Modern loudspeaker production

[cloth-ears]

Following on from a recent post of mine (that was posted in the wrong place) I have decided to share with those interested the thoughts and findings of a good friend of mine who works as a consultant technician for a well known and respected loudspeaker manufacturer.

Over the last ten years or so the technology involved in designing and building speaker drive units have come along in leaps and bounds. New super strong lightweight materials coupled with advance forming and computerised milling techniques are making speakers that are better than ever

Years ago the British company KEF used to check and measure every drive unit. Those that were very close to spec were matched up and sold as “reference” speakers. With today’s technology every speaker is the same. You could take one off the line on a cold Monday in January and another on a hot day in June and both would be classed as reference speakers

So tight are the gaps that if one were damaged it would not possible to have it re-coned. A whole new drive unit would be required. Provided the other speaker was undamaged you would only have to buy one driver. And this is a very interesting point. Despite the fact the replacement drive unit was “new” it should still match the older undamaged unit

When I questioned him about the “burn in period” he said that any tests they had done did not support this phenomena. In essence, if you don’t like the sound straight out of the box you were unlikely to change your mind at a later date. This must be taken in the context that other smaller manufactures who to do not have access to the same level of technology may well need time for their drive units to settle down. That may indeed be a matter of opinion

Another interesting point is that of bi-wiring. All of the tests his company have done do not show any relationship between bi-wiring and an improvement in sound quality. The only reason they supply multiple binding posts is because the customer demands it. Although it ultimately adds unnecessary cost to the speakers they do not want to be the first manufacturer to remove them. They do however supply professional monitor speakers to industry without the bi-wire option.

So that’s about it, make of it what you will:scratch::scratch:

[Joe]

None of this particularly surprises me. Modern production methods centre on reliability and consistency of products. It's only loony audiophile types who hunt down examples of ancient obsolete technologies, which they discuss in hushed, reverent tones.

I thought bi-wiring (as opposed to bi-amping) had long ago been debunked as a gimmick dreamed up by cable manufacturers wanting to double their already massive profit margins?

[Ammonite Audio]

Another interesting point is that of bi-wiring. All of the tests his company have done do not show any relationship between bi-wiring and an improvement in sound quality. The only reason they supply multiple binding posts is because the customer demands it. Although it ultimately adds unnecessary cost to the speakers they do not want to be the first manufacturer to remove them. They do however supply professional monitor speakers to industry without the bi-wire option.

So that’s about it, make of it what you will:scratch::scratch:

Hardly earth-shattering news. Focal eschew bi-wiring; so do Avalon and (I believe) Wilson. I'm sure that there are others.

[Clive197]

I


..........Another interesting point is that of bi-wiring. All of the tests his company have done do not show any relationship between bi-wiring and an improvement in sound quality. The only reason they supply multiple binding posts is because the customer demands it. Although it ultimately adds unnecessary cost to the speakers they do not want to be the first manufacturer to remove them. They do however supply professional monitor speakers to industry without the bi-wire option.

So that’s about it, make of it what you will:scratch::scratch:

An interesting point. I have just replaced my Spender S8e's (4 binding posts for bi-wiring or bi-amping) to Spendor D7's which only have 2 binding posts. I am no longer able (which I used too) bi-wire. Do I miss that. HELL NO.

So your friends company would not be the first to not bow to fashion.

[struth]

Some speakers can benefit from it depending on their xover designs I am told, but Bi Amping, both vertical and horizontal, I have found to be of benefit

[fatmarley]

Driver tolerance can still be quite bad depending on the material used and the manufacturer. I know this first hand because I've measured a few drivers and also because I use a lot of diy speaker forums. I think it's easier to get consistent results with plastic and metal drivers but paper can be a bit more variable.

As for burn-in, It's measurable and mostly related to the spider loosing up. Subjectively, I think I've heard a driver run-in but I'd imagine it depends on the driver and I wouldn't be surprised if some don't sound any different.

