Thanks to all readers and respondents throughout 2010. It's been exiting to view what some have made from the constructions suggested on these pages. I was pleased to see a few building the ScanSpeak Jensen 1071 speaker, and even more exited following those cutting sheets for the DTQWT. If you have read the entry level construction, TQWT, you will know why I'm particularly fond about this.
Last winter was spent making a lot of experiments on 1st order filters and of all the speakers made, only the PRELUDE and TQWT/DTQWT mkII have seen some of the fruits from this. The QUATTRO and OBL-11 took a lot of time as I wanted the finish right. After many, many hours have been spent in the workshop, uploading the intro page with a link to a new construction is always with mixed feelings. Is this speaker really the best I can make from these drivers? Are readers really going to pay some interest at all? What will I learn next year that might have made me do differently? Will people wire the drivers and crossovers correctly, damp the cabinets properly, etc., to hear how it was meant to sound? I could go on.
The final question is this: What's Next? The thing to do is relax and wait. Ideas will start popping up even before I think they should - and didn't I promise myself to leave speakers for a while and remodel the workshop this winter? And last winter?
Happy New Year!
When did a planned project take the scheduled amount of time? The vintage greencones certainly didn't. Making the cabs was more difficult than expected and once the drivers were mounted, unforeseen trouble and possibilities occurred. As is often the case with vintage drivers, these were in reality made for a specific application (radios) and bringing them into a modern world may call for modifications to make the best of it.
What also could be learned from this project was that it is indeed possible to make decent bass from a driver really not suited for bass at all by applying a resonant, thin-walled box. Adding ribs to the thin panels to better control resonances might be a never ending story and the use of lightweight solid wood, like spruce or cedar used for guitars, would be even more interesting. We're into instrument making and I'm sure several years could be spent in gaining sufficient experience to control the technology.
The final two month of this year will be devoted to finishing my next open baffle construction, again based on JA8008/TW034/Eminence DeltaLite 2515 drivers, this time with a full passive crossover between bass and mid. The crossover is finished and my OB-9 test panels are playing in our living room right now and most of the panels for the final construction are just waiting for glue and lacquer in the workshop. Sound radiated from a large panel - being close to an infinite baffle - is certainly different from any usual slim-line speaker with their inevitable baffle edge diffraction, and the wide panels have qualities that makes you wonder if this is not really what it should be like. Visitors' opinions are devided.
The OB-11 panels
will be part of my stock speakers. On one side these
panels are a gift to head-bangers as they can play
immensely loud with low distortion. On the other hand
I've never heard as much opera and live recordings before
as one these panels. They provide a live-feeling that is
The use of front grille is interesting I think. Ideally I would like to have a larger demo room with a room-wide, acoustically transparent curtain in front of the speakers, elimination the listeners ability to see what he's listening to. I'm sure this will cause some confusion - if not frustration. When we see a speaker and we see the driver set-up, we may already have made up our minds on what we think of this speaker and whether we like it or not. Obviously the listening experience counts as well, but my postulate is that we might choose quite differently if we had to base our preference entirely on listening. As humans we are - for very good reasons - enormously focused on visual appearrance. All our lives we've been trained to make judgement on visual appareance in order to know if what we see is something we should worry about - or not. Sound, odour, taste and feeling comes next.
|There are things we buy regardless of
price, just because we want to own it. Not because the
"thing" will provide us quality or pleasure no
other product couldn't, but because we have set our minds
to owing this particular product due to age, history,
show-off value, or whatever.
I recently followed an auction on eBay on a single (!) Altec 755A driver including a quite well preserved cabinet. Actually I bid up to 500 USD thinking this was my max bid to satisfy my curiosity on this driver. My goodness! This driver ended at more than 2,000 USD. Crazy!
Making an up-grade crossover for vintage Tannoy Monitor Golds, MG15, turned out more time consuming than anticipated. As always with vintage gear, we may run into surprises. The crossover data sheet from Tannoy prescribes a series coil to the low-pass section close to only one third of what was found in the actual crossover. Getting hold of the actual drivers also revealed that the schematics were far from ideal and in fact would jeopardize the performance of the driver. So, should the original schematics be maintained or should we tweak the crossover to a better performance? The fact that the Tannoy team made changes to the actual crossovers may suggest they were not satisfied themselves with the performance from the original schematics. In short it produces at huge bump at 1.5 kHz due to the steep impedance rise. Problem here was that even the change Tannoy made was in no way optimal, only taking the bump down to 1 kHz, albeit at lower level.
Peoples' attitude towards vintage equipment is very different. To some people vintage gear is almost sacred and shouldn't be changed in any way. If it only has to be kept on the shelf, this is fine with me. If the equipment actually has to be used we may leave things as they are and not really enjoy the full potential of an e.g. speaker driver. It's like digging out an old Gran Torino from the barn and leave it rusty and dented and never again enjoy the ride as it once was. Like old cars, speaker drivers may need new/restored suspensions to work properly. I've seen vintage drivers with softened spiders lacking elasticity and resilience and being pretty much useless despite pristine cone and voice coil. Towards loudspeaker drivers - usually made by the thousands - my feeling of reverence is modest. As the saying goes: It takes three hifi nerds to change a light bulb, one to do it and two to discuss why the old one was better!
