Diy Monitors - to build or not to build

[TomWaterman]

I think this was mentioned earlier but

http://home.hetnet.nl/~geenius/Proteus.html

is a similar idea to the Cubs and Joe's box. Avalon also make a D'App monitor with the same time aligned drivers...seems like a popular design style.

Tom

 

[mhelin]

Yes but these Wilson style speakers are more like mid-farfield speakers, not nearfield. Besides they are too big for that (next to your DAW display). The time alignment of the drivers is the strong point with these speakers. If you use regular dome tweater you could also use a waveguide like here:
http://www.zaphaudio.com/hornconversion.html
Benefits are the controlled directivity (better match with woofer's) and reduced distortion.

Though for nearfield listening this large waveguide is no good. You want to have the woofer and tweeter as close to eatch other as possible to avoid lobing effects (which are very strong in D'Appolites in vertical axis, practically this means comb filtering effects due to phase differences).

Ina ddition to Jordan full-rangers also Tangband,Aura and Hi-Vi have good drivers. John Krutke has a nice design here: http://www.zaphaudio.com/audio-speaker18.html

This too needs a subwoofer for low bass, but it sounds excellent in nearfield. There are also other interesting small speaker designs that would work fine in nearfield.
~Mikko

 

[Marik]

[quote author="mhelin"]
Ina ddition to Jordan full-rangers also Tangband,Aura and Hi-Vi have good drivers. John Krutke has a nice design here: http://www.zaphaudio.com/audio-speaker18.html

This too needs a subwoofer for low bass, but it sounds excellent in nearfield. There are also other interesting small speaker designs that would work fine in nearfield.
[/quote]

Those puppies are cool! It is actually very attractive idea to have a small wideband driver, without crossover in midband, and use it with a sub.
Now I got into speakers mood. Searching Partsexpress, I found TB W3-1364SB. They are 105Hz-20K and graphs look little better than Hi-Vi, but who trusts graphs? In a closed 4.4 liter box it looks like the -3db point will be around 125Hz. Who knows how they'd sound, but lots of fun for not much money. At least, for computer speakers would be perfect.

I want to finish my started 7 years ago D'Appolito, and got Audax AP130MO, along with AP170MO, which are on closeout sale for around $10 at Madisound.
Just want to see what they do.

For monitors, so far I am thinking of Seas W18EX001 woofers (used in Thor). For the top whether Heils, or make 3" ribbons, if I don't like first.

 

[bcarso]

"Ina ddition to Jordan full-rangers also Tangband,Aura and Hi-Vi have good drivers. John Krutke has a nice design here: http://www.zaphaudio.com/audio-speaker18.html

This too needs a subwoofer for low bass, but it sounds excellent in nearfield. There are also other interesting small speaker designs that would work fine in nearfield. "

I'd love to see the off axis response plots for these. My own leanings are to still smaller full range radiators.

I had a client briefly who lost their direction after the one visionary with the outfit quit, but we managed to come up with a prototype system using Aur*sound radial magnet transducers: two 1" and one 4" long-throw woofer in a single box. It was going to target the B(l)ose SoundDock after suitable industrial design.

It had a 50W switcher for the woofer (closed-box with 2nd order shelf) which in a typical domestic environment for some bass gain had very satisfactory response to about 45 Hz. About 12W each was available for the left and right drivers. The 1" drivers are wonderfully linear hence low IM distortion. The whole system sounded so good that I found myself wasting time listening to music on it. It of course totally blew away the competition.

