Fullrange Speakers

Horn Speakers

DIY Projects
  DIY Introduction
  Open Baffle :
  Diatone Big Blue
  Box Speakers :
  Magnificat Axiom 80
  Norelco 9762Ms
  Goodmans Axiette 8
  Aperiodic Boxes
  MCM Vented Box
  The Holy Grail
  MCM Revised
  Return of the Holosonic
  Radio Shack 1354
  Afterburner Project
  ACR / Fostex Project
  The 66% Solution
  The Japanese Experiment
  Pioneer with Piezo
  Deep Green
  Definition Gold 900 Tannoy
  Fostex 168S / BS-168
  TQWT / T-line :
  Fostex TQWP
  Hammer Dynamics TQWP
  Radio Shack TQWP
  Adventure with Voigt Pipes
  The Rat Pipe
  Eeehaah's Folded Voigt Pipes
  TQWT - Bigger is Better
  Jordan JX-92S TQWP
  Straight ML TQWT
  Fostex FE207 Voigt Pipe
  The Zigmahornet
  Rear Horn :
  Eminence Back Horn
  Fidelio with Fostex 208
  Fostex 208 Backhorn
  Fostex 108ES Side Horn
  Cheap Trick 164
  The Moore Horn
  The Less-Compromised Backhorn
  Nagaoka Swan D-101S
  Big Mouth Lowther horn
  Fostex FE 103 Memorial Backhorn
  Other :
  Laminate Horn Enclosures
  DIY Room Diffusers

Other Stuff


Cheap Trick 164

by Stefan Klose

A fullrange backloaded horn of living-room and wife-acceptable-size.

Cheap Trick 164 is a hyperbolic backloaded horn designed by Bernd Timmermanns for Fostex 10 cm fullrange driver FE103sigma. It was published some years ago in the German speaker building magazine Klang&Ton #5/96. Originally it uses two passive equalizers to flatten frequency response, but by some modifications - mainly a different driver and increased prechamber - I could leave them completely away.

I don't know if it is for the not so promising name ("CheapTrick" may not necessarily indicate "high end") or simply because most people will not believe such a small 10 cm driver can go fullrange, that means also being able to cover the bass range really satisfactorily; but it never became as famous as it's larger brother, the Jericho Horn (btw also designed by Bernd Timmermanns) using Fostex FE208sigma, which was published just two numbers earlier than CT 164 in K&T. I never compared these two horns myself, but those who did, referred CT 164 to have deeper and stronger bass than the Jericho Horn! CT 164 does not need to be placed in a corner, a wall in about 50 cm distance in the back is enough to bring the bass of this 280 cm long horn down to about 50 Hz.

The name Cheap Tick 164 comes from a series in K&T which was intended to show good speaker concepts on a budget; the different Cheap Tricks were simply enumerated through all the years.

Issue number 5/96 was actually the second number of K&T that I had bought. Through a longer period of time I had become aware that I was not satisfied by my current system's sound anymore, I was not involved in the music. But it was only a vague feeling, I could not tell what exactly was missing. So I started searching for information by reading K&T because I was sure, exchanging speakers would make the greatest improvement.

Even though I found CT 164 interesting immediately (I love "exotic" solutions), I could not decide to build it for several reasons: Too large, too cheap to be really good (FE103sigma costs only 80,- DM =40$) and I had already made some not completely convincing experience with another fullrange backloaded horn. My absolutely first pair of speakers had been Lowther Bicor 200 with PM6c/PM6MkII back in 1982. I bought them for their bright and detailed sound; but after a couple of years I went into another DIY speaker shop and had a listen to a 3-way transmission-line using Focal speakers. I discovered that I could hear tones so deep I did not even know they were on the record with my Lowthers. The Bicors definitely had no bass below 100 Hz or even higher! I took some time and listened to a lot of other speakers and after another year these Focal transmission-lines replaced the Lowthers. These again were replaced some years later by a small two-way combination with GIA speakers.

During this "speaker-rolling" expanding over a period of about 10 years I always had the feeling of getting better sound with every step. Clarity and neutrality were the most important factors to me. Strange, but I did not recognize a decrease in soundstage or articulateness with the newer (more-way) speakers, which should have been the strength of the Lowther Bicor 200, as I know now.

Starting from issue number 5/96 I had to buy and read a great number of Klang&Ton's through the last years to finally come back to CT 164. One day I happened to hear Quad ESL 63 and this way I became aware, that deep, wide and precise soundstage and great articulateness were the things I recently was missing in my system. From what I had read in K&T I was convinced, that the easiest way to achieve this for me would be fullrange drivers; the Quads themselves were not within my budget and they are not very comfortable when you have two little children at home. Still I was afraid of lacking bass - Bicor 200 and CT 164 are about the same enclosure-size!

