Audio Note AX-Two stand-mounted loudspeakers

Audio Note AX-Two stand-mounted loudspeaker

I have to confess that initially I was impressed neither by their look nor by their sound. It was a year ago when I climbed the stairs up to my apartment carrying the 15 kg heavy box that Andy Whittle handed me after a short stop at my place on his way to Copenhagen. “These are your new speakers, give them some time to properly break in,” were his words.

At the time they arrived I was playing with the Micro Block DIY speakers and I really thought Whittle’s creations will have hard time against the Timmermanns’. And my expectations were confirmed on the first listen. They were disappointment and didn’t come nowhere near Micro Block’s dynamics, bass extension and sense of gravity. I didn’t want to waste my time with them but to continue to experiment with stuffing, port and crossover adjustments of Micro Block. So after a short listening I put them back in their box, sealed it off with a tape and called it a day. I realized that I don’t want to spend dozens of hours in front of these unimpressive speakers expecting them to improve with time when I have such an interesting project going. So I wrote to Peter that I am done with them and actually don’t like them:

I gave them AX Two a try yesterday, but got rather unsatisfactory results, they sound pretty unspectacular and cheap, rather big step backwards in comparison to my current speakers. They might improve after some more hours of use but I doubt that they will ever reach near the level of performance of Micro Block MK2.

But however impressive Micro Block sounded I eventually grew dissatisfied with them, there was a sort of glare in the midrange region I was not able to tame. Another thing that bothered me was the inconsistency in the detail retrieval across the frequency spectrum, the bass was lacking resolution and sounded sort of a one-note, they were simply putting some sort of signature upon the music that was drawing my attention constantly. And I confess I made few mistakes by deviating from the original design proposed by Bernd Timmermanns. The first one being the choice of enclosure material in favor of plywood instead of MDF. My advice to anyone who decides to build Micro Block is to keep to the origin and use MDF. It might sound counter-intuitive but in certain cases MDF or even chipboard is preferable to plywood. I had a conversation on that topic with the loudspeaker designer David Haigner and his opinion is that especially because of the inconsistency of the wood quality, plywood as an ever changing target with strong and narrow resonances is harder to “tune” than the higher damping MDF. He also considers that the popularity of plywood within the DIY circles is due to the fact that it presents a finished surface requiring little effort to look good. So I am inclined to think that the aforementioned coloration could also be contributed to the choice of enclosure material. This sort of glare effect became even more pronounced after I decided to glue together the internal braces. So my advice is – try not to be smarter than the designer but keep to the original concept.

Fast forward – few months later

Eventually I reached a state of burn-out when I was listening no more to music but to defects of the reproduction and that was the point I had no other choice but to give the AX-Two a second chance. And this time it worked out. I guess I should have gone through this struggle in order to appreciate the innate musicality of AX-Two. In fact they felt like a breeze, stepping out of the way, not drawing attention to themselves they simply allowed me to appreciate music. And I had no wish to fiddle with them at all. Not that they were perfect, but they simply let the music be, without putting a stamp on it. I felt sorry for my rash conclusions and I couldn’t believe I’ve judged them so badly at the beginning. Although they still lacked the sense of weight and sheer slam of Micro Block they sounded extremely coherent, almost like a single driver speaker. I also noticed the very well defined soundstage projected in a natural perspective way behind the speakers. I learned a lesson this time, an important one which Peter summarized with the following words:

Listening is a learning process, often what impresses initially is not right, so no need for excuses, I am happy you got to the right conclusion eventually, your journey should be written down in your blog as well, as it us a useful reminder to readers that first impressions are not always that valid against a longer term.

Peter Qvortrup

With careful set-up and more run-in time behind their back AX-Two inevitably started also to impress. What I discovered and would like to share is that a key role of how extended a stand-mounted speaker will sound plays the stand itself. I use a heavy steel stands – Audio Note AN-K – that are not a welded design but assembled. Top and bottom plates of quite thick steel are attached to the hollow legs with bolts. The alloy from which the legs are made is quite inert and result of extensive experimentation. I got caught by surprise when I knocked them with a fingernail and there was no ringing whatsoever but a rather dull sound one usually does not associate with metal. I use them empty, coupled with the delivered spike sets to the floor and to the speakers respectively. I tried also Blue Tack and other putty material between the stands and the speakers but nothing came close to the dedicated spikes provided with the set in terms of bass response extension, clarity and natural build-up of harmonic overtones. The role of the spikes is to allow the stands to couple from one side to the floor, from the other to the speaker in order to serve as a mediator and channel the energy transfer between them. Using Blue Tack seems to impede this energy transfer by effectively cutting the speaker off the floor which to my opinion affects negatively the sense of realism. This effect is especially well observed when natural recordings with plenty of spatial cues are reproduced. But at the end, I think the way we set our speakers is part of the creative process most of us enjoy and see as an opportunity to express oneself and one’s understanding of what a good audio reproduction actually is. So I don’t want to claim this or that way is better but to encourage you to experiment in case you still feel there is a potential for improvement.