By the way:
Check out this website if you want to protect your DTQWTs
from tiny fingers :-)
Tweaking a website layout seems more tricky than tweaking a speaker! Trying out varios designs created responses from "excellent, thanks!" to "remove the crap". Thanks to those taking the time to constructively commenting.
Making some kind of order and overview of the material accumulating over the last eight years is quite a task and I hope the new front page - being the key entry to all information - may help you find what you seek.
response from a professional website designer
and one of the things these guys have to fight is the
appearrance of websites on various browsers and display
settings. If my website pages look funny on your computer
and if you bother, it would be a great help if you would
send me an image of my intro page on your computer
|28 August 2010|
Tweaking for the better is
hard. Whenever we invest a lot of effort on improving
things, we start preparing our minds for the success and
may have a hard time evaluating the objective outcome of
our project. And at the end of the day, the objective
outcome is only a matter of statistics. If a majority of
people think it is a failure, there's not much we can do
about it. Have we made the best car, wine or amplifier if
nobody buys it?
tweaks is a page describing some of the things
I've tried over the last couple of years to possibly
improve the sound of the TQWT and DTQWT constructions.
Launching the TQWT mkII and DTQWT
mkII has indeed been given some thought. The
kits have sold in considerable numbers and customers may
wonder if they have an out-dated product. My best advice
is this: If you like what you have, stay where you are.
If you have box of coil/caps leftovers, the tweak is easy
to try out before buying expensive caps and coils
although the investment is modest.
The launch of the TQWT mkII and
DTQWT mkII follows closely the QUATTRO release. If 40
liters is more acceptable than 120-150 liters, this may
be a way to get the 8008/TW034 sound without upsetting
What interests me the most right now is whether people will build the QUATTRO. The construction is based on a lot of requests, but the diy-community is a tricky niche segment. It may want a lot of things, but when it comes to actually buying, things may be different. Time will tell. Sure is that this 40 liter bas-reflex design is going to be one of my - few - stock speakers.
16 June 2010
The AudioTechnology based PRELUDE was a long
journey into some new territories and as should-be, a
learning experience. Based on history, very few people
will build this speaker due to overall cost but maybe
some of the ideas can be built into similar designs based
on lower priced drivers. Time will tell.
Appearance is critical
for any speaker design. If it looks crappy, nobody will
build it. The Jensen speaker has that
problem. Being the best ScanSpeak construction on these
pages, it has zero feedback from people actually building
the speaker. Raw MDF panels simply don't turn diy people
The Chario Sonnet "review" was an interesting break from speaker building and gave some new ideas that will be tested in the future. A 34 mm dome has been modified by vented pole piece and voice coil to produce a smooth response down to 1 kHz and will be used for experiments with points of crossover in the 1-2 kHz range. Something I never found 1" domes being able to do properly.
Vintage speakers still keep me interested and the greencones are being investigated for the time being. Reproduced music can be enjoyed in numerous ways and as long as we don't push light-weight paper cones too hard, some enjoyable moments can be had.
21 April 2010
I'm pleased to have added the NOMEX 164 construction to my website. This speaker delivers a midrange that qualifies the midbass drivers for use in even bigger speaker systems as a dedicated midrange driver. Pity Peerless doesn't make an 8ohm/90dB/10inch version of the same. A 10" NOMEX, the 5.25" NOMEX middriver and the HDS tweeter might make a powerful, classic 3-way. The 10" SLS driver doesn't quite make it.
Having set up the NOMEX 164 construction, four SEAS CA18RNX drivers were at hand and would fit the front panel without further woodwork, thus the SEAS CA18RNX construction. Another 2½-way featuring SEAS 27TDC tweeter. As for the NOMEX 164, we get an awful lot of sound from a modest investment: Twice the power handling, twice the sensitivity and overall lower distortion compared to a single 6" driver having to do it all.
Audio Technology PRELUDE is in progress.
3 pairs of different 15H52 middrivers are being tested
and I think I have a candidate. Having the middriver in
place more tweeters have been simulated and there may be
changes. Hopefully this project will be launched over the
summer. So far, it's going to be the most costly speaker
I've ever launched.
drivers: I often have requests on specific
drivers and given the vast amount of drivers available, I
can only cover a fraction of them. New drivers are often
brought in by diy'ers needing a crossover for their dream
speaker and a 3-way from the SS Illuminator 18WU may
appear some day, possibly with the new SS Discovery 10F
for mid and HIQUPHON OWI for treble. Based on history I
know a lot of diy'ers give up when it comes to cabinet
construction, so let's see.
A center speaker based on JA8008/TW034 drivers is in preparation. Quite a few people have asked for a center speaker for their DTQWT speakers and not being a surround fan, I've been reluctant - or just plain lazy...sometimes I need a break from building speakers - to find new directions.
01 March 2010
been a long and cold winter and people in
Scandinavia and the Northern Europe - based on response
to my website - have spent a lot of time in the workshops
building speakers. Apparently there's no economic
recession when it comes to diy speakers.
