[ Home | Staff & Contacts | DIY & Tweaks | Listening tests | HiFi Playground | Music & Books ]
Products: Audio Nirvana "8" driver.
Manufacturer: Common Sense Audio - Made in China
Cost, approx: $120/95 Euro a pair...
Reviewer: Geoff Husband - TNT France
Reviewed: May, 2006
Two things prompted me to chase up these drivers the first was that the BFB was obviously being built by a lot of first time speaker builders, and that though good, some people were finding the Fostex 204 impossible to get (out of production) and the 206 difficult to get I needed an alternative.
The second reason was that the T-Amp revolution meant that many TNT readers were looking for a cheap high-efficiency speaker to match.
When chasing Fostex suppliers for a reader who had emailed asking if I knew a distributor, I stumbled across the http://www.commonsenseaudio.com/. Based in the US this outfit sold only high-efficiency drivers such as Lowthers and the aforementioned Fostex, and also sold them as completed speakers or as drivers with free plans. BUT they also had just introduced their own range of drivers that were considerably cheaper than the Fostex equivalent and it was claimed, superior. Well, as a way to make the TNT BFB even cheaper, and to make a new 'T'-speaker this seemed to be a great opportunity so I begged a pair of the 8" drivers from David Dicks and decided to get my circular saw out.
The AN drivers are very similar in construction to the Fostex 206. The chassis are almost identical, the cones and centre cap seemingly the same except that here they are black paper rather than the cool cream cocoa fibre of the 206's. The big difference is the surround which is a foam roll on the Fostex and concertina'd doped cloth the former somehow looks neater but the latter certainly has the edge in longevity if the number of drivers around with shot foam surrounds is anything to go by.
Round the back the other difference is even more obvious the AN has a much bigger magnet in fact I've never seen such a lump on the back of any driver of remotely this price.
Spec? well the two drivers are similar but the most important figures as far as this article is concerned are the QTS and the X-max as these define some of the characteristics of both drivers. Lowthers, the 206 and the AN have very, very low QTS's 0.18 - 0.24, in the case of the AN it is at the low end of this range 0.19. X-max is how far the driver moves and in this case the AN is more extreme than the 206, and well in Lowther territory with an X-max of +/-1mm.
So what does a very low QTS and a very small X-max mean? The latter is simple enough, it means that an 8" driver isn't going to shift much air if it can only move forward 1 mm at most. So low bass is a no-no. But then put it in a ported box? Nope because according to our old friends Thiele-Small, a low QTS driver should not work properly in a ported box, you'll end up with a bass hump at port resonance, then a big dip followed by a nasty rising output through the upper midrange - modelling confirmed this. With the BFB the 206 has an X-max of +/-1.5 mm and a QTS of .18, and that design came with a health warning about it's PA nature and the need to toe-in to roll off the rising treble the AN will increase the problem.
The advantages of this low X-max is that the coil is underhung and will always operate linearly in that short range i.e. very low distortion, and a lightweight cone that is powered over such a tiny distance by such a big motor is going to be both incredibly fast and detailed, but also hugely efficient AN claim 96 dbl and for once I think the claim if anything understated. For some people this sort of design is as good as dynamic drivers get.
So what do you do with a driver with such an extreme QTS and a low X-max? The Lowther is the granddaddy of such designs and was specifically made to work in horn systems, either front or back loaded never in a ported box.
So everything about the AN driver screams "horn!" and as such will be the cheapest, 8", quality driver available for such an application, though it's not an ideal cabinet for a beginner.
So what do AN supply with their driver? A set of plans for a ported box, and a lot of arguments about how used in this box, the driver will be far better than in a horn ditto Lowthers. So in one foul swoop AN throw out 70 years of accepted wisdom ho hum...
But who cares about excepted wisdom especially as a simple rectangular ported box is exactly the ticket when it comes to getting DIY virgins on the first rung of the ladder so I chose to follow David's instructions for his AN 2.8 speakers, a medium sized (2.8 cubic foot) floorstander with a big hole 6" across acting as a port. Every bit as simple as the BFB and of course the plans supplied made it even easier.
Designed (in my case) to work with a $30 amplifier I wanted to make this speaker as cheaply as possible money every object if you like, just to see how small a budget I could build them for even I was surprised.