After a careful set-up what impresses the most is how transparent and open they sound while maintaining considerable bass extension and bloom – it is the transparency one usually gets with open baffle speakers combined with the full low-frequency extension of a ported box. It is really the key here – this sort of well balanced mixture of the best of both worlds, and I couldn’t stop wondering how is that possible. On the one hand it is the woofer itself – the venerable 6.5″ Vifa C17 featuring curvilinear paper cone, stamped steel basket and foam surround. This woofer has manifested itself in a range of commercially available loudspeakers during the 1980’s and 1990’s but most notably in the highly regarded Dali 104 loudspeakers. In fact these very same speakers were predecessors of the Living Voice Auditorium series of floor-standers started back in the early 1990’s. There is no wonder that this peculiarly looking woofer with its stamped metal chassis and short wave-guide extrusion is found also in Audio Note’s AX-Two stand-mounted loudspeakers. On the other hand it is that particular implementation I find very interesting and deserving a closer look.

Vifa C17WG-01 speaker driver

Trying to figure out what is the secret here I decided to check if the boxes are acoustically sealed, so I blew through the port and noticed that the air was escaping all around the woofer’s flange. And as it turned out there was no gasket between the woofer and the baffle. So the enclosure was “leaky”. But according to the theory in order to function properly a ported enclosure should be air tight. I realized that AX-Two behaves partly as bass-reflex, partly as aperiodic box and this should have been the main reason for the elegant amalgam of qualities it exhibits. Out of curiosity and to confirm my notion I decided to check how the speaker will behave if it is properly sealed. So I used the self adhesive foam gasket tape I had at hand to seal the woofers off and as you might guess – the magic was gone. Deprived of its aperiodic behavior this speaker sounded, let me say, quite typical. For me it was clear – we are talking here, as with everything Audio Note, for a rather unorthodox and well considered design that obviously leaves no place for dogmas. By carefully combining two distinct approaches to speaker design Audio Note has achieved what I perceive as an extremely transparent, dynamic and extended sound. The “leaky” enclosure seems to provide the Vifa C17 with ideal conditions to perform to its best, and on the other hand reduces the impedance peak at resonant frequency which makes it more amplifier-friendly.


The enclosure itself is made of 18 mm thick MDF finished in wood-grain vinyl and use a single sheet of 2 cm thick low density polyester felt covering the back and the sides. Sides, top and bottom panels are rabbeted so that they wrap around the front and the back panels with half of their thickness. There is no bracing of any sort. With a net volume of 11.5 liters and reflex-loading port dimensions of 3 cm in diameter and 10 cm length the enclosure is tuned at around 40 Hz resonant frequency which lays quite below the low frequency cut-off of the speaker which starts at 55 Hz. This is pretty unusual considering the fact that the vast majority of compact bass-reflex loudspeakers including Micro Block have their ported enclosures tuned at or slightly above the cut-off frequency. The fact is that AX-Two, as well as any other Audio Note speaker, is designed to work near room boundaries, thus the additional bass reinforcement from the nearby walls is part of the calculation. I can attest that this design concept works. Placed 15 cm from the wall, 2.40 meters apart these speakers produce an extremely clean and tuneful bass that makes Micro Block’s bass performance to sound kind of a one-note and lacking in resolution. True, the enclosure’s net volume and the woofer’s effective radiating area are other factors one have to consider, and they are all in favor of AX-Two, but here I talk about some substantial difference in quality that I think reflects the design principles behind.

Speaking about the bass performance I should not forget to mention the twitter – the 3/4″ silk dome Vifa D19 – also a long time favorite with long production history. I can’t help but think that this kind of perfect coherency I experience has something to do with the integration of both units within the crossover design. Throwing an eye on the circuit board I can see a neatly conceived dividing network featuring a second order 12 dB/Octave low-pass filter and a third order 18 dB/Octave high-pass one. Pretty straightforward design, no Zobel network or anything of that sort, 6 elements in total (in fact 7, due to space limitation two parallel connected capacitors are used instead of one). What impresses me most is the placement of the attenuation resistor in the high-pass section which is at the very start before the filter group. I have never seen Bernd doing this, in all of his projects the resistor or the voltage divider is always the last element placed after the filter group and just before the twitter. Judging by how seamless the integration feels I think placing the resistor at the very beginning is aimed at achieving a maximum phase linearity. I am impressed.

