My first horn experiments

[qwad]

LOOKS AS THOUGH YOU ARE USING A PULSE MEASURING SYSTEM WHICH NEEDS TO BE GATED SO THAT YOU ELIMINATE THE ROOM REFLECTIONS ALS O TRY MEASURING 200 Hz up... the top end of the tweeters looks smooth enough, try a little smoothing say 1/3 oct if the programme allows for it, forget about blaxkman/hamming et all for you and me and everybody else these have no relevance ANYWAY WELL DONE CHEERS, TJB

[till]

qwad, the system produces a MLP (whatever) signal throug soundcard. It sound like (white?)noise. It mesures this through soundcard an mic and gives me plots like waterfall, this one shown, pulse response. I don´t know to much about most of these... Would it help to measure outside in free air?



Ok, i had more time now to listen, and i did not much more than listening the last days. (as you may have expected) I'm no poet so please don't expect poems or so about how it sound. Direct, impressive, sometimes brutal. I never before experienced a speaker produceing these kind of hights, but i did not listen to too much different speakers in my live. As expected sound is very directional. And room is to small. I played with xover - i'm biamping with Zen lite (ca 1,5W at the moment) and V4 now. I tested some combination of high/low pass parts and ended without calculation only by sound with 22uF Film in series with TD2001 and big???H diy coil for TL1601a. Higher order made me mad. I know this is very unprofessional and to write this here has the risk of some flames, but i don't know how to do better right now.

Of course i tried out to play loud - and sound is very clean and clear at levels all the neighbours are entertained too. No problem to listen in the garden outside or 200m down the street - until they send police.(Did not happen until now) I must try these speakers outside some day, an event in the forrest or so, filling a valley with music.

[Elso Kwak]

Hi till,
For a hornloudspeaker the response is not bad at all I believe. Peak-trough values of 10- 15 dB are quite normal when you measure with a sweeped sinus.
Don B. Kiele has shown that you can do a nearfield measurement, thus not having any problems from the acoustics of the room.
Good luck!

[OMNIFEX]

Hey Till!

Things look mighty good from over here.


My question pertains to the A7 copies.


I've noticed your using a fixed port, where,
as the original gave you the option to
adjust the port to the desired frequency.
(Removable panels)

Have you selected the desired frequency,
why you chosed to make a fixed port?

Also, are you using the original Altec woofer,
or, one compatable in the TS Parameters?

Most nowadays woofers reflect reflex/sealed
enclosures, and not straight horns, so you
really need a woofer that will load in the
horns throat, and, not only the enclosure.

You should check out the unofficial Altec site,
to make sure your using the right driver for
that box.


Regards,

[hancock]

Originally posted by till
qwad, the system produces a MLP (whatever) signal throug soundcard. It sound like (white?)noise. It mesures this through soundcard an mic and gives me plots like waterfall, this one shown, pulse response. I don´t know to much about most of these... Would it help to measure outside in free air?

Hi Till,

That's probably MLS (minimum length sequences). An MLS signal is designed to look like white noise, but actually is deterministic. The way MLS works is you input the MLS signal to the device under test and record the output. You then do an autocorrelation between the input signal and the output signal. The autocorellation picks out just the component of the recorded signal that is correlated with the input signal, so noises like your wife yelling at you to turn the damn thing down get thrown out (or at least suppressed by quite a bit). The end result of that mathematical operation is an impulse response, which shows the response of your system to a unit impulse (a one followed by lots of zeros). The impulse response shows first the direct sound coming directly from your speakers and then a combination of reflected sound and direct sound. When you convert the impulse response to a frequency response, you do a math operation called an FFT. If you want to see the frequency response of only your speakers without the room response, you have do the FFT on just the first part of the impulse response, before any reflections. If you look at the impulse response you can probably see a smaller second spike that is the first reflection. You can get a rough idea of where that first reflection is by figuring out the distance the sound has to travel from the speaker to the first reflective surface (usually the floor) and then to the mic. Since sound travels 344 m/s divide the distance of the first reflection (in meters) by 344 and then multiply by the sample rate (probably 48000 samples per second or 44100). Now if you do the FFT always cutting out the first reflection, you can see what your speakers are really doing.

How about posting the impulse response you measured for the TAD's? I've got 2001s as well and I'm curious to see how smooth a response you get with your beautiful horns.

Best regards, John

[till]

John,
I will post all this measurement stuff when i manage do get some good results. To make things comparable, what FFT resolution is to choose? There are parameters like signal trigger, "FFT Fenster Anfang - n Messerte vor Impulsantwort max.", low pass. Else what distance speaker to mikrophone?

If you want diagramms before i post them i need your email.

I´m interested in what kind of system you use the 2001s.


OMNIFEX,
The size for port was estimated for something inbetween 40 - 45Hz tuning.
Its no problem for me to change it to a variable port but at the moment i have no base to choose another size. As you could see above in this thread or the threads leading to this -
recommendation for ZEN compatible speakers
speaker companies
There was a process leading to this kind of system. Thanks to Kent / cyclotronguy he told me what to do and what speakers to use. See this hint
http://www.diyaudio.com/forums/showt...067#post120067
Kent told me TL1601a would be the right speaker to use in the altec clones.
http://www.pioneerelectronics.com/Pi...45,873,00.html


Elso, Thank You.

Till

[OMNIFEX]

till,


That seems like a fine choice. It will give you the best of
both worlds.


If you ever decide to build a straight 15 inch sealed horn,
the TL 1602 will make some beautiful noise


Best Regards,

[phase_accurate]

Hi Till

First of all, your horns look higly professional (and marketable as well).

I also did measurements on my Mangers with Hobbybox. The in-room response shows a lot of peaks and dips as well. If you want to know the response of your mid/high horn alone, take a reasonably short measurement window as mentioned by John and measure them close-up. I bet it would look even better than your current masurement (which is in no way bad !!).
The drop at the higher end could be exagerated by the microphone's and soundcard's FR. There is a function within Hobbybox to compensate for the FR deviations by the soundcard/amp combination. If you know the properties of your microphone you can include them as well.

Now that the days are getting warmer I will go and measure my boxes outside to see if there is any difference (which I assume there is ! ¨).


Regards

Charles

[till]

First of all, your horns look higly professional (and marketable as well).

Thank You. I will build some more for improvements in production process and because i want to make a larger set for the TD2001 (250 Hz) . These are not perfect, i know how to do somethings better now, as mostly if you build somethings you know how to make better afterwards. I don´t think they are marketable because i need high quality (expensive) wood, use some not so cheap tools, and a lot of time. So i don´t want to go in price competition to Bruce Edgars saladbowls. If there is interest maybe i will build on request and trade with with other audio enthusiasts. But not to make a real business.

Measurement: I can´t compensate my microphones nonlinearity because its diy, with diy preamp. I will try measure outside only the TD2001 (big box to heavy to carry outside alone). What distance to speaker to choose? 1m ?

[phase_accurate]

Hi Till

I'd go even closer. Regarding the compensation of the measurement equipment: At least the response of the soundcard, power amp and preamp can be compensated quite easily with Hobbybox.

Unfortunately I forgot the respective cables, to setup the compensation measurement, when I did my measurements. Thats why I have the same HF drop on my MSW measurement (which should go higher than 20 kHz) as you see with your horn.

Regards

Charles