For most people this
will involve choosing a panel material, MDF, Chipboard or best of all
plywood, all in 18 mm thickness. Then go to your local timber supplier
and you'll find that they will almost always have a massive, wall
mounted frame saw.
Take the plans, stand over the guy doing it, and get him (or her) to cut all the panels out for you. These saws will cut exact 90 degrees and to an accuracy of under a mm. So once cut, your DIY project is a simple case of cutting an 8" and a 6" hole and gluing the thing together. Now I know loads of people out there are saying nope I need to buy off-the-shelf", but you should be ashamed of yourself. Done this way this is easier than screwing an IKEA shelf together so GET A GRIP! To this you need to add a bit of fibreglass stuffing and of course buy banana socket and solder the wires up again don't wimp out on me here...
But as I said before I wanted to be cheap... A trip to a local DIY store found a load of pine shelving on "special offer". Not ideal I know, but it was good quality, would just need a coat of varnish to finish and when all's said and done it's a lot stiffer than chipboard even if it is a little inconsistent. This, the wood for my speakers, cost 15 euro... Add a euro for the varnish, glue and a few screws and I had some cabinets. At the same DIY store they had some sound insulation panels that had been damaged, 2 euro, and I lined the box with them. I had some long fibre wool lying about (I hate fibreglass) so that went in too. Obviously I couldn't afford binding posts so I just drilled a hole in the back of the speakers and ran CAT5 (cost 10 Euro for 10 metres) cable from the drivers, through the holes and to the amp coloured wires for +ve, white for ve. A dab of hot melt glue to fill the hole and voila! A cable solution better than ANY binding post (i.e. no joint at all). Lastly three rubber door-stops went under each driver (2 euro each) and I had a pair of complete speaker boxes for under 40 Euro.
This 'cheapskate' approach involved compromises, but some will improve the sound, some change it, however inevitably any DIY project such as this will not sound like the fully finished speakers produced by AN but they'll give some idea of their capabilities.
All the following was backed up using measurements (see below) which are reproduced, with explanation at the end of this article.
Drivers such as this take a lot of running in, sometimes up to years so to be frank at best we're listening to partly run-in speakers. On axis 1 m results were close to what you'd expect, a bass hump, a dip then that rising output through the midrange. Treble held up incredibly well, being well up even at 16 khz, this is a driver that is truly full range in the treble, easily keeping pace with most tweeters. To listen to, the sound was sadly what you'd expect, a very hard 'shouty' quality dominating again as you'd expect in this respect they were worse than the BFB with a 206, to the point that I found them unacceptable.
Why bother if they sound so bad? Well because in some ways they are very, very good... First: efficiency. I had a pair of Supravox 164's as comparison, also rated at 96 dbl and the AN's were a street ahead for the same volume setting 4 dbl in fact. Now assuming the Supravox efficiency rating is the usual fiction and thus boosted from a true figure of say 93 dbl, then the AN's on a ported box were pumping 97 dbl at 1 khz. This is an astonishing figure well up with a Lowther, the rising frequency response takes some of the credit but in the end it means disco volumes with a 5 watt T-amp...
The other thing was that the bass was not quite what I expected, it was deeper (especially 'in-room') and very fast and tuneful. Here I wonder whether the box was in fact acting more as a mass-loaded TQWP rather than a simple ported box.
But there's that rising mid/treble... The solution, at least partial, is to toe the speakers in. Look at the trace of the 30 degree off axis and that rising response is greatly reduced treble energy is directional in all speakers, but in full-range drivers with whizzer cones it is doubly so.
That's better! With the speaker driven now by the hugely expensive Son-of-Pharao EL34 valve amp the result was much more to my liking this in my 'b' system set up in my conservatory. Vinyl sounded pretty good, but my dear old Onkyo CD player had it's mid range laid open cruelly and again wasn't really listenable. The T-amp also worked brilliantly, it's a bit harder than the Son-of-Pharao, but that fabulous soundstage width was well to the fore.
This was gratifying, the main motivation for the review (and I know other writers for TNT are also looking) is to find a cheap speaker solution worthy of the incredible capabilities of the T amp.
