Speakers measurement..need your advise !

[kanaddict]

Hi all,

At the beginning of last winter I ended my first speakers project. They are MTM tower using Vifa P17wj mid/woofer and McIntosh/Phillips NOS tweeters. I was very satisfy with the way they sound & had good comments from two friends who like me are music lover. But, I was wondering if the frequency response was realy close to be flat. So I build Eric W. preamp II, a jig and a DIY microphone. At first I try to take some measure using only the power from my SBlive card, but it was not working well. So, I tryed with Jolida Integrated amp. (hybrid) and it look like it work fine. I would like to know if there is a way to know if my setup is working well ? I want to be sure that what I see on a frequency response graph is "about" what my speakers deliver.

Take a look at the measurement I made and let me know what you think !

Best regards,

Daniel

An externally hosted image should be here but it was not working when we last tested it.

 

[Andypairo]

What about posting photos of your babies?

Are they BR, sealed, TL or somethng else?

Measurements don't look bad at all, but care must be taken to avoid reflections, resonances and other odd things that might spoil things out.

More advice will surely come from measurement experts

Cheers

Andrea

 

[kanaddict]

Hi Andrea,

What about posting photos of your babies?

Her are the pictures :

http://www.oricom.ca/d_bouchard/fredon2.jpg
http://www.oricom.ca/d_bouchard/fredon1.jpg

Are they BR, sealed, TL or somethng else?

They are double BR.

Measurements don't look bad at all, but care must be taken to avoid reflections, resonances and other odd things that might spoil things out.

I don't have a dedicated room for measurement, but I was carefull & my setup was "descent".

More advice will surely come from measurement experts

I hope 😉


Best regards,

Daniel

 

[kelticwizard]

Very, very nice, Dan. 🙂

They are double BR.
Daniel

Just to be certain, you mean a Bass Reflex with two woofers, correct? You don't mean a Double Chamber Bass Reflex, do you?

 

[SY]

I'm no expert, but I've measured a few speakers over the years. Can you describe how you got your frequency response curve? Is it a spliced 1m plus nearfield? Averaged over positions? Excitation method? Etc, etc, etc?

 

[qwad]

Hmmmm, not bad, for a first effort but if you want a more accurate pic of what your speakers are really doing the 8dB resolution of your graph is a bit high, you should try 5 dB to really see where the problems if any really lie...🙂 oh, and switch any smoothing out altogether because by combinig the two it is possible to make even the worst speaker design look kosher, which is what a lot of the smaller manufacturers do with their published graphs as far as dedicated rooms for measurement is concerened don't worry too much, as long as you have plenty of soft furnishings in the room and you have the speaker elevated off the floor about a ft and then you can measure away to your hearts delight, of course the results are open to interpretation, you should study published graphs in magazine reviews etc, and in your mind try and work out why certain things are the way they are, for instance that dip at just below 2000 Hz is probably a woofer characteristic which you can confirm by using the manufacturers graph as a reference apart from that it is quit good, well done!, but the response at the high end, 10K where the tweetershows why I personally don't like phillips tweeters... there are better ones to be had for not much money off ebaycheers and keep plugging away😀

 

[kanaddict]

I'm no expert, but I've measured a few speakers over the years. Can you describe how you got your frequency response curve? Is it a spliced 1m plus nearfield? Averaged over positions? Excitation method? Etc, etc, etc?

Hi,
it was only at 1m. I have done the near field & it look very good to me....it's why I didn't posted it

😉


here are some print screen that will give you a good idea of my setting. I realised that even if the label are in french, you'll be able to understand what is what ;o)

http://www.oricom.ca/d_bouchard/preference_mesure.jpg

http://www.oricom.ca/d_bouchard/preference_marqueur.jpg

Let me know if there is other parameters you need !

Regards,

Daniel

 

[kanaddict]

Very, very nice, Dan. 🙂

Just to be certain, you mean a Bass Reflex with two woofers, correct? You don't mean a Double Chamber Bass Reflex, do you?