Then I had the chance to listen to CT 164 (thanks Thomas Nowoczin!). I compared them to my GIA's and was blown away! There was plenty of bass (no compare to Bicor 200!) and soundstage and clarity were a new dimension to my ears! No question, I wanted them.

But now there was still its size: CT164 is small compared to most other horns, but still it is large for most wives - at least I thought my wife would think so.

I had just discovered James' wonderful site including the Fullrange Driver Forum and started to search for smaller alternatives using Fostex 10 cm driver. I came across the links to pictures of Tetsuo Nagaoka's D-101s (they have disappeared now) and was amazed - I found it much more living-room-suitable. I didn't know anything else about it and started to search the web for information. Finally I got in contact with a nice Japanese guy who helped me with lots of information about Nagaoka's designs in general, D-101s in particular, and differences between 10 cm fullrange drivers from Fostex (thanks, Nobu!). And this is an important point: up to then I did not know, that Fostex has some highend version speakers of their fullrange-drivers, which are built in limited numbers and only sold in Japan. This is what I found out about the 10 cm fullrange line from Fostex (there are similar special versions for drivers of other sizes):

The grandfather of the whole family is the FE103. With a double sized magnet it was introduced as FE103sigma. About 1980 Fostex replaced the FE103sigma with a new version completely different from the original: larger cone, cone surrounding made from rubber instead of coated cloth, different magnet AND different sound. This version was disliked by the Japanese DIY scene. Nevertheless this FE103sigma was for years the only Fostex 10 cm driver available in Europe traded by ACR and is just discontinued now. In Japan Fostex introduced due to public demand the FE106sigma, which was more or less the OLD FE103sigma but with a more solid aluminium frame. Tetsuo Nagaoka designed his "Swan D-101" for this driver. The recent version of this driver has a round frame and is sold now worldwide as FE108sigma. In the late 80's - early 90's, Fostex handbuilt a limited version called FE108super with much stronger magnet. The same driver with a voicecoil made of 99.9999% pure copper was sold as 6N-FE108super. For these drivers Mr. Nagaoka modified his "Swan D-101" to "Superswan D-101s" considering their shorter amount of bass because of lower Qts/stronger magnet.

Years later Fostex exchanged the paper cone of this driver for a non-wood based cone and called it 6N-FE108ES (picture here). The recent version is called 6N-FE108ESII and has again a stronger magnet, a different cone and a different cone surround (picture here). In Japan the "super" and "ES" versions are said to be much superior to the sigma version especially considering their ability to resolve details.

To show you how strong and fast these drivers are, let me compare the Fostex 6N-FE108super to a Lowther PM4A — the strongest driver in the Lowther-line: Flux density of a PM4A is 2.4 Tesla and a moving a mass of about 12g. Flux density of the Fostex 6N-FE108super is 1.48 Tesla moving a mass of 2.7 g! If you set these values in relation, you'll see… (2.4T/12g = 0.2T/g compared to 1.48T/2.7g = 0.55T/g) OK, I know magnet strength does not say everything; the sonic result of course depends also on the driver geometry, which determines how effective the magnetic force is transferred to the voicecoil. But I suspect the Fostex engineers will have regarded this detail too.

Sadly these drivers are hard to get outside from Japan. (Maybe worldwide public demand could change this…)

You can imagine how my heart made a jump, when my friend Nobuyuki Nishiyama mailed me he could sell me a nearly unused pair of 6N-FE108super he had left from some project and how happy I was, when I received that little package from Japan !

Having these drivers at home my dear shouted out: "What? You really want to build these ugly speakers? They look like two submarines with periscopes." She was talking about my beloved D-101s. I said "Yes" and if I were John Wayne that would have been the end of the discussion. But I am not, I am a loving man and so I was brought back on earth. I built a model of D-101s made from some old cardboard boxes and have to admit, she was right: it really spoiled our living room ! So I dragged out my old number of Klang&Ton and showed her the pictures CT 164: "What do you think about this one?" "At least it looks much more elegant !"

Well, an enclosure designed for FE103sigma with it's frequency response needing to be flattened by two passive equalizers, one for upper bass and one for midrange, using this enclosure with a 6N-FE108super — would that work? Would it be worthwhile to build these complicated enclosures to find out it does not work? Gladly I remembered some years ago, some time after CT 164 was published in K&T, I had seen the raw enclosures for sale in a DIY speaker shop here in Cologne. They had done some experiments with their ceramic drivers but failed. I called them and, yes, they still had them hidden away somewhere. They were so kind to lend them out, so I could try them first with my Fostex drivers. I screwed the drivers on, connected the speakers to the second pair of speaker terminals of my amplifier and turned on the music. Switching forth and back showed how much bigger and precise the soundstage became with the Fostex horns and how much more detail there was. It sounded so light and easy. But there was a problem with boominess. Nevertheless I thought this would be solvable and bought the enclosures. 450,- DM for a pair of horn-enclosures made by a carpenter, no woodwork for me!

But now removing the boominess! It was audible only with a few records for example deep female voices such as Cassandra Wilson or Cesaria Evora. The most painful piece of music was track #1 on Air's "Moon Safari": "La femme d'argent" - I liked this very much before but couldn't stand listening to it with CT 164 anymore!

First I tried it with different amounts of damping wool in the horn mouth. With the horn mouth completely filled, the boominess was lowered strongly, but that lowered bass output strongly also. Two layers of that wool on the bottom plate seemed to be a good compromise but still it sounded boomy. Moving the enclosures around showed similar results: Far away from backside wall cleaner but leaner, nearer to backside wall stronger bass but boomy. Filling the enclosure's large central cavity with sand showed some effect, but not much. Playing around with the tone controls of my amplifier resulted in a setting, that made the sound acceptable to me: Pulling down frequency response @ 400 Hz and pushing it up again @ 60 Hz. But that of course was not the way it was intended to work - I want to keep the signal pass as clean as possible: Better no tone controls at all! Still I had the passive network as possibility but for the same reason I disliked this idea too.

Another possibility was to increase prechamber volume to lower the cut-off frequency of the bandpass and thus lowering the amount of midrange getting into the horn. So finally I still had to do some woodwork: I opened up the enclosure's front plate completely over the whole area of the prechamber. Then I glued a second plate of 24 mm birch plywood — also having a whole covering the complete area of the prechamber — on the original from the plate. Finally I attached another plate made from 18 mm birch plywood which now takes the driver. That way the originally only 1.1 liter prechamber grew to a volume of about 1.8 liter, comparable to the size Mr. Nagaoka uses for these drivers in his enclosures. The edges of the additional plates were angled with 9° which refers to the angled side plates of the enclosure's bottom segment. This way the front plate's width was reduced to 130 mm, just enough to hold the driver.

While letting the glue dry, I made some new speaker cables (I tell this here because it had some effect on the boominess, as you'll see). I needed longer cables than before because of different speaker placement (2 x 11 m - I know this is not an audiophile length, but no other way to do it in my flat). Even though I never believed in cable voodoo, I chose to build a simplified version of Thorsten Loesch's UBYTE speaker-cable (original article), made from low-loss satellite cable, since it is really cheap, at least not more than the standard stranded copper cable you usually get in the mass markets. So there was not much risk to waste money and a chance to get a better cable. My previous cables had been such stranded copper of 4 mm^2 and they were 8 m long. In the same turn I also removed the internal wiring of the enclosures, which also was stranded copper and I exchanged it for the same UBYTE type. Furthermore I removed all internal damping wool from the prechamber that was still in there from the previous owners.

I reconnected one of the speakers with the old cable, the other one with the new UBYTE. Turning the balance knob clearly showed BIG differences: Treble and bass more extended, everything much more articulate with the UBYTE. Using UBYTE cables for both channels exposed another major improvement: The precision and depth of soundstage increased much more than I had thought it was possible! Furthermore no need for tone controls anymore, boominess problem solved — I thought. I can't say which of the steps (increasing prechamber, removing damping wool out of the prechamber, completely UBYTE cables from amp to driver) mainly made this success, but I personally believe it was the cables! I really can recommend these especially for fullrange drivers, because they seem to help in their weak areas: Extension of treble and bass; but also they improve their strong side: Most of all you gain a lot of overall precision with UBYTE cables!So far so good, I thought.

Looking back I must say, boominess was much bettered at this point but was not completely away. Through the success I had with CT 164 and UBYTE, I acquired a taste for upgrading the other components in my system. Next step on my wish list was a good tube amp. To make another long story short I finally found a second hand Audio Note pre/power combo M1Phono/P1SE, which I could buy for a fair price, and which replaced my old solid state Harman Kardon amp. When I was at this dealer's store, I brought my CT 164 with me. No boominess at all. Back home it was there again and even stronger than with my Harman Kardon! This opened up my eyes finally: My system is housed within a wooden cabinet. I found this cabinet starts to get in resonance at certain frequencies and makes a perfect feedback loop with my CD player and obviously even much better with my tube amp than with my old solid state amp! I placed those components on some squash balls for decoupling them mechanically and that's it - boominess is gone and away !