What else has been going on in my workshop these recent months? From the most recent files, Study on stepped baffles with little impact on tweeter performance., Siri's Killer Note, it may shine through I've taken a deep look in the practicalities of making 1st order Butterworth filters. Actually it seems clear that it's easier to make a 3-way 1st order crossover compared to a 2-way. Nevertheless, the last month I've been listening to eight crossover versions of a small "6+1" monitor. In reality this can go on forever trying to improve phase integration and frequency response from changing the stepped baffle, tilt of cabinet and subsequent crossover mods. What seems mandatory is proper phase integration between 1-10 kHz. Reversing tweeter polarity we need a decent suck-out between these two extremes and things appear to start sounding right. One octave below and above point of crossover isn't quite enough. My best guess at this stage is 1½ octave and we're safe. It either takes drivers with exceptional linearity - or a lot of equalisation via notch filters. Don't be frightened by 3 notch filters in shaping the roll-off of the midbass. Fortunately these are in parallel to the driver and sometimes do very little - but may improve phase integration considerably. What also seems clear is that amplitude irregularities (linear distortion) are much less noticeable when drivers are driven from a 1st order Butterworth filter. Well in accordance with the findings of Vandersteen, et al.
I have this monitor in our living room run by
the JungSon and I like what I hear. I've listened to more
classical music from these speakers than I use to do.
Audience clapping - as always a great tool for evaluation
of speakers - are improved compared to high-order
filters. Clapping sounds darker with reduced impulse
distortion. Sibilance is much more tolerable, which is
great! Sibilance is a much too common problem due to poor
stage microphones and singers having a much to close
relationship to their mics. Next to this any sound of
acoustic instruments seems improved and more natural with
regard to timbre.
Oh yes, I've read Michael Fremer's review of Vandersteen Model 7 speaker (Stereophile, March 2010). The ultimate impulse and phase coherent speaker system with proprietory cones made from balsa wood and carbon fiber. I've been looking forward to this. Fremer says he could live happily ever after with this (45,000 USD) speaker system and only reservation is the lack of visceral slam from ochestral climaxes. No wonder, this speaker claims 85 dB sensitivity and Atkinson measures 84 dB. All this work and 84 dB sensitivity...Vandersteen must be really keen on keeping size small and bass extension deep.
01 Dec 2009
It's been two months since last speaker's corner and November was indeed a busy month of diy. Many, many mails with no particular theme, rather comments and questions regarding a wide range of constructions found on this website. Quite a few people have visited the 10C77 pre-study and the PRELUDE is up and running although not yet with the final midrange. The target sensitivity of the midrange has been handed over to Per Skaaning of AudioTechnology and hopefully I'll have the drivers before Christmas. In the meantime I can sand and oil the cabs.
Regarding frd and zrd files for speaker simulation: I do not provide these files from my LspCAD library. I have done a few times and people sometimes get all sorts of funny results and I do not have time to discuss how to use the files correctly. Sorry!
Looking at website statistics is always interesting and I'm surprised how many hits the L100 and Spendor BC1 have. The BC1 scores high month after month and apparently holds a huge audience seeking information. I have two front panels waiting for a BC1 refurbishing, which will probably turn BC1 lovers away in disgust, because I'll transform the cabs into solid, well-braces cabs, flush mount the drivers and make a new crossover.... How about that? I took the speakers to my livingroom recently and what a sluggish and un-precise bass performance. We should be able to do a little better here. For all the thin-walled cabinet is supposed to do, don't forget it also meant low weight, reduced production cost and not least, reduced transport costs. I don't think we should be too blue-eyed here.
Asking for contribution to the
website was certainly not something I'd ever
considered until my webhotel gave me three options:
Reduce file size/upgrade to superpack/get lost! This is a
little more polite than last time where my former
webhotel just shut down my website without any warning.
Who are these people running webhotels? Robots?
Apparently they don't have names and they like to express
themselves in one-liners. I mean, I have no interest in
annoying these people and something like "Mr.
Gravesen, your traffic is exceeding so and so many
GB/month and you need to udgrade you hotel." No
problem! Of coarse I will.
11 Oct 2009
Making a crossover and connecting
the drivers right may not be as easy as it seems
from a quick look at the crossover schematics. I steady
flow of emails from first-time diy'ers suggest that
reading this file may be a good idea: the
importance of wiring crossovers correctly.
I'm pleased and proud to have the Audio Technology 2-way back again on the website, now with the SWANS RT2H-A planar tweeter in replacement of the Raidho planar tweeter. Read here about the new AT-SW.
And I've been visiting Audio Technology to pick up the 10C77-25-10-KAP 10" bass drivers with the new sandwich cones. Initial measurements suggest a 45-55 liter cabinet volume and a decent response down to 30 Hz. The sensitivity of the driver suggests a system sensitivity of 90 dB being possible. A pre-study has been done to gather information on performance before cutting sheets: 10C77.htm
These bass drivers are some chunky beasts weighing more than 8 kgs each. Measurements suggest a flat response up to 1.5-2 kHz and for fun I tried modelling a two-way system with a point of crossover at 800 Hz. No problem at all. Next I'll do some measurements to compare the drivers with ScanSpeak 26W/8861T00 and SEAS W26FX002.