Audio Note AX-Two crossover board

Interestingly enough the Micro Block dividing network uses the same constellation of second order low-pass and third order high-pass filters but while Micro Block uses 12 elements to achieve that (without counting the elements in the recommended impedance linearization network) AX-Two as I said needs only 6. Where Micro Block resorts to Zobel network to electrically address phase angle anomalies AX-Two resorts to mechanical means by judicious mix of aperiodic and bass-reflex driver loading. Chasing simplicity by keeping the parts’ count low as the only way to achieve perfection is reflected in the whole philosophy of the British company and it could be traced across its entire product range. It is something that gives the brand consistency and characteristic signature. As a user I can attest that keep it simple but not simpler principle works here, AX-Two wins against Micro Block in all disciplines that are essential to communicate the emotional impact, the artistic message and aesthetic value of a record.

My amazement continues to grow by the fact that even though the cheapest electrolytic capacitors are used, ferrite core inductor and a ceramic resistor, despite all these cost-cutting measures these speakers exhibit such coherency and dynamic prowess. I can’t help thinking – what if the crossover boards are out of the enclosures where the constantly changing pressure, relentlessly pounding airwaves and mechanical vibrations are taking their lot, what if parts of higher quality are used, foil instead of electrolytic capacitors, air core instead of ferrite core inductor and metal-oxide instead of ceramic resistor, what if … I simply cannot stop myself, I know it is not reviewers job poking his nose, and especially modifying a component sent for review is a heresy, it is something one should never do, but I can’t stop myself, it’s that passion that burns and tells me these speakers have an enormous potential that should be released. I ought to give them what they deserve and in return they might allow me to have a glimpse of heaven.

Bringing the crossovers out

As I already said I couldn’t resist the temptation to try the sonic outcome of just one simple move – bringing the existing crossover boards out of the enclosures. Generally speaking putting the crossover board within the enclosure is a serious compromise dictated by production costs and convenience considerations. For that experiment I decided to use the same Kimber 4PR bi-wired set of cables I already used for a week. What I did was to unsolder the boards from the binding posts and the speaker driver’s cables and then solder the speaker driver’s cables directly to the binding posts. Then I soldered the Kimber cables to the crossover boards at the respective points (the filtered out signals for the twitter and woofer) and replaced the original short pieces of cable that used to connect the crossover’s inputs with the binding posts while maintaining the bi-wired configuration and taking care of the correct polarity. The only difference now was that the crossover boards were out of the enclosures, closer to the amplifier and further away from the speakers.

Audio Note AX-Two crossover board

I remember the first listen with the outboard crossovers as if it was yesterday. It was Chet Baker & Paul Bley collaboration Diane (LP, SteepleChase SCS 1207). I was simply shocked by the effect of this relatively simple tweak. The tone of the Baker’s trumpet was shimmering in the air with real brassy sheen while the deep bass notes produced by Bley’s piano, rich and dense with harmonics, were unfolding in the space around, an electrifying, goose bumps experience. I instantly noticed the gain in clarity and immediacy, it was simply scary. I couldn’t stop to listen before the needle hit the lead-out groove, it was mesmerizing. I got so excited that few hours later I was writing an email to my pal Oscar telling him about the outcome of the experiment:

I just want to tell you that I brought the crossovers out of the enclosures, this was the only change I made, using the same Kimber double run cable I was using lately. I am speechless, these speakers are performing now on a level I didn’t think it’s possible, amazing stuff. It turns out something I suspected a long time ago is true – bringing the crossovers out gains a massive improvement, bigger than simply changing capacitors or wires.

Getting rid of the printed boards

Once you get on that road it is hard to stop, the gain in purity and immediacy is addictive, it brings you closer to the artist’s intentions. The interaction between both musicians became so palpable and transparent that one can get sense of every hesitation they encounter, every uncertainty, of every shade of expression. So after a few weeks of enjoyment I decided to move the crossover elements out of the original boards and by using point-to-point soldering to eliminate the printed circuit paths. I kept the original coils and one of the capacitors, but the rest I replaced with cheap film ones I had at hand while the ceramic resistor was replaced with a metal-oxide equivalent. I was anticipating some improvement to come but this time the effect was even bigger than expected. The music seemed to flow unbridled as if liberated, it was like a dam on the river has burst open and the stream was rushing with all its energy, the last obstacle was out of the way. Believe it or not but I had to reduce the loudness which now I perceived increased. I can’t tell why this happened and how exactly the dispense of the printed board affects the perceived loudness but I have the feeling that the speakers now are slightly more efficient.