But as already mentioned the result was far from evenly balanced. The solution is thankfully very simple. Designs such as this can be made to work with a 'notch' filter composing of an inductor and a resistor, parallel both and put the result in-line with the positive terminal of the speaker.
At low frequency the inductor offers no resistance to the signal. As frequency rises the coil begins to act as a resistor at a certain point, the larger the inductor the lower this frequency is. This resistance then parallels with the fixed resistor so that the total resistance is frequency dependent and dependent on the size of the fixed resistance. So the larger the fixed resistor the deeper the notch, or reduction at higher frequency.
The only way to do this is to experiment and by using simple screw terminal blocks you can put various 'notches' together using cheap resistors and a couple of cheapo inductors. Once you've hit on the ideal then you can break the bank and buy high quality components, in my case the 10 ohm hi-power resistor and 4.0 mH inductor cost 30e from IPL Acoustics.
This took the 'sting' out of the sound and added a little warmth, but read on...
During all this fiddling about I became more and more convinced that here we had one solution to the T-speaker problem. So now we get serious and pull the boxes into the main listening room.
This room is totally different to the conservatory it is very 'dead', walking in is almost like walking outside. There are a few resonances but they are pretty low down (primarily 40, 90 and 200 hz). The main system is £15,000 of high-end amps and turntable into speakers costing £4000 (Loth-x Polaris).
Against them I ran the turntable via the Gramamp Solo SE into the new 'Super' T-amp (smoother and more bass than the original) and then into the AN's. The turntable was obviously the same but the rest of the system cost around £200 :-)
And the result for the AN's? In some ways they were high-end speakers. What a thing to say but it's true. The sense of atmosphere and ambience simply stunning. Detail too was of a very high standard. How good? Well in all these areas they bettered my own Loth-x Polaris, specialist full-range horns. This though comes with a downside. That talent was apparent at moderate listening levels, wind up the wick and the lack of travel on the driver began to roll off the bass and further exposed the midrange.
The other problem was that so revealing were the drivers that the presence of the notch filter was not entirely benign. As well as throwing away 3 dbl of output the 'notch' put a veil between you and the recording. In fact much to my surprise I found that I preferred the sound without the notch, with the speakers hard in the corners to re-enforce the bass and toed in 10 degrees to roll off the treble/upper mid. Despite the still mid-forward sound, at moderate levels this sounded pretty special. True, with the wrong material it could be wearing but if you want a taste of the high-end for £150 then a T-Amp and the AN 2.8's will certainly fit the bill, just be prepared to put some time into placing, cabling, running in and maybe notch filters.
But... All this would equally apply if you had just splashed out £500 on a pair of Lowther drivers. The Audio Nirvanas really are Lowthers on the cheap. What that means is that these drivers, under £100 remember, will fit into any speaker designed for Lowthers. Just do a search on the internet and you'll be swamped by plans and advice for building Lowther horns, quarter-wave cabinets and Voight pipes. They are even perfect drivers for a go at Thorsten Loesch's 'Magnificat' project. The point is that once you've bought the drivers, built one of David's ported cabinets and seen what they are capable of you may agree with David and think them the best possible cabinet. Or you may lust to try them in a big full range horn etc. And if after making such horns and falling in love with them you can easily upgrade to Lowther drivers, though I have to say that so good are the AN drivers you may not find the improvement quite what you expected.
And lastly go look on Ebay and you'll frequently see people selling off old Lowther cabinets, often Acousta's, with no drivers. A decent pair can be as little as £50 so just drop in those AN driver and you have a pair of professionally-finished horn loudspeakers. The fact is that together with the T-amp a whole world of playing and learning is opened up for you and for the cost of a budget bookshelf speaker.
So do I recommend the AN's drivers you bet! But see them as the beginning of a journey, not necessarily a destination. Part two of this review (to follow when I get them into some horns) will document my journey with what I consider truly remarkable drivers for the money.