11 Sept 2009
In June's Speaker Corner I made
some comments on boring "6 +1 inch"
speakers and I won't change a word of what I wrote.
However, a recent standmount speaker project
incorporating an Audio Technology Flex Unit
6A77-25-10-KASD (4 ohms version) got me exited again on
small speakers. This driver features a generous 3"
voice coil former being a composite of both aluminium and
kapton. Now, having some 174 square centimeter membrane
area, this driver falls well in between a 6 inch and an 8
inch driver but I think it will be hard to find any other
8" driver that can truly match this driver when it
comes to power handling, dynamics and resolution. It's
phenomenally good and should you ever wonder if
high-priced raw speakers are worth while, not to forget
premium crossover components, this may be the one can
demonstrate that proper design pays off.
Up-dating the ATR and ATR25 constructions for the new SWANS planar tweeter obviously made we wonder if this high-sensitivity planar was suitable for the DTQWT design. For a couple of months I had the set-up running thinking it was work done! It took an old friend to tell me this really wasn't as good as the old one. Being your own hardest critic is not always easy. Thus the DTQWT goes on as-is and is my reference speaker. I have made drawings for 1.5 meter tall version for 10" and 12" bass drivers, but when they'll see light is hard to tell. Woodwork takes time.
And no, I
have not tried the RAAL ribbon and
I won't. I think I have a mail every second day if I have
tried this ribbon. I've tried a lot of alu ribbons and
generally find them suitable for metal cones and some
polyprop cones as well. For paper pulp cones - generally
not a good match to my hearing. Planars usually goes well
with polyprop and sometimes (heavily) coated paper. Alu
domes for alu midbass drivers and sometimes polyprop or
coated paper. Good softdomes goes with everything without
always making the perfect match. Very little
experience with ceramic domes.
08 July 2009
16 June 2009
Setting up the Summer-09 open baffle project, the OB9, from JA8008/TW034 + Eminence 15" bass drivers made me realise how utterly boring conventional speakers can be. How can I ever make a 6" + 1" midbass-dome-thing again? I will be doing it, actually a lot of them, but it takes a little time to recover and see the fun in small, boxy sounding and un-dynamic speakers again. Most speakers simply cannot even come close to rendering the true dynamics of an upright bass the way a 15" driver on an e.g. open baffle can do. Small speakers may to some extent be emotional engaging, but only when we hear a really large bass driver on an open baffle - or in a horn - do we also get the physical engaging sound we're longing for. We finally get the "punch and the drama"! The featherlight kick in the stomach - and we're looking at the bass-drivers to see if the cones are thinking of leaving the chassis' - and they're certainly not. They barely move at all. Magic! With a total of 1700 square centimetres membrane area, things start to work out. No wonder some 130-150 cm^2 from a single 6" driver can't do much. Even a 2½-way system from 6" drivers won't count for more than a total of 500-600 square centimetres.
A rather big ScanSpeak construction has been launched recently, the Jensen, from 10" ScanSpeak bass driver, well-known 18W/8531-G00 for mid - and the 18W has never sounded better - relieved of pumping air and at the same time trying to cope with upper mid and lower treble. Goes well at modest sound levels, but here we render the full potential of this fine driver. Last but not least I was pleased to get first-hand experience with the 9900 tweeter. It's great, the best Revelator I think and on level with the excellent SEAS T29 domes.
The DTQWT is doing well and
a lot of kits have been sold and I'm looking forward to
receiving comments and pics from the buyers. I've had a
number of visitors for hearing the DTQWT and the response
has been good. Whatever people bring of CDs and LPs, the
DTQWT will play it all. From classical music to the most
outrageous heavy metal stuff.
My new line stage project, 5687, is coming to an end, and the hum-war is on hold from the final version. The DTQWT system is not totally free of hum yet, but now it's one of the Audio Mirror amps that needs some attention. The blessing of home-made electronics...
I feel urged to comment on the
quality of some recent Verve vinyl pressings
from the US. Diana Krall, Anthony Wilson, etc. 180-200
grams "super" pressings and the vinyl surface
is crap. Visually blurred and sometime so noisy it's
looks like it's been taken hot from the mold and thrown
into its sleeve. How come? Who's pressing these LPs? Has
anyone experienced the same thing?
08 April 2009
It's been a busy winter with quite
a range of SEAS constructions and Scanspeak
crossover for speakers coming in from diy people here in
Denmark. Check ScanSpeak
constructions and SEAS
constructions on intro page.
These early days of April the temperature has risen to some 15 deg. C and it's time for spending time outdoor, digging the vegetable garden and getting rid of the garden debris from five long winter months. Speakerwork will slow down for some months. Not least due to amplifier building although an open baffle speaker is on the to-do list; the summer fun-project, like the OB7. This time OB9 from 8008/TW034 and a 15 inch bass driver.
