Expensive AVR vs Cheap AVR+Power amp

As far as i know there is a particular size transformer in an avr so when you use 2 channels say stereo all the power from the transformer goes to these two channels. When you use 7 channels then the power is divided so the load changes and power to each channel becomes less.

I'm sure venkat can explain better.

There is a lot confusion here and elsewhere on this. The confusion is in terms of understanding how power is used in an AVR. Let me try to explain this as easily as possible.

An AVR has a fixed supply rail that is fed by the transformer. In the older days, amplifiers used to have two power supply rails, based on a design that is roughly called Class G. One supplies lower power on a continuous basis for general listening such as dialogues, music etc. When the input supply demands a higher power supply, the amplifier would switch to a higher voltage supply rail, and then switch back to the lower rail. This worked well for regular amplifiers where the need for higher power would not be that often.

In an AVR, this design would create a lot of issues. The need for variable power is more and constant. Every time there a music blast, and every time there is a thud or an explosion, that channel would demand more power instantaneously. So modern AVRs loosely follow a design where the supply rails are variable, rather than two discrete steps. The input supply actually modulates the supply voltage. Why is this done? Simply because the amplitude needed for each channel is infinitely variable.

So, let us say we have 7 channels of 100 watts each. The transformer would be capable of supplying 700 watts of peak power on a continuous basis. In this scenario let us say we use only two channels? Does it mean that each channel will be capable of 350 watts of amplification?

No. The reasons are simple. The amplification is done independently for each channel. The specs are there, and not on the power supply. If a channel is capable of 100 watts, that is all it is capable of. In addition there is protection circuit built into each channel that will NOT allow more than the specified and required power to flow into that channel. If you feed 350 watts into an amplifier circuit that is capable of 100 watts, what you will see is an explosion and fire. So what happens to the 500 watts? Nothing. It just sits there to be used by other channels.

But my question is different. Music, and movies need extra power for speed, dynamism, headroom etc. So would I not be making it easy for the AVR if I use only two channels? Again the answer is no. When the amplifier is designed, the designer would consider speed, dynamism, headroom etc., and design the transformer. So instead of 700 watts, in general, the transformer would have a rating of say 875 watts, the additional 175 watts being capable of supplying the spike in power demand when needed. Each amplifier would be capable of taking in some 25% (or whatever it is designed for) power for handling transients. But this is it. Even if you have 1000 watts of additional power supply hanging around, it is not going to be used.

Cheers
 
Onkyo has highest power supply ratings in the same category.
eg.Onkyo 609 consumes 610w whereas Yamaha 667 only 400w.
 
Venkat has simplified the answer.
In short max amplification of each ch is fixed even though you put power supply more than required.(depending on the main transistor/Chip & circuitry)
In AVR power has to to be distributed in 7ch.Looking at the fact that may all ch not be working at same time ,powersupply used now a days will be under rated than actual required for all ch(continuous).Still for best sound Q & response,power supply is the key.
 
There is a lot confusion here and elsewhere on this. The confusion is in terms of understanding how power is used in an AVR. Let me try to explain this as easily as possible.

An AVR has a fixed supply rail that is fed by the transformer. In the older days, amplifiers used to have two power supply rails, based on a design that is roughly called Class G. One supplies lower power on a continuous basis for general listening such as dialogues, music etc. When the input supply demands a higher power supply, the amplifier would switch to a higher voltage supply rail, and then switch back to the lower rail. This worked well for regular amplifiers where the need for higher power would not be that often.

In an AVR, this design would create a lot of issues. The need for variable power is more and constant. Every time there a music blast, and every time there is a thud or an explosion, that channel would demand more power instantaneously. So modern AVRs loosely follow a design where the supply rails are variable, rather than two discrete steps. The input supply actually modulates the supply voltage. Why is this done? Simply because the amplitude needed for each channel is infinitely variable.

So, let us say we have 7 channels of 100 watts each. The transformer would be capable of supplying 700 watts of peak power on a continuous basis. In this scenario let us say we use only two channels? Does it mean that each channel will be capable of 350 watts of amplification?

No. The reasons are simple. The amplification is done independently for each channel. The specs are there, and not on the power supply. If a channel is capable of 100 watts, that is all it is capable of. In addition there is protection circuit built into each channel that will NOT allow more than the specified and required power to flow into that channel. If you feed 350 watts into an amplifier circuit that is capable of 100 watts, what you will see is an explosion and fire. So what happens to the 500 watts? Nothing. It just sits there to be used by other channels.

