Tuning KLS9

Article Index
Tuning KLS9
page 2
All Pages
This article is reprinted unaltered from Hi-Fi World DIY Supplement No29, June 1997 issue. It is published for reference purposes for those who use KLS9s today. We do not supply kits anymore, and cannot help with parts, service or technical queries.

Noel Keywood explains how to adjust our latest DIY loudspeaker, KLS9, to suit your tastes and how to tweak it for higher performance. 
In the April 1997 issue DIY Supplement we published plans for our new KLS9 two-way, floor standing loudspeaker. By using Audax's HM210Z0 High Definition Aerogel bass/midrange drive units in large-ish cabinets this loudspeaker produces a clear, uncoloured midband and plentiful deep bass. The HM21OZO crosses over to a TW025MO fabric dome tweeter, which extends output up to 20kHz. 
The idea behind our KLS9 DIY loudspeaker was that it should offer great results at a low price (just (230), a goal it successfully achieves. It is a relatively large (93cms high) floor stander that can be fine tuned according to taste. 
The 'speaker is sensitive to port dimensioning, giving constructors the chance to vary bass quality from rich and fulsome to dry and tight. KLS9 goes very low as well and can really shake the room. The precise quality of bass it produces can be adjusted, which is an interesting experience. It is one I performed taking into account comments from other listeners, by the way. I say this to underline the fact that although our test equipment provides a comprehensive picture of low frequency behaviour, allowing performance to be adjusted at will, there is virtually no literature relating operating conditions to perceived sound quality. And very difficult it would be to provide too. This being the case, bear in mind that no equation can give a notionally correct result. This is a misunderstanding, as I will explain. 
Unfortunately the wrong tweeter phasing was printed in the first article. The tweeter must be connected in-phase, not out-of-phase as originally shown. 
Bass sound quality was much as expected but I've been pleasantly surprised at how low this 'speaker will go. KLS9 plays bass lines cleanly and evenly, separating them from a mix with ease. It does so at low volumes too - this is not a 'speaker that demands huge amounts of power. Whilst bass quality can be changed significantly by reducing cabinet volume, I believe most people will find the variations produced by port tuning adequate. 
It's wise to bear in mind that when fine tuning bass subjectively you are not only working from your own tastes and expectations but are also listening to the combined behaviour of KLS9 and your listening room. Any 'speaker which goes very low (i.e. lower than 60Hz) will excite room modes strongly. If you tune the port so it hits a room mode, you'll get mighty bass at this frequency. It will only be apparent when an instrument hits and holds for a small time the particular frequency which coincides with this mode. So what I heard will not necessarily be what you hear because your room may be very different to ours. 
If you think this confuses the whole issue of loudspeaker bass quality in a room, then you're right! And that is yet another benefit to tuning your 'speakers subjectively in your own listening room. 
Which brings me onto one quick but important point. Some technically minded readers seem to regard theory as a way of achieving practical perfection. This isn't so. Because equations give an apparently 'intelligent' result does not mean the result is a truth. The result is valid only within the particular presumptions and simplifications which allowed the equation to be derived in the first place. 
So there is no one 'correct' cabinet volume or port size, as some appear to believe. There are various target values that come about in response to certain design criteria. You can, as I did, calculate port length for KLS9. It turns out to be around 3in. (7.5cms), according to the equation used. Audax on the other hand suggest 3.6cms. 
These are useful approximate values that get us into the target area. 
We know a port of this length should give decent results - and it did. But no port at all, just a hole in the cabinet, also gave very interesting results. Alternatively, a 6cm port with a small amount of wool damping gives another useful result. As does a port stuffed with wool whereupon KLS9 becomes an infinite baffle loudspeaker. All this goes to show how complex and variable port tuning is. 
For the purposes of this article I measured and listened to a short port, an optimal length (6cms) port and a long port (19cms). Port and cabinet damping were varied too, using Long Haired Wool. The general view is that reflex cabinets should not be internally damped since this reduces the effectiveness of the port. Like most 'rules' concerning loudspeakers, this contains a lot of truth. 
As always though the situation is sufficiently complex to defy simple rule making. In all cases below, our prototypes were lined throughout with half-inch thick fibrous carpet felt to absorb high frequency energy.