Identifying bottlenecks in
our systems is a never ending exercise. Few years a go I
had a guy visiting with all his silver interconnects and
silver speaker cables and we weren't able to hear any
difference comparing these to my rather mixed blessing of
cables. Not at all. The set-up at that time was a modded
Rotel CD player, Audio Research SP16L line stage and
Copland CTA505 power amp. I don't recall the speakers at
Once and a while we may be fortunate borrowing a piece of gear that may identify a true bottleneck in our system. The 5687 line stage with output transformers was such a piece of equipment, originating from diyhifisupply in HongKong as the TRAM w.o.t. line stage. Product is discontinued, probably due to cost of output transformers - or maybe hum problems, because the original TRAM has some serious hum problems only realised when you run real high-efficiency speakers. Cloning the TRAM isn't particularly difficult, but getting rid of the hum turned out to be serious trouble. So much trouble that I'm rebuilding the line stage with a separate power supply unit. This line stage is so good I'm ready to go all the way and I have even ordered a new power supply trafo to get all the secondary tabs I want.
with the silver interconnects, I thought it was time to
try out that silver foil for speaker cables I've had on
the shelf much too long. So, the foil was wrapped in
paper (painters' tape) and I ran bi-wiring to the DTQWT
having silver plated copper in teflon for the bass
Things like the above certainly need repetition! So, next day I installed the bi-wiring system again and it wasn't quite as bad as yesterday - late evening - maybe I was tired, but something certainly wasn't right. Midrange/treble integration still didn't work. So I ran the DTQWT from the silver foil alone and now things started to develop in the direction I had expected. Silver does something to the midrange in particular. It sounds less lush, less coloured, less smeared. It sounds very cool and clean and you wonder whether you would voice your speakers differently had they been developed with silver wires, etc. How about a DTQWT set-up with silver wires, Mundorf Silver/Gold caps, Jantzen Audio Silver Coils, graphite resistors and - - dream on. This would triple the cost of the kit. Anyway, foil wires are trouble. Fragile, trouble when you need cleaning around the speakers, low WAF, etc. What I really had in mind was making a multistrand silver wire from e.g. 12 x 0.4 mm silver wire = 1.5 mm^2. I only need 2.5 meters of speaker cable + 1.25 meter for internal cabling in the DTQWTs, thus (4x12) x 3.75 meters = 180 meters of wire = 235 grams silver = 1400 DKK (250 USD) from my local supplier. Not too bad after all. Let's see.
15 February 2009
Yesterday was Saturday and with
the lady of the house gone shopping, it was time for some
serious kick-butt, rock'n roll. Front end was Garrard 401
+ LINN ITTOK LV II arm + Dyna XX2 mkII
cartridge/Transcendent RIAA with Mundorfs/TRAM w.o.t. line
stage (DiyHiFiSupply)/modded Audio Mirror 6AS7
mono-blocks/Rotel RB981 for bass drivers in DTQWT
speakers. The TRAM line stage has out-put transformers
(w.o.t. = with output transformers) and can easily driver
two power amps.
Which makes me think of a visitor I recently had, bringing in a 20,000 DKK (~3,500 US $) TentLab CD player to compare with my slightly modded Shanling CD-T80 (I paid 650 US $ for this player from Hong Kong). Playing a wide range of CDs, we really weren't able to hear any difference at all. Now, my visitor also brought a CD which I happen to have on vinyl (Eva Cassidy) and this called for some A-B testing of vinyl vs. CD. Visitor was pretty shocked from the difference. Needless to say the CD sounded flat and congested with poor dynamics and transparency compared to what the XX2 cartridge could pick up from the vinyl grooves. Now, CDs can be good and bad. I have a (very few) excellent CDs where everything seems to have been done right to get as much information through the media as possible, but generally most of the CDs I have on my shelves are poor. In Stereophile, Feb-09, you can read about a demonstration at DIY Burning Amp Festival, where " - Jan played back samples of the same song released over a span of several decades. Each one sounded worse, more compressed, than the last. It was so shocking that it made people angry". No wonder!
Recent Months have brought a range of new speaker constructions : SP44, CNO-mkII, CNO-T25, Maya (SEAS CA18RNX/T25C003) and CNO-25 (CNO 2½-way), and I guess I'll take a break from SEAS constructions after this. I need to put together some plug'n play kits to finance new projects. And I'll be cloning the TRAM line stage and have started sourcing components. The project I'm looking forward to the most is an open baffle construction with the JA8008/TW034 drivers and some 15" at bottom. This OB should as close as possible look like the old Wharfedale SFB. I just love that design.
28 December 2008
Economic recession or not, people are building speakers like crazy these dark winter evenings and weekends - at least on the Northern Hemisphere - and coming home from work, I find my mailbox full. It takes an hour or so every evening and please forgive my short replies. Usually the Christmas holidays are very quiet with mails, so time for some writing and a few experiments.
There are a lot of speakers in pipeline
the coming winter months: CNO mkII, CNO-T25 (T25C003
tweeter) and a CNO-25 (2½-way CNO). I have reworked my
test cabs for these constructions.
A Happy New Year to all diy'ers visiting my website!
07 November 2008
Do you have the same experience
as I that sometimes your hifi system sounds crap and
sometimes just sounds great? Nothing in the system has
been changed and yet there are good days and there are
I'm tired after a normal working
day filled with mails that have to answered, meetings I
have to attend, talk with my colleagues, driving my car
and paying attention to heavy traffic, watching the news
and the Obama election, etc., etc. The everyday contains
an enormous amount of information that has to taken in
and/or taken care of - and after all this I'm really not
in the mood to be fully aware of subtle differences in
the performance of hifi components. Good or bad is
sometimes a matter of state of mind and not how the
system is performing.