 

[mhelin]

Aurasound drivers are pretty good, actually Linkwitz uses them also:

http://www.linkwitzlab.com/Pluto/construction.htm

John "Zaph" Krutke tested many 3" full ranges including Aura NS3-193:

http://www.zaphaudio.com/minitest/

He considered the Aura the best of the drivers which need no correction EQ as it's response is quite flat without one. For the Hi-Vi and Tangband he uses an eq circuit. Obviously you have to use the baffle step correction filter anyway. Zaph doesn't seem to like 2" drivers at all, but for me the NS2 Linkwitz is using (looks promising. I've used just TB's (W3-871s), it's ok but little bit colored and obviously at lower frequencies very warm due to 2n and 3rd order distortion components (built them into Cyburg's Needles for my daughter and ~3L bass reflex tuned to ~60-70 Hz for myself). Oh yes, these has to be listened on one spot and very close field, I think their off-axis is rolling off heavily towards high frequencies, but there may be single peaks at upper tones (due to phase plug).

The NS2 is only $16 at Madisound (http://www.madisound.com/cgi-bin/index.cgi?cart_id=5653076.18773&pid=102), and 24 EUR at SpectrumAudio (http://www.spectrumaudio.de) .For the woofer some Nomex cone DST (http://www.madisound.com/cgi-bin/index.cgi?cart_id=5653076.18773&pid=1956) might be good one. Here in Finland at least Genelec, Amphion and Chorus seem to favour Nomex drivers.

 

[Butterylicious]

Here's a picture of the speakers I want to clone.
http://www.peterpringle.com/visitgoldie1.html

 

[Marik]

I checked the large Heils and can already see they would not work.

I have some huge Neo magnets for ribbon tweeters. In 16 mm gap I'd get 6000 Gauss field, which together with 5um foil should give at least 95db sensitivity.
For good vertical pattern I make the ribbon shorter, with area 15mmX50mm. It should give crossover point probably around 1.5K-2K.

Can you think about cons of using ribbon tweeters for nearfield mixing work? I just don't want to make it just to realize that they wouldn't work for my purposes.

 

[mhelin]

I think ribbon is OK in nearfield, and Adam thinks so as well:
http://www.adam-audio.de/professional/
http://www.adam-audio.de/studio/system/s1a_data.htm
http://www.adam-audio.de/studio/nearfield/apple_data.htm
http://www.adam-audio.de/studio/nearfield/p11a_data.htm

 

[Marik]

[quote author="mhelin"]I think ribbon is OK in nearfield, and Adam thinks so as well. [/quote]

The tweeter is not a ribbon, but an air motion transformer, which is quite different. I read Adams uses Eton drivers:

http://www.madisound.com/cgi-bin/index.cgi?cart_id=3513088.22287&pid=1704

This one is quite small. The smallest original (much more affordable) Heils are 4" long. Although they are quite remarcable drivers, there are three main issues: 1) Becasue of their size, off axis is not good. 2) The sound is too "big" and right in your face, for using in nearfield. 3) They are kinda too sweet and musical, rounding off some "corners".

I am sure the Etons due to their smaller size, are free of most of these, but the major issue is their price--over $600 a pair+ shipping.
DIY ribbon tweeters would be well under $100 for a pair.

 

[mhelin]

[quote author="Marik"]The tweeter is not a ribbon, but an air motion transformer, which is quite different.
[/quote]
Ok. I've seen a design of a small speaker using Visaton driver, TI100 midwoofer and MHT12 ribbon. The designer recommends it for nearfield monitoring and for surreound systems. It's from a book "Tee itse hifihkaiuttimia" (you understand finnish, right?). The speaker uses a waveguide made from small aluminium sheets. It is this one:
http://www.hifitalo.fi/shop/sivut/hifi-52.htm
I cann't see why you couldn't use a DIY ribbon, but can it be made as small as the Visaton driver?

http://www.visaton.de/english/artikel/art_684_1_12.html

That's about 140 EUR btw.

ps. Oh my, they tell is a magnetostat, not a ribbon. What's a ribbon?