But my question is different. Music, and movies need extra power for speed, dynamism, headroom etc. So would I not be making it easy for the AVR if I use only two channels? Again the answer is no. When the amplifier is designed, the designer would consider speed, dynamism, headroom etc., and design the transformer. So instead of 700 watts, in general, the transformer would have a rating of say 875 watts, the additional 175 watts being capable of supplying the spike in power demand when needed. Each amplifier would be capable of taking in some 25% (or whatever it is designed for) power for handling transients. But this is it. Even if you have 1000 watts of additional power supply hanging around, it is not going to be used.

Cheers

theoretically, that correct, but practically its not. The confusion arises due to the way manufacturer's list their specs. A 7.1 channel AVR with 95w per channel will be having a trafo/ps of about 400-450w. How's that possible, have we invented a perpetual motion machine, thats where the AVRs defy the laws of physics. The logic is that the 95w rating is for 1 channel driven. If you drive more channels, the power handling per channel drops down. When you are driving 7 channels, its practically down to some 40-50 watts. With just two channels, you can get something like 60-70w. 95w advertised rating is almost always BS.
 
theoretically, that correct, but practically its not. The confusion arises due to the way manufacturer's list their specs. A 7.1 channel AVR with 95w per channel will be having a trafo/ps of about 400-450w. How's that possible, have we invented a perpetual motion machine, thats where the AVRs defy the laws of physics. The logic is that the 95w rating is for 1 channel driven. If you drive more channels, the power handling per channel drops down. When you are driving 7 channels, its practically down to some 40-50 watts. With just two channels, you can get something like 60-70w. 95w advertised rating is almost always BS.

If a manufacturer advertises 7x95, his transformer should have a minimal specs of 665 watts. The logic used by manufacturers is two fold. One, in any movie, 80% of the sound comes from LCR, and there is hardly any sound going to the other channels. At 95W per channel what you need all the time is 285 watts. Add another 115 to 165 for the other channels, and you are safe. At least that is the logic they use.

The other is of course, cost cutting. The market is so competitive that the margins are very small. Yamaha's xx5 series of AVRs became a dud as they screwed up on their power supply specifications. It is only with their xx7 series that they got back into the market. They lost 2 years.

Cheers
 
so - are you saying that replacing a transformer can convert a weak avr into a very powerful - taking it right to it's advertised specs?
 
Last edited:
So, let us say we have 7 channels of 100 watts each. The transformer would be capable of supplying 700 watts of peak power on a continuous basis. In this scenario let us say we use only two channels? Does it mean that each channel will be capable of 350 watts of amplification?

IMHO i think you have missed a very important point i.e. efficiency of the power transistors.

Mind you each discrete channel of most of the 5.1 or 7.1 recievers are powered by a pair of power transistors in Class-AB mode. Now this power transistors are roughly 50% effecient. That means if the powersupply feeds them with 200 watts it will produce 100 watts which will be delivered to the speaker output terminal and the rest will be dispersed as heat. That is the reason why amps/ AVRs get hot!!!

So a 7 channel amp will require a PS capable of 1400 watt to provide 700 watts of output on a continious basis.

Correct me if i am wrong:)
 
so - are you saying that replacing a transformer can convert a weak avr into a very powerful - taking it right to it's advertised specs?

Theoretically yes. But you need a lot of information and design capabilities to do that.

Cheers
 
So a 7 channel amp will require a PS capable of 1400 watt to provide 700 watts of output on a continious basis.

Correct me if i am wrong:)

You are right. But the assumption is that the designer will have taken that into account. For curiosity I opened my Onkyo 875, and it's PS occupies nearly half the cabinet! The front part is completely taken up by PS including it's transformer.

Cheers
 
I personally believe weight of an amp is agood indicator of its real power output capability.

Can you kindly elaborate how the receiver controls these 50 watts max per channel.

Rishi,

Agreed. weight is a decent indicator (Onkyo AVRs are often the heaviest and have the biggest transformers in their class but I am not enamoured with how they sound but that is another topic) of Power Supply size.

Say you have an Class AB AVR which is rated for 50W x 7 ch @ 8 ohms nominal 20-20khz and THD+IM less than 0.01% and if the same AVR puts out 100W @ 4 ohm nominal 20-20khz and THD+IM less than 0.01% and 200W @ 2 ohms nominal 20-20khz and THD+IM less than 0.01% now this is a AVR that has a beefy power supply. Most modern sub $1000 AVRs cannot do this. Why many sub $2000 AVRs fail at this too!