Thank you !

Yes, I mean Double Chamber Bass Reflex 😉

See :

http://www.oricom.ca/d_bouchard/hp_2.jpg


Regards,

Daniel

 

[Andypairo]

How did you calculate the volumes of the chambers? Which SW did you use?

I am interested since I own 4 Vifa P17 and I'm still unsure on the way to mount them.

How do they behave in the bass-midbass region?

Cheers

Andrea

 

[SY]

Daniel, reading French is no problem.

Is this an average of several positions or just one?

 

[kanaddict]

Daniel, reading French is no problem.

🙂

Is this an average of several positions or just one?

This is just one measurement took directly on axis. I'm new to speakers measurement & didn't knew that it's possible to do an average of several positions. I use Speakers WorkShop, but still don't know much about it capacity. I have not yet try to combine two measures (nearfield + 1m) to get the full SPL curve 🙁
The near field measure I took are very close to Vifa specification. My on axis graph look like the off axis measures of Vifa. The difference is due to the x-over point (it make the woofer rolloff earlier).


RegardS

Daniel

 

[SY]

Well, I'd take a look at the 30, 45, and 60 degree off axis responses to see if your crossover is doing what you hope. A neat trick is to use the on axis response as the reference, and ratio each of the off axis measurements to it.

 

[claudio]

Hello Daniel,
why did you use a low resolution (8192) in Speaker Workshop? Wouldn't be better, for precision, to increase it to 256K?

I would like to see the near field response: does it have the dip typical of DCR ? (even if your DCR is not the Weems/Augspurger's one).

Regards

Claudio

 

[kanaddict]

How did you calculate the volumes of the chambers?

The total volumes is calculated just like a normal vented box. The volume of each chambers was calculated by someone who helped me. They are 20 l (top) and 27 (bottom). If you are interested, I could provide you the diameter of the 3 holes (I have to look for it when at home). I can also ask the person who helped me how he made the calculation. I think it's a simple factor, but I'm not sure.

[Which SW did you use?

SW 😕 Speakers Worshop ?

[I am interested since I own 4 Vifa P17 and I'm still unsure on the way to mount them.

They are great drivers for the price. You can do a lot of interesting design with them. Someone used them in a MTM monitors (sealed) with some SS tweeters (9500 ). I have the x-over deesign...just in case I would like to try something different
😉

How do they behave in the bass-midbass region? [/B]

The bass and midbass is very good. My MTM (the Fredon) have a lot of punch, very alive. They make you move on your chair. I have some very good B&W floorstanding and they don't bring life to music like the Fredon.....But, it's a matter of taste !

Regards,

Daniel

 

[kanaddict]

Hello Daniel,
why did you use a low resolution (8192) in Speaker Workshop? Wouldn't be better, for precision, to increase it to 256K?

Hello Claudio. You will find my answer stupid.....simply because the little SW guide I have say that the resolution should be set at 8192. I don't know if increasing it to 256K would give better results. Anybody can answer this question 😕

I would like to see the near field response: does it have the dip typical of DCR ? (even if your DCR is not the Weems/Augspurger's one).

You asked for it 😉

http://www.oricom.ca/d_bouchard/fre_06062003d_proche2.jpg


Let me know what you think !

Regards,

Daniel

 

[kanaddict]

Well, I'd take a look at the 30, 45, and 60 degree off axis responses to see if your crossover is doing what you hope. A neat trick is to use the on axis response as the reference, and ratio each of the off axis measurements to it.

Next time I'll install my measurement setup, I'll do the off axis measurement. I don't understand much how this will tell me if my x-over do right 😕 Could you please try to give me some bsic explanation. I'm sure that I'm not the only one who will learn when reading it

Thanks !