 

[Marik]

[quote author="mhelin"]
ps. Oh my, they tell is a magnetostat, not a ribbon. What's a ribbon?[/quote]

Yes, these are magnetostats (or isodynamics). The difference with ribbon is in magnetostat there is a coil printed on a mylar base and the diaphragm is suspended around the whole perimeter. Very often they call it (wrongly) "ribbon". However, the true ribbon is a piece of corrugated aluminum foil suspended only ot the top and bottom. Some of the true ribbons are Decca, Raven, Fountek, Dick Seqquera, etc. Magnepan made both.

Both of them have their own limitations as tweeters. The magnetostats are more robust, but since the tension of the diaphragm is much higher and excursion is much smaller, they cannot go as low as the true ribbons. They are much cheaper to make, but it is easy to call them "ribbons" and charge premium.
The magnetostat's diaphragms are not nearly as compliant as of the true ribbons, so the transient is not as good.

On the other hand, the true ribbons are more fragile and much more expensive to make commercially. However, as I said, to DIY it would cost well under $100.
Should the ribbon break-up, it is not nearly as tricky to re-ribbon it as ribbon microphone.

The Visaton ribbon size looks like 18mmx63mm. I am thinking of the 15mmx50mm ribbon, so yes, it is smaller.

 

[mhelin]

So these are true ribbons:

http://www.transmissionaudio.com/

Almost full-range ribbons?

Btw. They are looking for test pilots, anyone with a studio over there?

Nice square wave response!

Edit:
Is this design diyable? What I can see he (Bengtsson) doesn't use transformer but connects multiple diaphragms in series to get the resistance an amplier wants to see. It's also told in the patent description that the sound from these multiple diaphragms integrates well in short distance. They also must be less expensive than usual ribbons which have to use the transformer.

 

[Marik]

After long consideration of everything, finally I desided on drivers. They are Eton 5-880 Hexacone woofers (right now on sale for around $80 each), and Seas 27TDFC/TV (H1210) shielded textile domes ($ 29.50). $220 for drivers + shipping... not that bad.

The Etons supposed to be exceptional and are used in Adams monitors. For a long time I was considering going for Seas Excel line, but although they are pretty much the best available (twice the price of Etons), it seems the crossover might be problematic and need a notch filter.

The only problem with Etons is they are not shielded, so I will need to find right magnets and caps.

From what I have gathered these Seas tweeters are supposedly better than Millenium ones, for third of the price and are shielded.

So next week is time to order things.

 

[barefoot]

[quote author="Marik"]After long consideration of everything, finally I desided on drivers. They are Eton 5-880 Hexacone woofers (right now on sale for around $80 each), and Seas 27TDFC/TV (H1210) shielded textile domes ($ 29.50). $220 for drivers + shipping... not that bad. [/quote]

I've tried to build speakers with the 7" Eton's and I wasn't very happy with them. I don't like the break-up characteristics of Kevlar cones. Maybe the 5" driver will be easier to use. In any case, Eton has mediocre motor and spider technology. Nothing to write home about IMO.

Did you look at the 5" Scan-Speak Revelator? Seas also has a new 5" Excel driver with a polymer coated paper cone and an aluminum phase plug - looks pretty sweet. I'm going to try it out.


Thomas

 

[Marik]

[quote author="barefoot"]
I've tried to build speakers with the 7" Eton's and I wasn't very happy with them. I don't like the break-up characteristics of Kevlar cones. Maybe the 5" driver will be easier to use. In any case, Eton has mediocre motor and spider technology. Nothing to write home about IMO.

Did you look at the 5" Scan-Speak Revelator? Seas also has a new 5" Excel driver with a polymer coated paper cone and an aluminum phase plug - looks pretty sweet. I'm going to try it out.


Thomas[/quote]

Thomas,

I appreciate your comments. Yes, I looked at those drivers. However, I'd like to stay with harder cones, as generally they are more detailed than paper ones and IMO, more suitable for studio monitors use. Besides Seas Excells I was also looking at L15 aluminums, which in performance are very close to Excells, but easier to work with.

Here are the actual graths for Eton 5-880:

The first resonance falls at right above 3K. I am going to use bi-amping and variable electronic 4th order crossover. To me it seems it would be easy to cross.