Why? Because W = V * I. At 50W@8 ohms x 7 channels the power supply is delivering 17.5A (total continous). At 100W @ 4 ohms x 7 channels the power supply has to deliver 35A (total continous) and at 200 @ 2 ohms the PS needs to be able to deliver 70A (total continous). Now this is where Power supplies fail.

Now let suppose a power supply is capable of delivering 25A (total continous). Lets say we are driving 7 speakers from it. 50W @ 8 ohms x 7 speakers no porblem (if the speakers are pure resistive loads). Now your speaker's impednace drops to 4ohms and hence the amp is required to put out more amps it's power supply will be stretched to cope with the demands of all 7 speakers. If however 2 of the main speakers were not enforcing their demands on the AVR's power supply the AVR's power supply could then use the 25A across 5 channels instead of 7. Since the front 2 speakers are those that put maximum demand on a AVR removing these 2 to a seperate power amp would free up the AVR's Power supply for center (also demanding) and rear (relatively less demanding) use.

Yes the Transistors that are part of the front 2 channels will be unused. But all 7 channels share a common transformer/recitifier and filter capacigtors so these will not be unused.

I hope this helps.

The solution to this is to buy a decent AVR, one with a hefty power supply and one with a good measured performance, to begin with...
...it is better to invest in a better overall AVR than an entry level AVR plus a power amp. .

very well put. That's what I wanted to say. Instead of adding up two low end equipment, get one very good no (lesser) compromise product where everything is top quality.

Ofcourse a $1000 AVR will outperfrom a $500 AVR even when it comes to power delivery (forget all the other features) but say you have $1000 to spend on an AVR. Woudl you spend it on a $500 AVR and $500 power amp or a $1000 AVR? A NAD T757 or Rotel 1550 will outshine a Denon 1919 or Manratz 5006 but lets compare the Yamaha A810 vs the A2010 or Denon 1912 vs the 3312 or Onkyo 609 vs 1009.

An AVR has a fixed supply rail that is fed by the transformer.

In an AVR, this design would create a lot of issues. The need for variable power is more and constant....So modern AVRs loosely follow a design where the supply rails are variable, rather than two discrete steps.

Wow I did not know AVR used Bob Carver's Class H design. Remember the M400 cube, then the M200, M1.5, M1.0t, etc....only he called it a commutator control or Adaptive Class G.

With a class H design, the supply rails are variable rather than two discrete steps. The signal actually modulates the supply voltage. - Wikipedia

Pioneer, Denon, Marantz etc.. have gotten clever lately. Nonetheless even at full power and the rails at amximum the Power supplies are still current limited na?
 
You are right. But the assumption is that the designer will have taken that into account. For curiosity I opened my Onkyo 875, and it's PS occupies nearly half the cabinet! The front part is completely taken up by PS including it's transformer.

Cheers

Onkyo TX-SR875 is a TOTL(Top Of The Line) 7.1 receiver weighing 23 kgs. It definitely posses a beefy power supply.

I have studied my own and quite old 7.1 channel Yamaha DSP-A2070 amplifier with my humble little knowledge I posses and have reached to a conclusion that power output claims of the amps in the 80's & 90's era where more truthful to their actual output capabilities. Let me share my findings with this internal shot of DSP-A2070:

p1010680marked.jpg


Back in 1993 Yamaha stated this amp was capable of 100 watts @ 6 ohms for the main left, main right & center channels. Only 26 watts was provided for the rest of the four channels. Mind you this total of max of 100X3 + 26X4 = 404 watts @ 6 ohms was available through out the 20Hz to 20KHz audio frequency range with a THD even less than 0.015%. When a company rates their amp at 1kHz, this usually means a full bandwidth(20Hz-20kHz) measurement will be about 15-20% lower. Also keep in mind A2070 weighs a hefty 21 kilos.

The Power Supply

Small One

At first I noticed the A2070 have two discrete power supplies. The smaller one is marked in white rectangle(pic above) in the extreme right bottom corner is dedicated to power the pre-amp module, DSP circuits, amp's protection circuits & all the LED's.

Big One

The second much bigger power supply consisting an EI core transformer in the violet rectangle(pic above) is completely dedicated to the power amp section alone and are assisted with dedicated power capacitors as shown in sky blue & blue rectangles(pic above).