I hesitated a long time to write a review about Cheap Trick 164 because of that boominess-problem. After this is solved I can wholeheartedly recommend it to everyone. It was a major improvement over my previous speakers and brought the sonic abilities of my system much further. Even though these are "small" horns, they are able to fill my living room, which actually consists of two rooms of 22 m^2 and 3 m height each, both connected together by a large opening in the wall (so the effective area is about 45 m^2), more than loud enough for me even with orchestral music. I don't feel being limited by this small drivers; far from it I never experienced music so light, easy, naturally and powerful, as I do now, and still plenty of bass to me except for the lowest octave (<50 HZ) ! I don't miss that lowest octave right now and the rest of the bass range is reproduced much better than anything I've heard up to now.

At the dealer's store, where I bought my tube amp, I had the chance to listen to his "reference system", which consisted of electronic components from the German high-end manufacturer "Horch" and "Avantgarde Acoustic Duo" speakers! — I honestly can say from the short impression I got there, this mega$$ system did not sound better to me than my AudioNote/CT 164 system. It had some more bass due to the subwoofers of the Duos, but the overall detailed presentation, the clarity, the resolving abilities, it was also there the same way with my CT 164 !

Also don't underestimate the influence of those UBYTE cables which seem to synergize very well with fullrange drivers. They made more obvious difference to me than changing from solid state to tube amp (the latter was more subtle - not less important but it was not so eye-catching like the cable's improvement) !

I see it will not be easy to get a pair of FE108super, 6N-FE108super or 6N-FE108ES, which all should work fine with the slightly modified CT 164, but try it, it's worth it. Maybe it is also possible to order a pair of the brand new 6N-FE108ESII directly from Fostex in Japan, even though because of its once again stronger magnet its driver-parameters are more different from FE103sigma than those of the above mentioned drivers. I'd guess, it might also work with FE108sigma, maybe even with FE103 / RS 40-1197… Give it a try and modify.

My personal next steps will be some homebrewed interconnects and power cords based on Thorsten's UBYTE-cable-family plans (thanks Thorsten)… Then, as soon as possible, I hope I will leave all this dealing with technical stuff behind me and simply do, what the whole thing is about: LISTENING TO MUSIC (but at a much more pleasant and joyful level than before) !

Thanks to Nobuyuki, everyone here at this forum, at AE68's Japanese forum and elsewhere for your help!


Images of the CT164 drawn by Stefan from the Klang & Ton magazine:
Original plan inside
Original plan outside
Stefan's modified plan inside
Stefan's modified plan outside

Cutting dimensions:

19 mm plates:
2 Side plates for upper segment 772 mm x 450 mm
2 Side plates for bottom segment 460 mm x 230 mm
1 Bottom plate 450 mm x 300/150 mm (trapezoid shape)
1 Top plate for bottom segment 450 mm x 300/150 mm with a cutting of 320 mm x 112 mm
16 mm plates:
1 Front plate for upper segment 772 mm x 112 mm
1 Front plate for bottom segment 230 mm x 117 mm 9° angled sidewise
1 Backside plate 756 mm x 112 mm
1 Top plate for upper segment 434 mm x 112 mm
Internal plates (16 mm):
Plate #1 194 mm x 112 mm, 45°//22,5°
Plate #2 253 mm x 112 mm, 22,5°/\1°
Piece #3 (triangular strip) 23 mm x 23 mm x 112 mm
Plate #4 693 mm x 112 mm, rounded on both ends, radius16 mm / 8 mm
Piece #5 (triangular strip) 28 mm x 28 mm x112 mm
Plate #6 578 mm x 112 mm, 75°, rounded on one end, radius 16 mm
Plate #7 88 mm x 112 mm, 45°/\44°
Plate #8 88 mm x 112 mm, 4°/\24,5°
Plate #9 82 mm x 112 mm, 24,5°/\26,5°
Plate #10 574 mm x 112 mm, 26,5°//18°
Plate #11 236 mm x 173/117 mm (trapezoid shape, 45° angled on parallel sides, 7° on sloped sides)
Additional Front plates for the modified version:
Additional Front plate #1 1040 mm x 150 mm, 24 mm thick
Additional Front plate #2 1040 mm x 150 mm, 18 mm thick

Both plates glued together get 9° - angled sideways

Images of the CT164 from Klang & Ton:
Article text (420 KB)
Picture of speaker
Measurements 1
Measurements 2
Measurements 3
Measurements 4
Measurements 5
Diagram with Dimensions
Overall Dimensions
Entire Article on one image (659 KB)

Also see Fernando's implementation of Stefan's Cheap Trick.