Any comments?

 

[chrissugar]

Thomas,

What speakers you consider of very good quality. I'm interested to know about speakers that you had the chance to use and listen to.
Thanks.

chrissugar

 

[Giganova]

I'm with barefoot and vote for the Madison:

http://www.madisound.com/bigavkit.html

The room for improvement could make this a great project since it starts at a very modest budget and everyone can improve as they see fot and we could compare our results.

I'm in!

 

[mhelin]

I don't about the Etons, but I recently heard a new Amphion speaker (Ion) which uses Seas H1216 5" latex coated paper cone woofer, it was really nice together with the custom (similar to Audax used to make) metal dome (titanium coater aluminium) tweeter:

The H1217 is it's 6.5" version, it looks like this:

I would absolutely use the MCM H-65 waveguide (zaphaudio.com link above) with the tweeter if you are going to build a passive crossover. There aren't practically any other ways to control the directivity at the crossover frequency and besides the tweeter and woofer should be time aligned (tweeter moved a little bit backwards).

Aluminium cone or kevlar are little bit difficult because of the breakup modes, you should really use at least 24 dB/oct or steeper crossover with them (something like a BruteFIR based PC crossover might be fine).

 

[burdij]

I drove out to Madisound on Friday looking for some tweeters to replace blown units in an old pair of KOSS CM523s (two way system with 1" dome tweeter and a 10 inch woofer and cone phaser) and talked with Larry. He recommended the SEAS H1189 (27TDFC) textile dome units for 30.80 each. I put them in the cabinets using the original crossovers. That didn't sound very good and there was still considerable midrange boominess. Larry recommended this unit because it has a small chamber in back of the diaphram which reduces the resonance below 1KHZ. Remembering that, I modified the wiring to allow bi-amp drive and using an old Heathkit 40W transistor amp for the tweeters, a Marantz 80W transistor amp, and a Bheringer CM2103 crossover adjusted to 2KHZ (original crossover was above 3.5KHZ) I now have one of the most transparent (and loud) monitor systems I have ever had, only surpassed by a bi-amped pair of A7s I once owned (when I could still lift them).

 

[Marik]

[quote author="Giganova"]I'm with barefoot and vote for the Madison:

http://www.madisound.com/bigavkit.html

The room for improvement could make this a great project since it starts at a very modest budget and everyone can improve as they see fot and we could compare our results.

I'm in![/quote]

This is a D'Appolito. Personally, considering very big range of opinions on using them as studio monitors, I wouldn't think about them until I have chance to audition in my enviroment.

The second consideration is purely financial. Instead of paying for two sets per channel for midwoofers, getting one but of a higher quality more expensive driver might be a better option.

If you want to go safe, avoid box building and port tuning, and without getting into a hell of crossover breadboarding and fine tuning, then kit like this:

http://www.madisound.com/vifastudio.html

with Vifa PL woofer might be a very good option.

Or some well documented designes from Zaph might be excellent and will be much cheaper.

[quote author="mhelin"]

I would absolutely use the MCM H-65 waveguide (zaphaudio.com link above) with the tweeter if you are going to build a passive crossover. There aren't practically any other ways to control the directivity at the crossover frequency and besides the tweeter and woofer should be time aligned (tweeter moved a little bit backwards).

Aluminium cone or kevlar are little bit difficult because of the breakup modes, you should really use at least 24 dB/oct or steeper crossover with them (something like a BruteFIR based PC crossover might be fine).[/quote]

Thanks for the MCM pointer. I was trying to find waveguide without luck:

LINK

For Aluminums or Kevlars, besides higher order crossovers you might need to use a notch filter, as well. This is one of the reasons I want to use biamping with electronic crossover (at least for now).

[quote author="burdij"]
original crossover was above 3.5KHZ .[/quote]

The 3.5K point sounds very high for 10" woofer.