This transformer being over 5 inches tall, 6 inches wide is rated at 800VA or 800 watts weighing 8.5 kilos alone. In contrast the whole Yamaha RX-V471, 5.1 channel receiver with 105W (6ohms, 0.9% THD) (1kHz, 1ch driven) rated output power claim, weighs just 8.2 kilos. :p

This power supply unit is assisted with four power capacitors. Those bigger pair of 4 inch tall lug type capacitors(clamped with the frame) shown at the sky blue rectangle(pic above) in the upper top corner are rated at 22,000uF each take care of the main left & right & center speakers only. So there are 44,000 joules of energy readily available to handle the dynamic peaks of music. Equates to 44000/3 =~ 15,000uF per channel rated at 100 watts.

To provide an analogy my Norge 1000 stereo amp rated at 125 watts per channel uses just 9,400uF capacitance per channel. So I found DSP-A2070 have almost two times the capacitance to handle the dynamic peaks per main left, right & center speakers than my Norge 1000 provides per channel.

The rectangle in blue(pic above) shows the other pair of capacitors rated at 8,200uF each, dedicated to the four effect speaker channels rated at only 26 watts each. So we have 16,400uF capacitance for these 26X4=104 watts.

In the end I found a respectable 800 watts transformer and 60,400uF of capacitance dedicated to power the power-amp section alone which is claimed by the manufacturer to produce just 404 watts in total from all its 7 channels.

The Power Transistors

For the Main left, Main right & Center channels

The power amplifier module of this DSP-A2070 is discrete/solid state as it uses a pair of Toshiba 2SA1302/2SC3281 power transistors in push pull mode for each of its right, left & center channels. These transistors are marked in light brown rectangles(pic above) as pairs.

Looking at their spec sheet I found each of these pairs are rated at 150 watts of peak dissipation and can easily pump out a genuine 100 watts of continuous power without crossing the SOA limit. Considering these transistors have roughly 50% efficiency, each of these pairs will require 200 watts of power from the power supply to produce 100 watts of power for each channel while the rest 100 watts being dissipated as heat.

So, these 3 pairs of transistors for the right, left & center channel will require 200X3=600 watts of power from the power supply to produce 300 watts of max power in total which equates to 300/3 = 100 watts per left, right & center channel all being driven simultaneously.

For the Four Effect Channels

Yamaha used a pair of Sanken 2SA1726/2SC4512 power transistors in push pull mode for each of the four effect speaker channels shown in the brown rectangle(pic above). Each of these pairs can easily pump out 26 watts of continiuous power without even comming close to the SOA limit. Again if we consider these transistors have roughly 50% efficiency, each pair will require 52 watts of power from the power supply to produce 26 watts of power to each of the channels, the rest 26 watts being dissipated as heat.

So these 4 pairs of transistors for thsee four channels will require 52X4=208 watts of power from the power supply in order to produce 104 watts of max power in total, which equates to 104/4 = 26 watts for each of these four channels all being driven simultaneously.

The Heat Sinks

Even before I considered that this amp is truly capable to deliver 404 watts, I had to consider how the rest 404 watts of heat generated from all these 14 power transistors can be dissipated. Looking at the dual row of heatsinks, each above 4 inch tall and nearly 12 inches long I came to the conclusion that this amp must be producing a lot heat, and since it produces a lot of heat a equal amount of lot of power is provided to the speakers.

Calculations

Total continuous power capability of the power supply = 800 watts (excluding the power capacitors.)

Total power requirement from the power supply to produce the manufacturer rated 404 watts of this amp = 600 + 208 = 808 watts.

So this amp meets the demand of its power amp section in full measure and are actually capable to deliver true 404 watts of continuous power, throughout the entire audio bandwidth (20Hz~20kHz) @ just 0.015% THD with seven channels being driven simultaneously.

Until now we have not even considered the 60,400 joules of capacitance waiting to lurk whenever & ever if more power is required. The outcome is with this amount of dedicated capacitance the A2070 comes with 2.0dB dynamic headroom @ 6 ohms. Very few TOTL AVR's can match this figure.

This simply means that while this amp will provide 404 watts of steady continuous power @ 6 ohms with all channels driven simultaneously, it can under extreme conditions jump up to say 600 watts for a moment (with out even considering the transformer's internal capacitance) to handle transient peaks of a complicated music composition when heard in full blast.

IMHO the points to consider about an amp's real power output capabilities are :

1) The weight of the amplifier/ AVR in question. Heavier means more power.
2) The size & weight of the transformer. Bigger & heavier means more power.
3) The amount of capacitance available from the power capacitors. More is better for better dynamic headroom.
4) The heat sink size. Bigger means more power.
5) The power transistors in use & their max power output capabilities.

---------------------------------------

Please feel free to add your own inputs.
 
Last edited:
You are right. But the assumption is that the designer will have taken that into account. For curiosity I opened my Onkyo 875, and it's PS occupies nearly half the cabinet! The front part is completely taken up by PS including it's transformer.

Cheers

Yes ,I have taken picture of it ;)-

 
so - are you saying that replacing a transformer can convert a weak avr into a very powerful - taking it right to it's advertised specs?

It can make AVR stable in terms of power,but not powerful.
ie. 90w powered AVR can not output 100w simply by changing power supply, but it can deliver true power as mentioned in AVR specifications.
 
RSX-1057 claims 75w continuous (Total 5ch) & power consumption around 450W.

Don't go by the power consumption figures, I have read in many forums they can be very misleading.

A specific amp/AVR can have 10 types of power consumption ratings specified for 10 different countries even though all are capable to produce same amount of power.
 
Don't go by the power consumption figures, I have read in many forums they can be very misleading.

A specific amp/AVR can have 10 types of power consumption ratings specified for 10 different countries even though all are capable to produce same amount of power.

I agree,it depends on reading method & frequency used.(1khz or all spectrom).
Still Rotel ,Marantz specifications can come close to real performance.
 
Still Rotel ,Marantz specifications can come close to real performance.

I cannot comment about Marantz, but Rotel specs comes close to their real capabilities.

About Marantz, all I can say when an audio enthusiast sold me his 15 year old 7.1, Yamaha DSP-A2070 he told the power output capabilities of his recent 7.1, Marantz SR6003 weighing just 13 kilos rated 120 watts @ 6 ohms for all channels are farcical in comparison with the DSP-A2070.

According to him, at low volumes both were good, but at high volume A2070 was always the clear winner having better sound-stage & SQ.
 
Last edited:
At 95W per channel what you need all the time is 285 watts. Add another 115 to 165 for the other channels, and you are safe. At least that is the logic they use.

My contention:
AB amps are about 50% efficient. Add to that the PS supplies of AB amps have heat losses too so for every watt we have at the speaker terminals we are probably consuming 3W at the power cord. Hence a AV amp that has a 900W power supply will in reality only deliver about 300W (be it 50W x 6 or 100x 2 + 30 x 3 or any combination there of) at the speaker terminals.

An AVR with a 900W PS will hence be able to keep those 300W for center and rear if a stereo power amp relieves it of front channel duties.



Now this power transistors are roughly 50% effecient. That means if the powersupply feeds them with 200 watts it will produce 100 watts which will be delivered to the speaker output terminal and the rest will be dispersed as heat. ...Correct me if i am wrong:)

Exactly only you forgot that Transformers also convert energy into heat so we have more losses there too.

According to him, at low volumes both were good, but at high volume A2070 was always the clear winner having better sound-stage & SQ.

A big transformer and big power supply is not guarantee of good sound. Powerful sound yes; but "good" sound No. THis thread is only discussing power not good but since you bought this up....

Every generation there are companies whose products outshine thier peers. Some 25 years ago Harman Kardon was very well recieved. One of their integrated the PM655 or 665 (I am old and I forget which) was highly rated by the late J Gordon Holt. Today the NAD T757 and Rotel 1550/1560 would be my best "audiophile" VFM bets. VFM is very subjective though.

Both the T757 and 1550 dont have Airplay, MultEQ XT, HDMI 1.4 etc....but they have excellent audio credentials.

My idea was to take an AVR that has Airplay, Mult EQ XT, HDMI 1.4, etc...and mate it with a stereo amp and get the best of both worlds on a budget that is less than that of a T757 or 1550.

Since I know very little about audio and absolutely nothing about video or home theater I figured the experts on this forum could help me. So back to my original question (See post 16 on this thread)...

So which AVR would qualify?
Pioneer VSX-2021? Marantz 5006? Yamaha Aventage 810? Onkyo 709? or something else?

What other specifications would one consider essential? I have listed 5 below.

  1. Front channel (2.2, 5.2, 7.2 etc...) pre out (thanks again Dharshan)
  2. Airplay (Does Airplay use Bluetooth?)
  3. Network capability (does this mean I can connect the AVR to the Internet and play music of some Cloud?)
  4. EzSet-EQ/YPAO/MCACC/MultEQ XT or similar DSP based (with manual override) TA/EQ system for 2 independant subwoofers.
  5. HDMI 1.4
 
Follow HiFiMART on Instagram for offers, deals and FREE giveaways!
Back
Top