[Barry]

Some speakers can benefit from it depending on their xover designs I am told, but Bi Amping, both vertical and horizontal, I have found to be of benefit

http://theartofsound.net/forum/showthread.php?5818-Basics-amp-General-Bi-amping-and-Bi-wiring

[struth]

http://theartofsound.net/forum/showthread.php?5818-Basics-amp-General-Bi-amping-and-Bi-wiring

trying to tell me summat Barry:eek:

[hifi_dave]

Few of the speakers I have had through my demo room, the past couple of years have been bi-wirable. It was a fad - a long running fad but now no longer fashionable. I have never found any benefit from bi-wiring except for getting rid of the (usually) brass links. Replace these with a piece of decent wire and you have bi-wiring for free.

Several years ago, Revel (Harman group) did an exhaustive study of bi and tri wiring. Their findings, which they published, confirmed that bi-wiring had no benefit and under some conditions can make the sound and performance worse.

I agree about the running in bit. Any 'improvement' is small and usually occurs within 24 hours or so. If a speaker sounds horrid out of the box, it ain't gonna sound wonderful after 'running in'.

[Arkless Electronics]

Bi-wiring may be of some benefit if the circumstances (domestic?) require long and thin speaker cable.... but then you obviously need two runs which may negate any WAF increase! Other than that no it's a fad.

[Reffc]

What you've written Bob coindices with my own findings, by and large anyway. I have tested new speaker drive units for consistency and on average most are matched closely with tolerances on acoustic response of a few percent or less being typical of even budget drivers.

The discrepancies come when drive units from some manufacturers are tested against manufacturer's data sheets when it becomes apparent that a surprising number do not tally, yet remain consistent between drive units (I wont mention the companies). This can be a real pain for DIYers who rely on accurate information to design cabinets and crossovers where they may have no specialised tools to help them.

On most modern rubber surround midwoofers, the differences between out of the box and say 100 hours playing can be just one or two Hz at resonance. In one case, after 200 hours, they measurements for one driver were identical to brand new.

Where some can require more bedding in is dependant upon spider suspension design and materials as some do in fact take a little bedding in period but the changes can be slight and bedding in not that long (modern Tannoys for example).

Some manufacturers still insist on 100 hours or more run-in and oddly when the drive units are checked against suppliers data, those drive units should require little or no bedding in so this may be more a marketing ploy.

As far as bi-wiring goes, I agree with you. There's some technical merit in not bi-wiring because single wiring with common connections back to the amp offers some electrical damping of the crossover, albeit at probably inaudible levels, but you could still say (accurately) that bi or tri-wiring actually creates more (tiny) distortion. Bi-amping is different and can be very beneficial.

[cloth-ears]

Unfortunately Paul there is no real solution to this. Manufactures rarely if ever tell you which equipment they used to come by their specifications. As you are all too aware different amplifiers will have different results. My friend for example uses a Quad modified 606 and I believe that at some time B&W used Myryad but I may be wrong on that. You then have to take into consideration the microphones and other ancillary equipment used during testing. The permutations are almost endless

I’m just happy that we agree on something. I'm welling up :)

[Barry]

Bi-wiring may seem to be a fad, but sombody must have noticed an improvement and reported it for it have caught on. (Or is it a marketing scam by cable manufacturers to sell more cable? :D) I suspect that it all depends on the design of the speaker and of the crossover; the better designed speakers being less amenable to improvement through bi-wiring.

For a possible explanation as to why bi-wiring might be beneficial, see the relevent section in: http://theartofsound.net/forum/showthread.php?5818-Basics-amp-General-Bi-amping-and-Bi-wiring, and references therein.

[cloth-ears]

Bi-wiring may seem to be a fad, but sombody must have noticed an improvement and reported it for it have caught on. (Or is it a marketing scam by cable manufacturers to sell more cable? :D) I suspect that it all depends on the design of the speaker and of the crossover; the better designed speakers being less amenable to improvement through bi-wiring.