DIY outboard crossover for Audio Note AX-Two loudspeaker

Honestly the AX-Two outboard crossover project exceeded all my expectations. I enjoy these crossovers for a few weeks now and they give me a glimpse of what AX-Two are really capable of. I am definitely going to continue this experiment and plan as a next step to replace the ferrite core coil and get rid of the last electrolytic capacitor. Afterwards I may try some better capacitors and cabling, finalize the layout and replace the provisional boards with dedicated ones.

Listening impressions

Balanced is the most accurate adjective I can come up with trying to describe Audio Note AX-Two. These speakers do not favor any particular genre but do equally well with acoustic or electronic music, vocal or instrumental, natural or studio recordings. Listening to Jan Garbarek’s classic I took up the runes (LP, ECM 1419) I am taken on a journey into fantastic landscapes created by Bugge Wesseltoft’s synthesizer, peaks and valleys traversed by Eberhard Weber’s unexpected and bold bass lines, rays of sunlight peeking through the clouds radiated out of Garbarek’s saxophone. The sound is spacious and densely saturated, very solid and liquid at the same time.

To check how AX-Two are dealing with human vocal range I give the Kenny Wheeler’s suite Mirrors (LP, Gearbox GB1513) a spin. Featuring Norma Winstone backed by the London Vocal Project this is a contemporary record conducted in the best analog traditions using Studer tape machines, vintage Haeco Scully cutting lathe with Westrex heads and all Audio Note monitoring equipment. Norma’s voice, soloing or backed by the 24 voice chorus, is as spooky present as I ever heard from a record. There is an unmistakable warmth radiated from a living being that only the best records played under the best conditions could deliver. I am enthralled, it is so charming and comforting to listen to this record, and it fully shows what a properly made vinyl has to offer.

Audio Note audio system with DIY outboard crossovers


AX-Two are exceptional speakers. Especially with the outboard crossovers they deliver on par with some of the best stand-mounted transducers I have ever encountered at a fraction of their price. Their relatively high sensitivity of 90 dB (2.83 V, 1 m), benign phase angle characteristics and 6 Ohms impedance make them perfect partner for low-power tube amplifiers. Featuring pair of posts per speaker driver the terminal plate allows for bi-wiring which I find beneficial. When bi-wired these speakers maintain the seamless integration of both driver units while bringing up some additional clarity and composure to the presentation. Power handling is pretty good and there is plenty of headroom before they start to compress dynamics which in fact I have never experienced. They are capable of playing very loud while keeping the whole picture intact. They keep the balance also at very low listening levels which make them perfect companion for the late hours when the neighbors are asleep. All in all an excellent speaker that has become my reference.

Thank you for stopping by and
Happy New Year!

2 thoughts on “Audio Note AX-Two stand-mounted loudspeakers”

  1. Hi
    I am listening to my AX Two as i read your blog. Fantastic results you got there! Have you changed the remaining caps and which ones do you use? I must try that mod! Thanks for a very interesting read!😃

    1. Hi Fredrik,

      Thanks, I am glad you liked the read and got inspired!

      You should definitely try the outboard crossovers for they will give you an idea of what AX-Two are really capable of. Better keep the original crossover boards intact and build new ones from scratch. This way you will be able to restore the speakers to the original state if you decide to sell them later, which I really doubt. The parts used in the original crossovers are of basic quality and it doesn’t make sense to reuse them. Here is the schematic of the original network you can base your new build on. You can start with any budget foil capacitors first and later on switch to some more expensive ones. I usually use MKP foil capacitors from Audaphon during the prototyping stage, but for the final build I would go for Mundorf MCap Supreme. Two of the capacitors used in the original boards have values of 6µF and 12µF that are not standard, but you can use 5.6µF and 10µF instead. This will shift the tonal balance slightly, but later on you can always add some small caps in parallel to achieve the precise values. The 1.33mH coil in the low pass-filter has also non-standard value, but you can use 1.5mH one (any air core coil with wire diameter between 1.2mm and 1.4mm would do) and unwind few turns until it reaches an inductance value of around 1.35mH. For this you will need an LCR meter that doesn’t need to be expensive. Another alternative is to use two standard 0.68mH coils connected in series (something like this), which will bring the total inductance to 1.36mH and sounds pretty good. The 0.33mH coil in the high-pass filter is made out of 0.7mm wire and has a standard value. Hope this helps.


Leave a Reply

Your email address will not be published. Required fields are marked *