 

[kanaddict]

Hmmmm, not bad, for a first effort but if you want a more accurate pic of what your speakers are really doing the 8dB resolution of your graph is a bit high, you should try 5 dB to really see where the problems if any really lie...🙂

Happy to hear that it's not bad 😉
I'll try to use the 5dB resolution next time....if I'm able to find how to change it. The 8 dB setting was set by SW.

oh, and switch any smoothing out altogether because by combinig the two it is possible to make even the worst speaker design look kosher, which is what a lot of the smaller manufacturers do with their published graphs

Since I don't plane to sale my speakers, I'll avoid the use of the smooting function.

as far as dedicated rooms for measurement is concerened don't worry too much, as long as you have plenty of soft furnishings in the room and you have the speaker elevated off the floor about a ft and then you can measure away to your hearts delight, of course the results are open to interpretation,

I don't have soft furnishings in the room. I will have to cover a few things to see if it could help. My speaker was put on a stand to increase the distance between the bottom woofer and the ground.

but the response at the high end, 10K where the tweetershows why I personally don't like phillips tweeters... there are better ones to be had for not much money off ebaycheers and keep plugging away😀 [/B]

The Phillips were very cheap (12.5 USD each) and I can say that for the price I like their sound a lot. Roger who worked for McIntosh for many years, told me that these were some of the best tweeters McIntosh used. That said, you are right, there is some great deal on e-bay.

My first project was a learning one and my next project will be a better one. Probably Some Neo3 ribbon tweeters + Neo8 ribbon (mid) + Focal 10' woofer. I still strugle with myself. .....Doing this big project (100% DIY) or building a design made by someone who have a very good reputation (Ribbon tweeters + Vifa XT woofers) 😕

Thanks !

 

[SY]

Next time I'll install my measurement setup, I'll do the off axis measurement. I don't understand much how this will tell me if my x-over do right Could you please try to give me some bsic explanation.

Whew! That could fill volumes, but I'll try to condense it to a few sentences. Your crossover and the physical configuration of the drivers (diameter, placement, baffle geometry) will all affect how the off-axis response relates to on-axis response. There are various schools of thought on what that relationship should be, the two most common being that the response should be relatively constant off-axis, or show a smooth HF rolloff. But whichever you choose, you need to know whether or not you're hitting that target.

It's pretty well-established that if the horizontal polar pattern is ragged, the spectral balance and imaging will be problematic unless radical room treatment is done. A common flaw is a two-way that has an off axis response that droops toward the top of the woofer passband, then suddenly rises up as the wider-dispersion tweeter cuts in.

 

[kelticwizard]

Dan:

I am not sure that is a Double chamber Bass Reflex that you have. I think it might be a brace which is tuned high enough to prevent interfering with the bass.

Judging from the front of the speaker, I am guessing that you have three 4" diameter holes there, or thereabouts. That is because the diameter of the holes is about half the size of the front baffle.

Three 4" diameter holes tune a 27 liter box, (your larger chamber) to 165 Hz. I don't think such tuning will yield satisfactory results.

Just guessing here. I ran the tuning for a 27 liter box with three 4" diameter holes and for that figure on Win ISD. 165 Hz was the calculation.

I don't think you have a Doulbe Chamber Reflex. but I could be wrong. 🙂

 

[claudio]

Hello Claudio. You will find my answer stupid.....simply because the little SW guide I have say that the resolution should be set at 8192. I don't know if increasing it to 256K would give better results. Anybody can answer this question

Hi Dan,
SW help was written many years ago, when Pentium CPU was the fastet one, and a 256 k dimension could give some headache to the computer. If you look at the bottom of the options window, you will see that the RESULTAT DE PRECISION is +/- 5.86 Hz, while using the 256k you will get 0.18 Hz: That's a big improvement!

About the near field response, how did you measure it since you use 2 woofers? Did you took the single drive response and then jointed the 2 ? The useful response is till 800 Hz, since it is a 6,5 woofer; the near field response you show doesn't look too good to me: did you add the port response?

Regards

Claudio