor - Wikipedia, the free encyclopedia) of 3:1, You need to buy one with highest crest factor to minimise this issue. Other points like smps polluting the supply will be there , i dont have any data to tell you how much impact this makes in audio effect,
so you can buy the ups with least e.m noise. i am aware of apc SURT's are cispr22 class A,which is not the best ,but it is difficult to reduce the noise to minimum with switching powersupplies

In CVT if you see my earlier curve, that was 15A rated ferro output dropping to zero V at 25A, if you can find the actual peak requirement from amp mfgr , then get a cvt designed to meet this peak demand, also get a minimum load connected always to avoid the stability issue i had explained earlier, you can get this parameter as well from the cvt manufacturer

Borg, this is really informative .
I have only seen 2 online UPSs till now and both have a crest factor of 3...not sure if there are any other.

Also why would an Online double converting UPS not be Isolating ? since there is a conversion to DC and then back to AC, isnt there an inherent isolation here ?

 

[ajay124]

A Mumbai based company manufacturing both Servo Stabilizer's and CVT's.

Continuous Variable Voltage Auto Transformer Variac | Servo Controlled Low Voltage Stabilizer | Toroidal Auto Double Wound Enclosed Transformer | Dimmer | Electro Plating And Anodizing Rectifier | Voltage Compensator | Current Auto And Constant Volta

 

[borg]

since there is a conversion to DC and then back to AC, isnt there an inherent isolation here ?

Did i say otherwise anywhere? I cant remember saying so,
come to think about it , may be in the bypass path there is none(in a tru online ups, the mains is automatically bypassed directly to the output incase of an inverter fault/overload with the help of a static bypass switch, in this scenario an isolation tfr is needed for true galvanic isolation)

Also there is a possibility of common mode noise passing through in case of a transformerless smps designs,transformer based designs provide a magnetic isolation from dc component of battery circuit and provide a stable neutral reference point,

However above scenario(non isolation during static bypass) is applicable to both tfr and tfr less designs, hence i prefer from a technology standpoint tfr less ups(high efficiency, less footprint) with a high efficiency, low impedance toroidal output isolation transformer

 

[ajay124]

@Borg,Arj,Hifiashok,others
More enlightenment please!
Specifically on
Isolation
Dynamic Load
Smps v/s Transformer
One reason why an online UPS seemed attractive to me was that I assumed,isolation would be inherent because of the double conversion involved.I lost interest when a couple of dealers recommended that I would separately need to buy an Isolation Transformer.What is the correct approach?
Other factors that I don't like in a UPS
At 70-80% efficiency a 2 kva UPS would have app.1.5 kva output.May not be enough for my system.A 3kva would not be vfm for me.
A transformerless power may impact on dynamics???
Fan noise would make it neccessary for the UPS to be placed near the distribution box,necessitating further internal wiring to the listening room.
Similarly in a CVT I am worried about
Loss of dynamics
Magnetic leakage
Sudden voltage 'dips' caused by what Borg termed as 'occilations'.
When a power ( ful) amp demands an instantaneous 'burst' of power the CVT may offer a 'compromise' solution.
A local manufacturer I contacted yesterday,who builds Servo's and CVT's categorically stated that a CVT should not be considered in an Audio set up as it is 'unstable' when the load requirements are dynamic.
I seem to be returning full circle back to a Servo+Isolation Transformer+ Surge Protector as the 'safe and simple ' option:sad:

 

[sidvee]

Ajay124 - I recall reading in other posts - where-in Audire - a mfr. of power conditioning equipment recommended against a CVT, specifically for high current audio applications. You may have to do a search.
Cheers
Sid