1) Short Port 
This comprised an 8cm diameter hole in the cabinet wall (19mm thick). 
With no cabinet damping it gave massive bass but there was a bit too much of it we felt and control was lacking in the lower registers. However, some people might well like the massive impact this gave drums and bass lines. 
Our analysis shows forward radiation from the HM21OZO reached down to 60Hz before diving downward rapidly. Port output forms a fairly high Q peak centred at 50Hz and its output extends down to 30Hz or so (-6dB).
2) Ideal Port 
This comprised a 6.6cm diameter port, 6cms long, the size given by various equations. Audax also suggest it is the right size. So what did we get? 
Bass again seemed a little too prominent and it was more resonant and less apparently controlled than the other options. However, just a small amount of wool in the port drastically damped down bass level and cured its boomy quality. Putting wool in the cabinet had a similar effect and gave a slightly cleaner overall sound. 
Measurement showed this port gives a rounded peak centred at 38Hz, extending port output down to 20Hz (-6dB), a very low value. Forward radiation from the HM21 OZO rolled off less steeply, allowing the unit to reach down to 50Hz.
3) Long Por
This comprised a 6.6cm diameter port 19cms long. Some authorities, like Vance Dickason, recommend using long ports. 
This gave the best results in our situation. We were looking for bass of a sensible level that had good definition and control. Adding wool to the 
bottom 20cms or so of the cabinet usefully reduced internal standing waves at 180Hz and 480Hz, attributable to the cabinet's height and depth dimensions respectively. This made the 'speaker sound just a fraction cleaner. Putting more wool in gave an unpleasantly boofy quality to the bass though. 
A small amount of wool in the port also helped control bass level and tightness and it was like this that we achieved the best general balance for KLS9. 
Measurement showed forward radiation reached down to 45Hz, so there was little change here from the 'Ideal' port. That was surprising considering port output changed a lot, the centre moving down to about 34Hz, the -6dB point measuring an amazingly low 12Hz! Our room reaches down to around 25Hz (-6dB) so such downward extension is a bit lost. I notice many readers have 20-24ft long rooms, 12ft wide. In a room like this KLS9 would have no trouble reproducing 16Hz organ notes at their correct level. 
The position of the tags changes on the tweeters; some are 180degrees apart, some 30degrees. This is why we left out clearance holes in the original cut-out diagram. Drill 'em when you've got the tweeters! 
Audax tell us, by the way, that they check tweeter phase with a 9V battery as they do with bass units. The convention is that when the cone/dome moves forward, the battery's positive terminal is connected to the driver's positive (red) terminal. 

As standard, KLS9 gives a flat frequency response with a little bass lift. Treble can be rolled off smoothly, if desired, to give a warmer sound by connecting a 1.5µF capacitor across the tweeter. This will drop output at 20kHz by -I dB or so; not a lot but quite audible all the same. 
The most striking feature of this loudspeaker is the sheer depth of its bass and the ease with which it reproduces low frequencies. There's little to touch it at present in this area. I can almost guarantee that it will astonish you in this respect but please bear in mind you will only get to hear and feel this in a room at least 15ft long. 
The High Definition Aerogel HM210Z0 bass/midrange driver offers superb midband clarity and good detailing. A smooth response and, absence of crossover suckout gives this 'speaker a very even tonal balance, which is also quite obvious subjectively. 
Like all our kit loudspeakers, KLS9 is a very easy load so it will work well with any amplifier. Most of the time I drive it from our K588 I Mkll valve amplifier, which delivers 20watts per channel. It was run from an Audiolab 8000S during development and for testing but we also used a Roksan Caspian. A 20watt NAD 310 solid-state amplifier was tried as 
well and I was pleased to find that it worked perfectly well with KLS9, making the NAD sound like a real power house. That's how it should be. At just £230 KLS9 is a lot 'of loudspeaker; I'm sure you'll enjoy it. 


Hi-Fi World, Powered by Joomla!; Hosted by Joomla Wired.