Loudspeakers are the most troublesome
components in our system. We not only listen to our
speakers, we listen to our speakers and the
room, and not only the speakers in our room, but the way
the speakers are placed in our rooms.
If you happen to have a listening
room of e.g. 20-30 square meters of reasonable
dimensions, square is often worst, take a day off and
move your system around including all your furniture. Try
every possible positioning of your speakers and hear what
happens. If possible try your system in a smaller room
and a bigger room too. Try the bedroom also! At least
here we may have some fabrics. Now, how does it all
sound? One room better than the other? One position
better than the other?
By chance I came about this
19 September 2008
"Could I use 2 x SS
18W/16831G00 in replacement of the single bass driver in
the SP38/13 construction and
use the same crossover?".
"Could you change the page
set-up in your files because I have an old
monitor and 800 x 600 pixel setup?" Now, that's a
new one! I have asked around and I can't find anyone
running this format anymore. I have a standard
laptop/14" screen and usually make my tables
1000-1100 pixel width. I use tables because it allows me
making the page layout I want (from my stone age Frontpage
version, I admit). I'm not particularly interested in
computers (- but I love what they can do) and usually
stick to my old software if it works well. There are
about 250 pages on my website and changing the format
would be an enormous task. So, "sorry, won't".
The poor devil buying my TJL3Ws hadn't checked with his spouse in advance and all hell broke loose when the speakers were installed in their livingroom. Wood finish didn't fit furniture veneer and to avoid sudden divorce, the speakers were wrapped in white plastic foil.... All that nice woodwork - for Heaven's sake! Point is: Don't forget to check with your loved ones before engaging in new speakers!
More WAF issues: I had a mail from a guy showing some of my constructions to his spouse to hear her opinion about possible designs. Now that's the way to do it! Her response was: "All these speakers look like coffins! Horrible!" She has a point. Most floorstanders may look more or less like upright coffins. We only need some flowers to bury our addiction to music! The single construction she did like was the Acapella. So, I asked my wife the same question and she was in favour of the Acapella too, comparing all the speakers she'd been living with over the years. Hmm... I always liked these cabs better than anything else I've made, but it will be tough squeezing the DTQWT into an Acapella design!
Thinking of coffins, these come in all
shapes and sizes and maybe this would be an options for
those engaging in speaker-building for the first time. As
can be seen above someone already got the idea. I have
seen speakers looking more terrible than these. Some
coffins come fairly cheap. Add braces and bitumen pads
and a pair of 12-15" bass drivers and I'm sure you
could impress you bodies - but I guess it's for
"singles only". To keep green-house gasses down
you can even buy coffins made from recycled materials
degrading without environmental damage. Fully RoHS
I suggest reading the whole article at
Stereophile's website once published: http://www.stereophile.com/thefifthelement/
An extensive test of
coupling capacitors has been performed on my 300B integrated amp and this is
a never-ending exercise. Mundorf silver/gold/PP and VCaps
are the winners here, but read for yourself.
Last but not least: Some bedtime reading on the state of hifi (Gordon Holt/1991 and Art Dudley/2004). Download here.
23 August 2008
Finally: The launch of the DTQWTs. It's been a long, long journey and hifi - to me - has never been more exiting. The DTQWTs are doing what I want my speakers to do: A huge soundstage, a midrange delivering the presence and detail I want without being harsh or aggressive, deep bass delivering the punch and drama we want from reproduced music and last but not least: High sensitivity allowing low-wattage triode amps to be used. With these speakers I can easily evaluate subtle differences in the quality of components used in my amplifiers. My Audio Mirror 6AS7 SET mono-blocks and the 300B SET amp are already revealing interesting - and different - ways of amplifying source material and soon a Transcendent T16 OTL will be finished and tell if omitting out-put transformers really is as important as Bruce Rosenblit says.
28 July 2008
The final DTQWT cabs are shaping up and only needs the final lacquer before finished. The DTQWTs pretty much do what I want my speakers to do and I might just quit speaker building here. My bank account to some extent reflects this as I spend more money on vinyl and less on new drivers. That's how it should be! Well, a few more amps are on my wishlist and besides a diy 300B SET under construction, a Transcendent T16 OTL has arrived.
The DTQWTs are sensitive and they project a big, wide and transparent soundstage and allow me to use all sorts of amplifiers, but I do rush to say I do not recommend 2 wpc SET amps. No matter what loudspeaker and amplifier producers may claim, there are limits. I do recommend 6-8 watts minimum. Often we see that 2 wpc amp producers may claim their products suitable for minimum 90 dB speakers, but be cautious, it very much depends. And speaker producers may claim their 90 dB speakers suitable for 5-10 watt amps. Again, be cautious. I recently made a crossover for a large 3-way speaker (ScanSpeak drivers) and I ran it from a 6 wpc SET amp and it sounded great - but indeed not very loud. If we run 2 wpc SET amps we need humongous horns with extremely light-weight paper cone drivers to render a true presentation of bass fundamentals.