For a possible explanation as to why bi-wiring might be beneficial, see the relevent section in: http://theartofsound.net/forum/showthread.php?5818-Basics-amp-General-Bi-amping-and-Bi-wiring, and references therein.

Unfortunately Barry your argument is flawed. Not because of the results but because of the way they were achieved. Bell wire for example is poor at transferring current, therefore doubling up will allow more current to travel. But that should not be seen as an improvement; instead it would be better to use a suitably large cable in the first place. Basically, if one small glass could fill a bucket in twenty minutes then two glasses would seem to be an improvement as they could do it in half the time. The real improvement would be to fill a bucket with a bucket in real time. In other words, if the cable is up to the job then no further improvement can be made. That is not an opinion but a provable scientific fact

[DSJR]

Drivers running in is a sore point with some driver manufacturers I think. The main deformation of the spider takes place in the first few cone excursions I'm told and 'shouldn't' change thereafter. ATC used to throttle their bass units at 20Hz for twenty minutes, the feeling being that if they lasted through this, they'd do for decades in a rough professional environment.

Tweeters are a different matter though and some older ones lose output over ten or more years of hard use. Once splashy Mission 770 mk1's should sound very sweet toned now (I have an experience with these but for another time) and I remember Rogers LS7T's were much better after ten years than when new...

[Barry]

Unfortunately Barry your argument is flawed. Not because of the results but because of the way they were achieved. Bell wire for example is poor at transferring current, therefore doubling up will allow more current to travel. But that should not be seen as an improvement; instead it would be better to use a suitably large cable in the first place. Basically, if one small glass could fill a bucket in twenty minutes then two glasses would seem to be an improvement as they could do it in half the time. The real improvement would be to fill a bucket with a bucket in real time. In other words, if the cable is up to the job then no further improvement can be made. That is not an opinion but a provable scientific fact

They're not my arguments, they are arguments made by the authors cited in the reference. It would seem the argument has nothing to do with optimal current transfer, more the damping of intra-crossover interference.

[cloth-ears]

Drivers running in is a sore point with some driver manufacturers I think. The main deformation of the spider takes place in the first few cone excursions I'm told and 'shouldn't' change thereafter. ATC used to throttle their bass units at 20Hz for twenty minutes, the feeling being that if they lasted through this, they'd do for decades in a rough professional environment.

Tweeters are a different matter though and some older ones lose output over ten or more years of hard use. Once splashy Mission 770 mk1's should sound very sweet toned now (I have an experience with these but for another time) and I remember Rogers LS7T's were much better after ten years than when new...

Well David, Not too sure if that was meant as a light relief or not. I don’t think too many people would buy a speaker on the off chance that it may sound better in ten years time but you never know. Today’s hi-tec speaker’s use non-pours thermally inert components for the suspension and are unaffected by outside influences such as humidity etc. What it sounds like out of the box is what it will sound like throughout its entire service life. As I said earlier, this basic rule of thumb may not apply to smaller manufactures who do not have access to modern production techniques or materials

[cloth-ears]

They're not my arguments, they are arguments made by the authors cited in the reference. It would seem the argument has nothing to do with optimal current transfer, more the damping of intra-crossover interference.

Intra-crossover interference would only apply if they were fed by separate earths. This is of course not the case unless you are bi-amping the speakers. The single amplifiers common earth will always apply

[Barry]

Suggest you read the synopsis for the case of bi-wiring that I cited in The Knowledge, and take it up with the respective authors referenced therein.

[cloth-ears]

No need, I know the answer. But thank you anyway.

[Reffc]

Intra crossover electrical damping: I would suggest, if a problem at all (a problem looking for a solution in terms of measurement methinks?) is better served, by single inputs allowing out of band damping across the crossover. Series crossovers especially benefit from this although I am not a fan. As someone who designs loudspeakers, I can say, hand on heart, there is no technical merit or argument for bi-wiring other than decreasing reactive and total impedance where single wiring cross section may be insufficient. If damping factor is that poor to demand this or things that marginal, buy a more powerful amplifer ;)

[Barry]

I am no champion or proponent of bi-wiring, I am simply curious as to why this technique became popular; to the extent that many loudspeaker manufacturers provided this facility. What was it that the original proponent (who he?) heard/observed that caused this idea to convince many that this was a 'Good Thing'?

