Amplifier Choice once and for ALL!

[Compo]

Dear whoever is in charge or has the know how or ability to consider my proposal and perhaps put it into action? Hopefully it may ease what seems to be a) the most asked question and b) the most annoying question to answer - over and over again?

Is there some way to set up a list where members can list Speaker type / Amplifier Choice / short reason for choice? If there could be some kind of easily accessible list available to review it may be easier for members and non-members alike to glean a "consensus of opinion" as to preferred amplifier choice for specific speaker types. If the member than has further "Amp Choice" questions he may contact another member directly who has made a similar choice for further advice, saving everyone having to read general “AMP” questions every other post. (I exaggerate!).

….or am I just a new member sticking his oar in? Perhaps it’s been tried before? Although I have been reading these forums for a number of years now!

{Moderator's comment: we're all for anything that lays to rest this subject but we do not want to become a free advertising window for products that we ourselves at Harbeth UK have no experience of ... Worth a try though!}

[DSRANCE]

This is so difficult to answer properly and I think this is the reason for the reams and reams written about such things.

I'd like a proper electronics design engineer to put me to rights on this, but my take is that amplifier damping factor (low output impedance a the speaker terminals) is of paramount importance when driving conventional speakers with passive crossovers. Some, slightly under-damped speakers love a good powerful amp with good damping, the bass in particular taking on more authority and clarity in my basic opinion and all else being equal.

On the other hand, older speakers with higher nominal impedance, lets say vintage Tannoys for example, which tend to have an absolute minimum of over eight ohms across the audio band I understand (more like eleven to fifteen ohms I think from memory), may not need this extra electrical damping and may sound over-tight and perhaps more coloured if driven that way, when a "looser" transformer coupled valve amp may well be better perhaps?

The above is just general feelings and I think provable, although I've never carried out detailed objective testing on this. I do remember some decades ago comparing a mid 70's Radford HD250 integrated amp with the then "top end" (and still worthy in my opinion) Crown "D" series of amps. Into some speakers, the radford lost bass in comparison and into others, had far too much with a "loose" quality to it. The Crowns sounding more even tempered into more speakers.

Hope this helps a little bit ;)

[denjo]

Hi Compo

Welcome to HUG! Interesting suggestion! Harbeth speakers are relatively easy to drive given their benign impedance. Under the FAQ, an amplifier with 25 watts (8 ohms) should be able to drive the Harbeths, 50 watts would be more than enough should you want to play louder.

The problem of having an amplifier/gear profile is that some members do change gear quite often (I have been guilty of that in the past) so there might not be a stable reference point in which to guide you with choosing the 'ideal' amplifier.

Having been on the 'merry-go-round', I am beginning to see the wisdom of Alan's advice to choose an amplifier that is reputable, reliable and has good after sales service and not base your decision on marketing hype. There are many brands to choose from in England (Rega, Naim, LFD, Arcam, Quad, Audiolab, etc) and most (if not all) should be just fine driving your Harbeths and providing you with hours of musical enjoyment.

Best Regards
Dennis

[A.S.]

This is so difficult to answer properly and I think this is the reason for the reams and reams written about such things.

I'd like a proper electronics design engineer to put me to rights on this, but my take is that amplifier damping factor (low output impedance a the speaker terminals) is of paramount importance when driving conventional speakers with passive crossovers. Some, slightly under-damped speakers love a good powerful amp with good damping, the bass in particular taking on more authority and clarity in my basic opinion and all else being equal....

Well, the working assumption in your opening remark is that there is, truly, repeatedly, universally, unequivocally, a significant audible (rather than emotional-subjective) difference between amplifiers. To even begin to decide if that is true or not, they would have to be compared under controlled conditions. As we've seen recently, not one single commentator stepped forward to take up the challenge and put that important belief to the test. So can we agree that, at this time, these claimed night-and-day 'sonic differences' are completely unproven and the validation of this claim remains a work in progress?

Second: as I recall, this damping factor issue has been thoroughly explored. I suspect that most/all/any solid state amp made in the last 30 years or so should have a very low output impedance and should, I'd strongly suspect, meet your target for adequacy of damping factor. Indeed, don't forget that there is (typically) an ohm or so of DC resistance between the amp's output terminals and the woofer/tweeter (in the crossover) so even if the amp has an output resistance of (typically) 0.1R, what the woofer 'sees' is about 1R, including the crossover and cable wiring resistance. As for tube amps which are critically dependent on their huge audio output transformer with miles of wire inside (all of which has a resistance) the output resistance is much higher and hence the damping factor inevitably lower.

Here in Windermere I don't have my full archive, but I'd be very surprised if your entirely reasonable focus on damping factor will yield the answer. What's next on your radar I wonder? We need to know!

P.S. If you think about it, it really doesn't matter to us ordinary music lovers what the technical reason for amp A sounding better than amp B is (if that is indeed provable). Surely all any of us want to do is get straight to that amplifier nirvana, get out the credit card and buy it pronto. The only way to reliably identify that golden amplifier I have exhaustively proposed, with nil take-up. So as this thread suggested at the outset, we are all still stumbling around in the dark in the same time frame that man has been to the moon and back. What engineering issues are holding us back from once and for all, nailing the amplifier argument? Can you think of any? If the instantaneous A/B test can identify one superior amplifier count me in as the first to buy it, cost irrelevant.

[Compo]

Phew!

My original reason for posting this thread was really for some reassurance on amp choice. Reason: I have dreamt of a pair of Harbeth speakers for the last 30 years making do with a pair of HHB Circle 5's in my humble studio for about the last 20.

Having saved and saved during this time I have managed to get 1 rung up the ladder and got hold of a pair of original HLP3s. Having first tried them with a consumer type Technics amp I nearly started a PayPal claim "not as described" !! I won’t even begin to tell you what I heard! However when a friend brought round a very old Creek 4040 they absolutely sang!!
Trawling the internet, most of the reviews of the original HLP3s always seem to mention its difficulty to drive. This combined with my recent experience concerned me!

Now the problem is again one of cost. I have limited funds (sub £200) to try and get the best integrated amp possible. (2nd hand of course). The only other requirement I have is it must have at least 4 line inputs and a Pre-out. I was looking at Musical Fidelity X-A1 or A2. So in a roundabout way I was hoping that I could find someone else using the HLP3s with a similar amp that I could afford, just to put my mind at rest after my "Technics" scare!!

I run a small vocal, sound effect, video studio from home!

{Moderator's comment: then there must have been a fault with the Technics or it was intolerant of the (original) HL-P3s 3-4 ohm impedance minima. It beggars belief that a credible brand like Technics (aka Panasonic, the world's largest consumer company) couldn't produce an amp with adequate performance. What model was it exactly? We'd like to buy one. BTW, the Circle 5/DMP1 is a really excellent speaker and an upgrade would be to the similarly sized M30.1 or C7ES3.}

[A.S.]

... My original reason for posting this thread was really for some reassurance on amp choice...

Let's restate our position again. I work damned hard to make sure that my whole life can be lived without anyone ever expending a molecule of anxiety about 'what amp should I use with my Harbeth speakers'. That's because (unlike many other speaker designers it would appear) I simultaneously optimise the perceived sonic balance, the frequency response and the impedance presented to the amplifier. I can virtually guarantee that any hifi separate amp made in the past 30 years that is working to original specification with work flawlessly with Harbeth speakers.

I truly believe that your experience with the Technics (Matsushita/Panasonics) amp is atypical. There must be a measurable reason why the sound did not appeal to you. I can honestly say that in 30+ years of audio listening to whatever amp is to hand or is loaned to us for an exhibition, I have never once been concerned about a mismatch and I have never rejected an amp for reasons of unacceptable sonic quality when the room and the music are unfamiliar. I have merely selected an amp on a) brand reputation for after care preferably from my personal experience b) power output c) facilities and features.

Therefore, I caution you about drawing universal conclusions about this inexplicably bad experience you have had. Technics is a brand from the best resourced consumer electronics company on the planet with thousands of skilled engineers to draw on*. They don't make basic mistakes in audio amplifier design, which is a highly mature technology and leaves no real room for innovation or significant performance variation and must be reliable to avoid warranty claims. If this particular Technics amp really did give you cause for concern (and I am more than prepared to believe you), I really would like to buy or borrow it. Using my relay change-over box it should be easy to sonically identify it in an instantaneous A-B comparison and if so, it could be a most useful 'control'. We could learn much from this.

*I would expect that Technics group has more design engineers available than the entire high-end audio industry multiplied by 50. Matsushita-Panasonic are an extremely serious, credible business.

[Compo]

Thanks Alan, that has put my mind to rest somewhat in my imminent amplifier choice.

However.... I shall be taking you up on the offer of that test at some point!
Even if it’s just for a free tour of the factory!! (maybe??)

The Technics Hi-Fi amp I used (SU500) is probably almost 30 years old itself. I use it with a pair of original Royd Coniston R's and its fine! Even when I push it a bit! With the HLP3s (just had another listen) anything above a low listening level makes the bass verrrrrry noticeably woolly, kind of like the drivers cones are too loose and bottoming out! Bubbling?? Very scary! Sounds like someone has gone mad with a bass processor in the recording studio and the speakers can’t cope!

Back on the Creek - and I can hear the bass end clearly and although constrained at higher levels, its clean and what you would expect from this size of speaker.

Is it West Sussex I come to or did I read somewhere about sunny Windermere? (Don’t worry I’ll give you fair warning!!)

[EricW]

I think what Compo's conclusion is an excellent example of how we humans are wired to reason inductively (start with a specific observation or observations, and extract from that a general principle or rule) without at the same time being mindful of the fact that inductive reasoning always has the potential to lead to the wrong conclusion, even if the observations and premises are individually valid.

Here the chain seems to be:

Premises:

1. My attempt to drive Harbeth HLP3s with a Technics amp was unsatisfactory.
2. My attempt to drive Harbeth HLP3s with a Creek amp was satisfactory.
3. Some internet reviewers say that the HLP3s were hard to drive.

Conclusion:

Any Technics amp is inadequate to drive any Harbeth loudspeaker.

The form of the reasoning is not wrong; however, as you point out, there are other more nuanced conclusions that are more likely true than the one reached by Compo.

I think that the problem of reaching valid general conclusions based on specific observations is always a complicated one, and may account for some of the misinformation that exists in the audio world (not to mention the world in general).

[Pharos]

This is a very interesting aspect of audio for me, and I want to describe some relevant experience.

In the 70s I was running a home built Nelson-Jones 10+10 class A amp, (from Wireless World and with 3 signal path coupling caps, one to speaker), fed from a Quad 33, and feeding Tannoy Gold Lancaster 15"s.
I bought a Quad 303 from a BBC engineer, and whilst there was a little more headroom, the bass was very sloppy and soft, and so I sold it.

More recently, in '02 after acquiring my current speakers second hand, and rebuilding them, I was without an amplifier other than my Nait 1 which I use for feeding headphones in my singing cubicle. I used the Nait temporarily until a friend brought round two amplifiers to try. One was a Marantz, I think about 25 or 50W, which was enough for a 91dB/W speaker, and the other an ATC SIA2 integrated (150W).

The Marantz had a 'glazy' fine 'buzz' which it superimposed on the sound, which the other amp did not. Both were good at controlling the bass in damping terms. I was surprised at the Marantz, which is I believe supposed to be a high end product, and so bought the other, (which has been praised widely).

Related add-on of experience. After that I went through several loudspeaker cables over some weeks. First I bought for financial reasons some very thick Concert Cable, (racing green hosepipe 2.5mm) for £3, then I tried QED Silver Anniversary which was bright in comparison, and then Kimber 4TC which was more balanced in spectral terms. Lastly I tried Kimber 8TC at £400.

Most surprisingly to me and my audiologist friend, this produced a sound which was like a 5dB bass boost, and with the top very unfocussed and seeming to have lost its transient capabilities.

I returned the latter, and was, and still am eight years later fairly bemused about what happened, and cannot even propose an explanation. I am confident in the accuracy of perception of all these changes even though the s.p.l.s were not tightly controlled.

My audiologist friend confirmed my perception, and, to validate it I cite the following. In the 70s, and whilst working in a professional broadcasting capacity, a fellow worker tested my capacity to hear small changes in the level of a 1kHz tone. The 'wisdom' at the time was that less then 1dB was inaudible on tone and 2dB on programme, and I reliably could detect 0.25dB change in tone level.

In '95 I bought an expensive pair of active 100L monitors, and, living on the sea front with no traffic noise, the ambient noise level is very low. I realised after a couple of days that I could hear a hum from the left speaker, and returned it to the manufacturer, they changing the mains transformer to reduce the hum. I guessed by ear that the hum was at about 30dB s.p.l., and the manufacturer said that he measured it at 33dB s.p.l.. I don't think I have cloth ears, and I think there are audible and unexplainable changes to sound which may seem irrational.

However, when my audiologist friend says that he can hear a change when someone replaces a standard mains fuse with a gold plated one, I have to gulp.

[A.S.]

I think what Compo's conclusion is an excellent example of how we humans are wired to reason inductively....

Thanks for these contributions overnight. At 4.03am my subconscious dragged up the recollection that from about 1989 through to 2011 - that is, during the period that the P3, M40, C7, P3ESR etc. were in production, the standard measuring amplifier in our lab was a Technics integrated. Following your tip I think it was an SU something - 5050? - in quite a large case, tone bypass switch, non remote. It was used to power the speakers to make the finest most intricate acoustic and electrical impedance measurements and worked perfectly - until last year when the A-B output switch became unreliable and with great reluctance I took it to the municipal recycling centre as end-of-life.

From your description of the bass sound when the Technics amp is driving the original P3s with their low impedance (= high current requirements) you may have a classic case of partial amplifier failure.

With the HLP3s (just had another listen) anything above a low listening level makes the bass verrrrrry noticeably woolly, kind of like the drivers cones are too loose and bottoming out! Bubbling?? Very scary! Sounds like someone has gone mad with a bass processor in the recording studio and the speakers can’t cope!

seems to correlate perfectly with severe crossover distortion in the amplifier. This is where the two halves of the audio signal no longer perfectly match at the central or crossover point between them, and what you hear is chronic distortion - 50% or more distortion would be expected. Because it is the bass frequency region which draws much power from the amp, this subjective effect whilst present to a lesser degree at all frequencies, is magnified at low frequencies, where, incidentally, the ear is not normally very sensitive. You can read about amplifier crossover distortion here. You can see the kinking in the audio waveform on this video - tube amps and solid state amps are conceptually the same in this regard.

So how does amplifier crossover distortion occur? Basically component ageing. That's why I frequently caution about buying old amps and set my primary amplifier procurement parameter as not "sound quality" but after care. Some amp designers, living in the real world of component ageing and conscious of long term customer care - such as QUAD amps from the 405 onwards - made them self correcting. Yes, that's right: they designed out the trimming controls that are themselves one source of drift and built intelligence into the amp to keep them correctly biased. And that philosophical approach really appeals to me. Unsubstantiated claims of night-and-day sonic quality leave me completely cold.

Incidentally, it is theoretically possible that your Technics amp could be nursed back to health. It may only need a 60 second adjustment to a variable resistor (it clearly isn't self-adjusting) but who will service it for you? Your local hi-fi dealer? I don't think so. Technics UK? Most unlikely - so it is now junk for the sake of apathy by its maker and representatives. Is that a good use of natural resources? I don't think so. So I am back once again to my mantra .... buy audio equipment with one eye on who will service it for you in 5, 10, 20 years.

Caution about your old Technics - sooner or later it will completely fail and destroy the speakers. But I think a perfectly functioning Technics amp of adequate power output "working to original specification" would comfortably drive any Harbeth speaker.

P.S. Here are some audible examples of crossover distortion - deliberately introduced to harden the sound of an electric guitar, here.

[Compo]

Blimey! I'm now fearing for the life of my trusty Royd's!!! Think I'll have to get a new amp for these now!

Anyone know of a Royd Speakers forum where I can ask what’s the best.................He He!

Anyway thanks for all the psychological and technical (or is an examination of how one might think technically called Technicology?) advice everyone I'm off for some psychotherapy... how are you fixed EricW?

Oh and I'm buying a Quad!!!