Anyway, there will be more speaker constructions in the future and more 2-way/6"+1" kits and some 3-way classic kits from ScanSpeak 21W/8555-01 + mid + 8513 tweeter. I have a range of mid drivers that look very promising from simulation, i.e. simple 2nd order filters and flat impedance curves. The D2010/8513 is special in this context and seems just made for 3-ways with a relatively high point of crossover between mid and tweeter. Modern 1" tweeters are often designed to extent the lower end - to make two-ways - and some of the 3-way simulations I have made got into serious trouble when replacing the 8513/OWI with e.g. SEAS 27TFFC, SS 9500, etc. These 1" dome tweeter often takes 3rd or 4th order filter to perform from e.g. 3.5 kHz in a 3-way set-up. The EKTA might do well with the 8513 tweeter from a simple 2nd order filter at 3.8 kHz - but I'm afraid people wouldn't believe it.
The summer sale quickly ran out of 2-way kits and I need a break before launching more. Selling a kit requires quite some hours of actually making the kit with presentations, etc., and often takes 10-20 mails before it's actually sold, not to forget packing the kit to ensure safe transport. But it's also fun hearing from people all over the globe and exchanging views on speakers building.
One of my summer holidays has been
spent reading reviews - speaker reviews - and gosh, are
reviewers and manufacturers one big family! I'm amazed by
how reviewers manage to write 3-4 pages on a new speaker
that is slightly different from the old model. We are
told it's a significant improved product, yet in no way
does it diminish the qualities of the old product. So,
we're all happy! A few times I may actually know first
hand the new drivers - don't ask how - and know the
changes were made to cut production costs and increase
profit. I wouldn't never blame manufacturers for doing
so, but the way reviewers are able to handle this is a
matter of linguistic artistry. Bravo!
Buying on eBay has been
low for the past months, but a single Wharfedale Super
8 FS/AL was found - and bought. And it's one of
the better drivers from the Fifties. Much better than the
terrible Super 8/RS/DD. As expected,
not much of the foam surround had survived transport but
I managed to find some fine felt material and have made a
new surround. What I don't get from these Wharfedale
drivers of the Fifties and Sixties is the acoustically
transparent outer surrounds. Why? Even some of the later
fabric surrounds were acoustically transparent. It causes
serious trouble in the midrange. It doesn't fit with the
response graphs found in Briggs' book. Any ideas?
17 May 2008
Launching a sale takes a deep,
deep breath! Space and finance is the reason
behind this and I know it takes tons of mails back and
forth before diy-people decide to buy anything. But here
we go! Take a look at the '08
Summer Sale' if you might be thinking about
the following items: Acapella NEXT, TJL3W, Compact Studio
Monitor Kit, JBL L100 matched drivers + crossover KIT,
ARC SP16L preamp, TubeBox phono stage, Rega
RB300/Incognito tonearm and W11XT speakers.
7 May, 2008
It's springtime and really not a time for doing loudspeakers, but you can always gather components for the next project while digging the vegetable garden and mowing the lawn. And so I do - like seen on the photos below. 10" bass drivers, wax coils, super caps, cables in teflon wrap, low-ohmic C-coils, new cable plugs, gold plated soldering tags, etc. All meant for the DTQWT project, i.e. Double-Tapered-Quarter-Wave-Tube.
The DTQWT is the JA8008/TW034 kit
with two 10" drivers on the rear panel loading a
common center horn together with the front driver. No big
deal really, but finding the right bass drivers and the
right dimensions of the horn has taken time. Some 5
different 10" drivers have been tested and two
candidates seem to do the job well. One is Beyma 10LW30N,
rather expensive but with excellent build quality. The
other is a 10" Pioneer driver derived from the US.
This one is cheap and does a great job too - at a
fraction of the cost of the Beyma drivers.
My system has been through a lot of up-grades lately and this has urged me to use the very best I can afford for the DTQWT. In addition to the components seen above I'm doing homemade silver foil speaker cables and graphite resistors for the crossover. I'm not sure all these measures will contribute equally to the good sonics, but for the benefit of the doubt - and to gain experience.
Gathering components also includes bits and pieces for a 8 wpc 300B SET amp. I need a really low-wattage amp for reference with the MAE now back in Norway. The 300B amp is based on the Oliver design shown here.
20 March, 2008
I've bought a new phono cartridge!
I never invested heavily in these things and I was pretty
happy with my Dynavector 10x5 until a wire from one of
the coils broke and these tiny wires are way too thin for
repair (I tried). So, a deep pocket dive and a Dyna
DV20X was in place. 400 £ for a cartridge was
serious money I thought, but it provided one of these
rare occasions where you - once more - have to go through
a lot of your vinyl collection to hear what you've missed
over the years. Self-pity is a really bad thing, but it
came close and it kept me wondering what I would get for
800 £ or maybe even 1600 £. Better start saving for
that next step.
The ATR25 was finished early March - took a long time to get the cabs finished - and these are "disturbing" speakers. Disturbing because they do a lot of things better than any other speaker I have. Still don't beat my JA8008/TW034 - a matter of taste and WAF - but at half the size - and twice the cost - they take midrange clarity and treble quality to new levels. No domes can do what these planars can do, that's for sure. My 20 wpc SET mono blocks have no trouble driving these transmission lines and I'm having good times in front of these speakers. Despite high level of transparency, they handle all musical genres equally gracefully. They don't make poor recordings sound worse. Low distortion is the key word.