Some theoreticians have offered an explanation:

How and why does bi-wiring work?

A very good article on this appeared in Hi-Fi News magazine in October 2005. Written by Keith Howard under the title ‘Double Take’ [1], it describes what bi-wiring is supposed to do and describes attempts to measure the differences between single and bi-wiring. The following is a synopsis of the theories about bi-wiring. The first theory, proposed by Jon Risch [2], makes the perfectly understandable observation that as well as acting as radiators of sound, drive units also act as microphones and therefore as signal sources. That is, the pressure differential created by, say, the bass unit will act on the tweeter and vice versa. Such signal sources will inject voltages back into their respective parts of the crossover network. If the speaker, or rather the crossover network, was connected to a power amplifier having an ideal output impedance of near zero, then the generated signals would be shorted out. However all power amplifiers have a finite impedance and worse, so too does the single speaker cable pair used to connect the speaker to the amp. The presence of a finite, non zero impedance upstream of the microphonic drivers means that the shared common input allows an attenuated version of the signals generated by one driver to affect another.

An alternative theory, suggested by John Wilkinson [3], is that all loudspeaker drive units are inherently non-linear, and as a result, the current drawn from the power amplifier contains harmonic and intermodulation products not present at the amplifier’s output terminals. If the speaker were directly connected to the amplifier and the amplifier had a near zero output impedance, the non-harmonic motion of the drive unit would be heavily damped. The presence of finite amplifier and finite speaker cable impedance means that the distorted driver currents set up a distorted voltage across the speaker terminals. Thus the non-linearity of one drive unit would modify the signal reaching the second, if both are fed through their crossover filters from the same input terminals.

What bi-wiring does is to increase the path length, and because the cable has finite resistance, to dampen the interfering signals as they travel away from the speaker towards the amplifier and then back again along the separate second cable path. This is similar to the damping of mains noise in a spur wired mains system: noise generated and injected into one mains socket is damped out by the finite resistance (voltage drop) of the cable run back to the meter then out again to a socket supplying the audio system. In the case of a bi-wired loudspeaker, using two cables of the same gauge will not degrade the effective damping factor of the amplifier; the amplifier’s control of the speaker is unaffected.

An additional advantage offered by the option of bi-wiring is that the user can select different gauge and styles of wire according to the frequency path. Thus the LF bass driver/ filter might use a heavy cable of large cross section, whereas the HF tweeter might use a smaller gauge, and even use a different conductor altogether. There is thus ample scope for experimentation and hair-tearing indecision given the range of options!

Many speaker manufactures offer the facility to bi-wire their speakers. In some, this is because it is genuinely though that bi-wiring can confer audible advantages, for others it is simply there because the market expects it; they are simply pandering to fashion. In all cases speakers with multiple terminals are supplied with bridging links or bus bars for those users who do not want to try bi-wiring. It has been claimed by some that there is an audible improvement if these links are replaced with jumper leads made of the same kind of wire used to connect the speaker to the amplifier. Unless the bridging links are made of (gold plated?) brass rather than copper, it is hard to understand how a couple of inches of multi-stranded wire can better a solid metal strip.

[1] Howard, K., ‘Double Take’, Hi-Fi News, October 2005, pp.96-101.

[2] Risch, J., ‘A New Class of In-Band Multitone Test Signals’, Preprint 4803, Audio Engineering Society 105th Convention, 1998 (available via www.aes.org)

and www.geocities.com/jonrisch/bi-wiring.htm.

[3] Watkinson, J., Studio Sound, 2004.



I should add that Keith Howard in the above cited article attempted to measure the difference: what differences he measured were minute and unlikely to have any significant or detectable audible effect.

I am sceptical of any benefit of bi-wiring, whereas I do believe in the benefits of (active, 'vertical') bi-amplification.

[Arkless Electronics]

Intra crossover electrical damping: I would suggest, if a problem at all (a problem looking for a solution in terms of measurement methinks?) is better served, by single inputs allowing out of band damping across the crossover. Series crossovers especially benefit from this although I am not a fan. As someone who designs loudspeakers, I can say, hand on heart, there is no technical merit or argument for bi-wiring other than decreasing reactive and total impedance where single wiring cross section may be insufficient. If damping factor is that poor to demand this or things that marginal, buy a more powerful amplifer ;)

An amplifiers power has nothing to do with its damping factor ;)

[BTH K10A]

Following on from a recent post of mine (that was posted in the wrong place) I have decided to share with those interested the thoughts and findings of a good friend of mine who works as a consultant technician for a well known and respected loudspeaker manufacturer.

Over the last ten years or so the technology involved in designing and building speaker drive units have come along in leaps and bounds. New super strong lightweight materials coupled with advance forming and computerised milling techniques are making speakers that are better than ever

Years ago the British company KEF used to check and measure every drive unit. Those that were very close to spec were matched up and sold as “reference” speakers. With today’s technology every speaker is the same. You could take one off the line on a cold Monday in January and another on a hot day in June and both would be classed as reference speakers

So tight are the gaps that if one were damaged it would not possible to have it re-coned. A whole new drive unit would be required. Provided the other speaker was undamaged you would only have to buy one driver. And this is a very interesting point. Despite the fact the replacement drive unit was “new” it should still match the older undamaged unit



Call me cynical but I see things in a different light

Drive units can now be built more cheaply using lightweight polymers and less labour.
Manual quality control and matching costs money.
Serviceability is no longer built into products and there is greater profit in providing replacement units.

One has only to look at the Vox Olympians to realise that producing quality engineered drive units is a very expensive process for loudspeaker manufacturers who have generally been dumbing down on engineering quality since the 70's in order to maintain price points.

[cloth-ears]

Call me cynical but I see things in a different light

Drive units can now be built more cheaply using lightweight polymers and less labour.
Manual quality control and matching costs money.
Serviceability is no longer built into products and there is greater profit in providing replacement units.

One has only to look at the Vox Olympians to realise that producing quality engineered drive units is a very expensive process for loudspeaker manufacturers who have generally been dumbing down on engineering quality since the 70's in order to maintain price points.

I can understand your cynicism and I’m sure that in some cases that may well be the truth. However, it would be wrong to tar all manufactures with the same brush. Although modern manufacturing techniques can and often do reduce costs it should not always follow that such cost savings should equal an inferior product.

As an example, the flat copper wire is bonded to a specially developed alloy tube that is designed to only expand along its length and not its width. This particular innovation allows for a much tighter tolerance in the flux gap. It also means that it is almost impossible to repair. You really can’t have one without the other. Providing the speaker is used within its design parameters it will have a long and trouble free service life.

[Reffc]

An amplifiers power has nothing to do with its damping factor ;)

I know Jez. I was referring more to the suggestion that "...if things are that marginal..." about lowering impedance by increasing gauge. I should have split the sentence up as it reads badly.

[Joe]

Call me cynical but I see things in a different light

Drive units can now be built more cheaply using lightweight polymers and less labour.
Manual quality control and matching costs money.
Serviceability is no longer built into products and there is greater profit in providing replacement units.

One has only to look at the Vox Olympians to realise that producing quality engineered drive units is a very expensive process for loudspeaker manufacturers who have generally been dumbing down on engineering quality since the 70's in order to maintain price points.

Call me cynical, but the Vox Olympians look more like a product built to a price than a reflection of material and labour cost.

[BTH K10A]

Call me cynical, but the Vox Olympians look more like a product built to a price than a reflection of material and labour cost.

They probably are but with regard to the drive units the point is still valid.