You'll see three traces, all taken at 1m at the height of the driver centre. The Green trace is that of my old IPL S3MTL's. These are pretty close to being "full range" demonstrated by that powerful bass at 30 Hz. The glitch at 40Hz is partly due to port suck-out a problem with most transmission line speakers, but one which is largely filled in in-room, otherwise they are a very good example of a speaker that measures "flat" anechoically. The AN's don't even attempt to go there but join the party at around 60 Hz. In room this can be boosted by wall, or even corner reinforcement, but it's asking a lot for the driver to really shift a lot of air that low down. Follow the red trace which is the AN's on-axis and you'll see the predicted hole in the upper bass and then the steadily rising mid and treble output. Though it looks good on paper to see that high treble output it is in fact misleading. In-room the trace should roll-off as with the IPL, as mid and bass is reinforced by reflections where treble is not. The result is a very bright pushy sound with no body. Now look at the blue trace. Here the speaker is measured at 1m but with the speaker turned 30 degrees off-axis. Just look how that rising treble is controlled to give a lovely roll-off just like the IPL. However the big hole in the upper bass is still there robbing the sound of body and warmth.
Here the red trace is the same as the blue trace in Fig 1, i.e. the best so far. The other two traces are with the notch filter in place reducing upper-mid and treble output and therefore raising the lower frequencies in comparison. With the gentle notch I used you can see how in both cases the hole in the upper bass is filled to a certain extent. Bigger notches would have a bigger effect, but perhaps not such a benign one. The Blue trace which is notch + 30 degree off-axis being the pick of the bunch.
Here we see the best result of all. This is the trace of the AN's, in-room against my own Loth-x Polaris. The room is dead, the notch was removed, and the measured result is pretty good compared to the Polaris. The mid peak is still high compared to the bass, but it's bearable and playing with position would no doubt improve it. The astonishing thing is how smooth that treble rolls off (at 10 degrees off-axis) and how high it goes - note that it's better than the Polaris... It's also important to note that all these measurements are taken at pretty high levels - I'd guess 90 dbl+ and so reflect the response at realistic listening levels rather than the very low levels some measure at.
These three graphs show just a few of the major changes you can make in response with very little effort or cost. But beyond that there's a mass of changes that could be done vis a vis the cabinet; materials, stuffing, ports etc and another world of different cabinet designs.
Generally, I feel it's quite fair and thank you for the fine recommendation. Of course, I have a few comments:
1. Lowther insists that their drivers work equally well in bass reflex OR bass horn cabinets and I agree. I think that Mr. Thiele and Mr. Small should be taken out and tarred and feathered. They're responsible for more, bad speaker engineering than everyone else put together. There is no magic to Qts, in my humble opinion. I have heard many, many speakers with excellent bass (in bass reflex cabinets) that the 'experts' say shouldn't work. I can't tell you why they work, they just do. My guess is that the same geniuses who accept the Qts nonsense don't know how to properly port a bass reflex cabinet. You'll note that our tuning flies in the face of conventional wisdom here.
2. Lowthers have plus or minus .75 mm excursion, Audio Nirvanas plus or minus 1.0 mm (total movement is 2.0 mm, not 1.0, right?), and Fostex have plus or minus 1.5 mm. In my opinion, and the folks at Lowther, the 206' 3.0 mm movement is too much to work properly in most bass horn cabinets. And we purposely chose plus or minus 1.0 mm so that the Audio Nirvanas would work equally well in either type of cabinet......and make a little more bass than the Lowther.
3. 'Hard, shouty?!' Well, as you say, the room was perhaps too live and I'm not a big fan of CAT5 cable or the T amp. Yes, the T is a good value for the money, but I much prefer a 'normal' amp.
4. Lowthers and the PAudio coaxials we sell are still significantly better. But the Audio Nirvana's are the 'best I ever heard' FOR THE MONEY. And I'm glad that you agreed they actually sounded better without the notch filter. It is, of course, possible to 'tone them down' in other ways, like using different damping material. However, the best way is the cabinet material you use. Baltic Birch ply will give the softest sound, industrial grade particle board--intermediate, and MDF will give a harder sound. Solid hardwoods (not recommended because of the inconsistency of the wood--knots, etc--will also usually give a harder sound). But I think the review is generally very fair.
You may want to try the Super 8's eventually. They're even closer to Lowthers. And we're working on something new that I hope will work out and be even better. Regards, David
© Copyright 2006 Geoff Husband - www.tnt-audio.com