Right now the JBL L100 Recreation is on the workshop bench. Cabs are finished and I need to add damping, install crossovers and drivers. I think I've spoiled the L100 kit as much as possible and I'm looking forward to hear if it all makes a significant improvement compared to the originals. I think it will. The original box is rather resonant and having the 12" bass driver handling most of the midrange, there just gotta be serious improvement gained from a well damped and braced cabinet.
On my way home from work today I picked up some 5 square metres of 19 mm DMF (pre-cut to some extent) for my double TQWT, the next JA8008/TW034 system. Have collected 10" bass drivers over the last half year and I'm expecting to have some 5 pairs for test. Two pairs of Beyma drivers, two pairs of SEAS drivers and I'm waiting for the last pair of 10" drivers from the US. You can find the principal "DTQWT" design here and here: Two conical horns feeding a common center horn mouth.
12 January 2008
The Acapella NEXTs are up and running: This project started like: "Well, I have the cabs and only need a bass driver" and turned out to one of the most rewarding projects I've made. You never can tell.
There's a long list of projects planned for 2008. After the Acapella NEXT, the ATR 2½-way is next. View ATR-2-way here. Initial test set-up will be a bass-reflex design and initial listening test from a quick simulated crossover is promising. The final cabs will probably be a classic transmission line with the same front panel dimensions, allowing the crossover to be the same for both versions.
the ATR-2½-way is finished, the JA8008 55-litre
bass-reflex is scheduled. The project has been on hold
for some time as initial tests in a classic 36 x 80 x 28
cm (W x H x D) cabinet didn't sound right. The crossover
was easy, measurements were good, but the sound was just
too forward. Too much midrange despite an overall flat
response. My main suspicion is the width of the front
panel and a more narrow design will be tried. Actually
similar to what can be seen on the sketch shown at the JA8008
Next project is something I'm really looking forward to do: A "W26-Classic" 3-way from SEAS W26FX002 + M15CH002 + T29CF001. Possibly the middriver is going on top in a separate semi-open cabinet. Having the mid on top is seen before in older KEF & B&W speakers. However, my initial inspiration came from the Spendor SP100 and Naim DBL. We'll see.
The "W26-Classic" cabs will also allow a test set-up of my JBL L100 drivers. Modelling the 123A bass driver in a 65 litre cab looks promising.
Did I say 2008? I also
look at my Spendor BC1 speakers from time to time and I
have 22 mm plywood sheets on the shelf for new front
panels... and I've had a pair of "vintage" SEAS
EXCEL W21EX001 drivers for a reasonable price. These
drivers have a hard paper cone and performs exceptionally
well although sensitivity is some 3-4 dB lower than
promised from the old SEAS data. An "EXCEL
VINTAGE" from W21EX001/W11CY001/tweeter could be
Speaker Talk 9 December 2007
SEAS has got
a new website and it's certainly better than
before. And they are about to launch a new
"alnico-whizzer" driver - with rubber
surround...Hmm... Have they heard the cry from the
SET people right?
While the Norwegians seem
fit for fight, I'm wondering how things are in Videbaek
town, not too distant from where I live. Apparently
people are fleeing the castle of Vifa/ScanSpeak and have
started new companies, e.g. SB Acoustics and GAMUT, based on former
dst employees. I'm sorry to say that
whenever you guys "over there" take over
smaller companies "over here", things seem to
rapidly deteriorate - from our perspective. RCF, Audax
and now Vifa/Peerless/ScanSpeak are examples. I'm
wondering if the Tymphany air-pump can compensate for the
departure of some key figures in the former dst
organisation. This is - to my eyes - the only radical new
product from the Tymphany company for a long time.
This Month's Speaker Talk
was meant to deal with a few reflections on how
differently we perceive sound. Not a new topic, but I
still wonder sometimes how people claim aural nirvana
from speakers I dislike - and know first hand - like the
vintage Philips 9710. I had a mail from a studio
engineer, seriously suggesting this driver as a great
tool for monitoring. I don't get it.
"I have found myself
100% captured by listening to 12 x 17 cm full range
drivers mounted in cardboard boxes with an open back. I
have also found myself 100% involved by listening to 40
cm coaxial drivers. Every speaker has something to say!
The beauty of the sound of some small driver may weigh up
the power of some big driver. On the other hand, the
sheer power and authority of the big driver may please me
more than the delicate sound of the small ones. It
depends on my mood and on the sort of musical material. I
have had lengthy discussions with a former Klangfilm
speaker engineer and we have both agreed that the whole
matter depends very much on proportion and on harmonic
and well balanced tonality. If you listen to a 20 cm full
range driver you will obviously not hear the authentic
impact of a full scale big orchestra. But provided
it is a good driver you will hear the
orchestra as an entire phenomenon. On a smaller scale!"
In line with these words,
I recommend reading an article by Art Dudley
(Stereophile, Nov-07, page 41) on a speaker system from Aurum Acoustics: