• Welcome to the all-new HUG. All content has been converted from the old system, and over the next days we will re-style HUG in a more familiar way.

Got my new baby P3ESRs - my experiences at Doug Brady (all good)

Kumar Kane

New member
The Naim sound - a logical explanation?

The Naim sound - a logical explanation?

I recommended the Naim XS as it does work very well with Harbeth speakers and just wanted to clear up any uncertainty.
No personal experience of Naim, but they are known for big and powerful sound. Could that be on account of their high sensitivity inputs compared to many other amps of the same power output ratings? Why else would a 50wpc Naim sound so, compared to another equally honestly stated 50wpc amp? Maybe the sound level difference also contributes to their reputation for pace and timing, whatever that means?
 

A.S.

Administrator
Staff member
Amplifier "gearboxes" and power, and changing preferences

Amplifier "gearboxes" and power, and changing preferences

No personal experience of Naim, but they are known for big and powerful sound. Could that be on account of their high sensitivity inputs compared to many other amps of the same power output ratings? Why else would a 50wpc Naim sound so, compared to another equally honestly stated 50wpc amp? Maybe the sound level difference also contributes to their reputation for pace and timing, whatever that means?
This is a very serious observation and deserves to be investigated.

When we read comments like "amp QUV although small, really punches far above it's weight sonically" we despair. Although I have tried my very best - and god knows how many hours and illustrative examples I have created - to convey in junior high school language how amplifiers actually work, it is just not penetrating to the logic centres of the brain. I just do not know how to simplify the already simple explanations even further. The (uncomfortable perhaps?) fact are that every amplifier ever built and ever to be built until the end of time is merely a gate which permits the energy stored in its power supply to dribble out to the speakers. At the most basic level that's all you need to know.

The tiny little music input signal appearing at the input sockets drives a circuit gate which fully opens, partially opends or fully shuts in time with the music loudness which permits some (or all) of the PSU to feed directly to the speakers according to how wide the gate is open. The PSU and the gate do not know that the signal represents music. Only humans know that the wiggly wave generates a sound we call music. The signal that appears at the speakers released from the PSU reservoir is not 'music': it is a modulated DC voltage from the charged-up capacitors in the power supply. Can we get the idea firmly in the head please that amplifiers cannot 'punch above their weight': when the PSU is fully drained (which it can and will be with loud music and a small amp) then that's it - game over - no more power available until and if the capacitors recharge over seconds. 10W is 10W. It is not 11W or 15W. 100W is not 105W or 120W. A watt is a watt is a watt the universe over. So if your amp is truthfully specified at 12W into an 8 ohm load both channels driven then it has 3W more available power than another amplifier that is truthfully rated at 9W into an 8 ohm load both channels driven.

So how is it that perfectly normal, otherwise rational people can make such convincing statements about amplifiers, and in this example, about how certain models seem to have a loudness advantage over other alternative amps? First; the opinions are truly held. They are not fantasy; the listener (usually) does hear something, some difference, but rarely does he hear what the thinks he hears. His paucity of technical appreciation of what goes on under the lid of technical equipment leads him to make incorrect assumptions which lead to erroneous conclusions which once worked-up into a pet personal theory in the brain can last for life.

I am told that some scallywags on the internet incorrectly summarise my position as "Alan Shaw says all amplifiers sound the same". BS. If only that was my position! The exact opposite is true. What I am actually saying is that:

  1. ... under uncontrolled casual listening comparisons listeners will truly and inevitably hear differences between amplifiers (see how that is the opposite of what I am quoted as saying)
  2. ... under controlled listening conditions where loudness is measured and controlled at the ear, it may or may not be possible to have a preference for amp A or amp B
  3. ... any preference, no matter how passionately and fervently held for amplifier A under uncontrolled casual listening can be reversed under manipulated comparisons such that B is deemed far superior by the same listener, on the same speakers, in the same room on the same day with the same music.
The point that Kumar Kane is making is that the volume control can be thought of as the gearbox of the car. If the car designer has manipulated the gear ratios such that in first gear the tiny engine seems to accelerate like a rocket, the illusion will be created of tremendous power at the smallest rotation of the volume control from the off position. That would be cunning marketing/engineering. No matter how cunning it could not disguise the fact that once the car was in second or third gear the reality of the limited engine wattage (car engines just like amplifiers are specified in watts output) becomes apparent - acceleration plateaus. If the little car is sold to the about-town young driver the illusion of zippiness may fool 95% of users 95% of the time. A more mature driver used to a much bigger engine with even power throughout the rev range would laugh-off the little highly-geared car for what it is: an illusion of power, not true power.

Bring me two amplifiers. First we technically measure them to establish that the have comparable technical performance (distortion, frequency response, overload threshold etc.). Then you set up amp A and B and in a casual, unhurried way alternate listening between A and B, powering down and swapping over the input/output leads for as long as you want. I leave you alone to do that, nip out for a pie and pint, a country stroll, an afternoon in an English garden. Eventually, upon my return, you tell me that you have a preference for A or B. I ask you to give me 30 minutes to set-up my test equipment and I very carefully adjust the volume controls of the amps, measuring the signal level at the speakers. I apply masking tape to the volume controls so that neither of us can nudge the volume. You listen, me present or not, taking as long as you want but you are not able to see the amps; a third person does the cable shuffling. My manipulation of loudness will have reversed your preference. I will have made the discounted A or B your now certain, unarguable preference. How did I do that? It's all about understanding how the ear attributes sonic quality by determining loudness.

If you take away nor understand nothing else in this post other than this one thought, I've done my job:

There can never be an absolutely superior amplifier because the test conditions, particularly loudness at the ear, is the dominant factor in defining "amplifier sonics" nothing to so with the shape, size, power, technology, price or any other factor in or about the amplifier itself. That is a demonstrable fact.
 

Kumar Kane

New member
This is a very serious observation and deserves to be investigated.
Extending this further, seeing how much input sensitivities of currently available amplifiers vary - 90 mv to 500mv is my research - could this be the reason for the different sonic signatures seen across brands for their amps of the same output power specs? After all, most people will think that balance levelling means leaving each volume control at say, 11 o'clock, but the resultant significant level differences arising from input sensitivity differences will make comparisons meaningless.
 

Kumar Kane

New member
The (real) illusion of amplifier power!

The (real) illusion of amplifier power!

Extending this further, seeing how much input sensitivities of currently available amplifiers vary - 90 mv to 500mv is my research - could this be the reason for the different sonic signatures seen across brands for their amps of the same output power specs? After all, most people will think that balance levelling means leaving each volume control at say, 11 o'clock, but the resultant significant level differences arising from input sensitivity differences will make comparisons meaningless.
Sorry to quote myself, but extending the above even further - given that most modern sources, i.e CDPs and DACs, have outputs in the region of 2 volts, does amp input sensitivity become a factor in amp selection in addition to the recommended things like features, build quality, power output and after sales support? All things being equal, for a 2 volt source, which amp would be better - one with 90 mv input sensitivity or one with 500? The latter number because I know of none that has 2 volts sensitivity at the input.

And to Alan's post above, I am sure that at least some makers would be engineering the high sensitivity to create the more power illusion. I don't know any modern source that needs it - phono needs it to be a lot higher than 90 mv anyway, so it isn't to accommodate the vinyl set.
 

mhennessy

Member
Amplifier technical specifications: engineering v. marketing

Amplifier technical specifications: engineering v. marketing

There are ways to explain amplifiers which apparently "punch above their weight":

1. Their rated power is very conservative. An amplifier which is rated at 50 watts by the manufacturer might actually be 70 or more when measured independently. According to the objective reviews in Hi-Fi News, this is actually very common indeed.

This is surprising; after all, why didn't the manufacturer simply sell it as a 70 watt amplifier? Well, perhaps he wanted to be absolutely sure no-one could accuse him of being over-optimistic (knowing the power of a negative review)? Or perhaps the manufacturer didn't want to provide the heat sinking and power supply for a continuous 70 watts? After all, these items make up a large percentage of the overall material costs. Alternatively, the marketing department might have had very specific ideas about the size of the box, which might restrict the size of the power supply and heat sinking. Or, quite simply, the manufacturer might reason that his 50 watter will be directly compared to other 50 watters; many people buy on basic specifications alone, and many hi-fi magazines do group tests of similarly specified powered amplifiers. In these situations, he's given himself a clear advantage, yet one which no-one could reasonably complain about it. Smart :)

2. There is a difference between continuous power and dynamic power. An amplifier can supply a bit more power for brief periods because power supplies are nearly always unregulated. A supply will fall when supply a continuous sine wave is produced, but for typical music, it will be able to give a little more for peaks. In-between the peaks, the power supply is quickly "topped up" from the mains transformer (100 times a second). There is a balancing act between the size of the mains transformer and the size of the reservoir capacitors - but in general, the smaller the PSU, the greater the difference between continuous and dynamic power. There are a few standards for measuring dynamic power:

http://books.google.co.uk/books?id=-5UPyE6dcWgC&pg=PA306&lpg=PA306&dq=ihf+dynamic+power+audio&source=bl&ots=wvImsoCIkV&sig=8CxJ_ZeS2Q4sttItdDf4PBp5PGM&hl=en&sa=X&ei=BXhVUfi2H8u8Pd7lgdAE&ved=0CGkQ6AEwCA#v=onepage&q=ihf dynamic power audio&f=false

3. The clipping performance - the exact nature of harmonic distortion that result - varies. Amplifier A producing 10% clipping distortion might sound considerably "nicer" than amplifier B producing the same amount. Of course, this is why people think there such a thing as "valve watts".

These are in order of significance in my experience, but others may have other experiences...

Cheers,

Mark
 

A.S.

Administrator
Staff member
...And to Alan's post above, I am sure that at least some makers would be engineering the high sensitivity to create the more power illusion. I don't know any modern source that needs it ...
It's the oldest trick in the amplifier designer's book: fool the inexperienced audio buyer by gearing the volume control so that he has the illusion that his 500cc amp has the power of a 3 litre. Is my analogy of gearing completely understood because that is exactly how the illusion works?

People: if you are 17 year old first time driver I'd expect the power/volume rotation (gearbox) illusion to hoodwink you but if you are a mature driver, I'd really expect by then you'd see that marketing game for what it is.
 

Kumar Kane

New member
The lasting illusion of amp power

The lasting illusion of amp power

It's the oldest trick in the amplifier designer's book: fool the inexperienced audio buyer by gearing the volume control so that he has the illusion that his 500cc amp has the power of a 3 litre. Is my analogy of gearing completely understood because that is exactly how the illusion works?

People: if you are 17 year old first time driver I'd expect the power/volume rotation (gearbox) illusion to hoodwink you but if you are a mature driver, I'd really expect by then you'd see that marketing game for what it is.
The auto analogy is excellent, but for one thing. Most autos do end up being driven in the gears and speeds where the running out of power would become apparent soon enough. But in home audio, there will probably be no chance or desire to run a modern solid state amp at full - it would be just too loud. So the illusion may last longer, maybe forever.
 
S

skutters

Guest
Simple request

Simple request

All i wanted was an Amplifier that to my ears made the music sound better/nicer, and after listening to that amp at home side by side with my current amp to me it sounded better. At the end of the day thats all i am bothered about.

Regards, Kev.

{Moderator's comment: Noted Kev but others would reach different conclusions which could, on a 50/50 flip of the coin be the exact opposite of yours!}
 

s.a.b.

New member
Frequency response differences of amps?

Frequency response differences of amps?

Is it the case that two different amps may cause a slight difference in frequency response and, putting the issue of loudness aside, that this slight difference in frequency response may cause one to have a preference between the amps?
 
S

skutters

Guest
My decison

My decison

All i wanted was an Amplifier that to my ears made the music sound better/nicer, and after listening to that amp at home side by side with my current amp to me it sounded better. At the end of the day thats all i am bothered about.

Regards, Kev.

{Moderator's comment: Noted Kev but others would reach different conclusions which could, on a 50/50 flip of the coin be the exact opposite of yours!}
Yes i agree someone else could well hear the amps totally different from me, i have not been asking anyone to agree with my decision. I have just been reporting on here on what i found suited me.
 

mhennessy

Member
Realities of solid state amps frequency response

Realities of solid state amps frequency response

Is it the case that two different amps may cause a slight difference in frequency response and, putting the issue of loudness aside, that this slight difference in frequency response may cause one to have a preference between the amps?
That would be relatively unlikely - I've never seen it done. It would be spotted in a quick test or a glance at the schematic, and roundly condemned. The only problems you might run into would be a slightly early roll-off at the frequency extremes, but while measurable, these would rarely be audible.

I'm talking about solid-state amps here. With valves, the game is slightly different.
 

A.S.

Administrator
Staff member
That would be relatively unlikely - I've never seen it done. It would be spotted in a quick test or a glance at the schematic, and roundly condemned. The only problems you might run into would be a slightly early roll-off at the frequency extremes, but while measurable, these would rarely be audible.

I'm talking about solid-state amps here. With valves, the game is slightly different.
I don't completely agree Mark. If you connect a real world speaker across a selection of solid state amps (incl. for example a QUAD 405) you will see that there are minor changes in loudness across the audio band, related (somehow) to the electrical load of the speaker interacting with the amplifier. Those variations could be, say, ± 0.3dB or so even on well executed amp somewhere in the middle frequencies (say, 300Hz - 6kHz) which makes it a little tricky to pick a reference frequency for 'average' gain. Whether this frequency dependent gain change is audible is another matter: it's close to what I have found is the audibility threshold under super-controlled conditions.

I have to admit that the least load-interdependent amp I have ever measured is a North American made unit of good repute. It was really in a class of its own for flatness when connected to a real speaker.
 

mhennessy

Member
Q405 performance

Q405 performance

I don't completely agree Mark. If you connect a real world speaker across a selection of solid state amps (incl. for example a QUAD 405) you will see that there are minor changes in loudness across the audio band, related (somehow) to the electrical load of the speaker interacting with the amplifier. Those variations could be, say, ± 0.3dB or so even on well executed amp somewhere in the middle frequencies (say, 300Hz - 6kHz) which makes it a little tricky to pick a reference frequency for 'average' gain. Whether this frequency dependent gain change is audible is another matter: it's close to what I have found is the audibility threshold under super-controlled conditions.
Interesting. A change of 0.3dB suggests a change of around 3.5%, and to see that with a nominal impedance of 8 ohms, you'd need an output impedance of around 0.25 ohms (very roughly - I suspect that a more rigorous analysis might suggest a slightly higher Zout would be needed - but time precludes that right now).

I'd suggest that 0.25 ohms is pretty high for a solid-state amp with normal levels of feedback. I've seen about that with some esoteric designs, and was surprised to see that the BBC AM8/17 (which was principally designed to bolt onto the back of an LS5/9) has an 0.47 ohm output resistor. Of course, low Zout means high "damping factor" - a rather pointless specification once it's got to a certain point, but one that manufacturers like.

I'll endeavour to check a 405 at some point soon (I have a pair waiting to get on the bench for overhaul) - thanks for the pointer.

As I say, with valve amps, things are quite different. High output impedance is the order of the day, which makes everything difficult :)
 

A.S.

Administrator
Staff member
Measurements of QUAD 405s - an example of ageing?

Measurements of QUAD 405s - an example of ageing?

Interesting. A change of 0.3dB suggests a change of around 3.5% ..)
Rather than rely on memory here is a comparative measurement from my 2009 archive. The North American amp (blue) v what I called at the time a 'good 405' (black trace) and 'bad 405' red trace. Vertical trace position (gain) adjusted for visual offset only. The same speaker load was used for all. I don't recall which speaker.

As you can see, the 'good' 405 is (let's say) a max. 1dB variation; the 'bad' 405 perhaps 4dB. Most likely to be audible. See attached.

What is the likely cause for the 'goodness' of one amplifier compared with another off the same production line? Well, the 405s could be 25+ years old, one may have been serviced recently the other not. The 405 has no adjustable parts (it is 'self-calibrating') so one must assume component ageing beyond the self-correction range ... i.e. electrolytic capacitors? So you see, in the real world amps even of the same design can be quite different! Just as well QUADs service dept. is incredibly inexpensive and efficient.

>
 
Attachments only viewable to members

mhennessy

Member
Amps ageing and servicing

Amps ageing and servicing

Fascinating! It's hard to guess what could cause such a marked rise in output impedance - almost certainly capacitors, as you say. I normally replace them all without worrying about making measurements as I go (time is money!), but next time I'll do just that.

The bootstrap capacitor (C10) seems like an obvious suspect. C5 is known to cause hum when it fails, but as it's part of a current source acting on TR2, it could also cause problems. Basically, anything that can reduce open-loop gain is in the frame. I do think the 405 is somewhat unusual in using capacitors in these positions - but it was designed in an era when capacitors were cheaper than transistors.

Thank you for posting that. Especially as it verifies my belief that a good service is definitely worthwhile with these... And yes, you can hear a difference - at least, that's what I've found with the BBC LS5/8 version (AM8/16). The last pair that we had in service were unable to produce a central mono image because they were so mismatched, and by substitution we found that it was mostly caused by the amplifiers (one of which was humming badly).

I assume it was a multitone test - the noise from a manually swept test would have been somewhat unbearable :)

Cheers,

Mark

PS: The "self-calibration" you speak of is related to the bias setting. This is a tricky problem with conventional amplifiers, neatly side-stepped by the genius of current dumping. The issue concerns crossover distortion, and wouldn't have much impact on the measurements under discussion here. I could say more about this in the future if you like, but preferably after I've serviced the pair in the workshop, as there are a few things I'd like to double-check first. It's probably also getting a bit technical here :)
 

BAS-H

Member
Valve-Transistor A-B Comparison: open circuit difficulty

Valve-Transistor A-B Comparison: open circuit difficulty

Is it possible to undertake a controlled A-B comparison of a valve and a transistor amplifier? I have been longing to do this for a long time, since I have been convinced of my preference for the former. I am now doubting myself.

An irritating problem with a test involving a valve amp is the fact that they don't like open circuits, so swapping cables is not straightforward. You'd have to switch it off (audibly) each time. I have been struggling to design a test that results in the listener being entirely unaware which amp is operating. The closest I have come involves the assistant being in the next room with the amps, complicating cabling.

Any suggestions would be much appreciated.

{Moderator's comment: No, no, no. You cannot 'compare' anything and expect a reliable conclusion unless the time gap between A and B is the tiniest fraction of a second. Swapping cables around simply isn't good enough. Save yourself the bother and flip a coin. Or construct a proper comparison. This has been stated countless times here. Solve the valve open circuit issue ...}
 

Pluto

New member
Solution to A-B valve amp test

Solution to A-B valve amp test

An irritating problem with a test involving a valve amp is the fact that they don't like open circuits, so swapping cables is not straightforward
Could you not use a double-throw relay so that instead of switching the 'inactive' amplifier into an open circuit, it drives a dummy load? So while amp A drives the speakers, amp B drives the dummy load and vice versa when the relay state changes!
 

DSRANCE

Member
Amplifier distortion etc.

Amplifier distortion etc.

I'm wondering just how "good" some modern amplifiers actually are. The other thing is clipping performance I think, in that one of the amps mentioned earlier in this thread isn't particularly good in terms of measured distortion (although it lacks the predominantly 3rd and 5th artefacts of previous models in its ancestry) and doesn't soft clip, the resulting sound truly horrid if taken past its comfort zone but nice and punchy enough if used within it.

The OP seems to have an attentive dealer "up north" and was able to try his preferred amp at home to confirm he can live with it. I hope he's bought one and is enjoying the music coming through, which at the end of the day is what it's all about. The amps my mate hifi dave mentioned are also worthy contenders (I've heard them many times into Harbeths of all shapes and sizes :) ) and of course, Alan tells us that all of his designs offer a very benign load to any amp you care to choose.

The plots above of a good 405 against a well worn one are an excellent thing to show. Quad are indeed very reasonable on service..
 

BAS-H

Member
Switch over and dummy load and why instantaneous

Switch over and dummy load and why instantaneous

Could you not use a double-throw relay so that instead of switching the 'inactive' amplifier into an open circuit, it drives a dummy load? So while amp A drives the speakers, amp B drives the dummy load and vice versa when the relay state changes!
Thanks Pluto. That is a good suggestion if the dummy load connects before the speakers disconnect. I have read that if one subjects a valve amp to an open circuit for even a fraction of a second, voltage spikes can cause arcing across transformer windings and fry the insulation.

(I still do not understand why an instantaneous switch over is required. If one can identify which amp is operating under genuine double-blind conditions, every time, then I don't believe the time between samples is of consequence.)
 

A.S.

Administrator
Staff member
Time gaps?

Time gaps?

Thanks Pluto. That is a good suggestion if the dummy load connects before the speakers disconnect. I have read that if one subjects a valve amp to an open circuit for even a fraction of a second, voltage spikes can cause arcing across transformer windings and fry the insulation.

(I still do not understand why an instantaneous switch over is required. If one can identify which amp is operating under genuine double-blind conditions, every time, then I don't believe the time between samples is of consequence.)
I'm curious how you could arrange a reliable comparison if there were significant gaps between listening to A, then some time later B. How big a time gap would you be using?
 
Top