WinISD reduce my SPL every time

[oaristys]

Hello,

I try to find what is the best driver for my enclosure. I found many spec sheet and "recommended" enclosure dimension with port size and frequency response on many website.

If i put the specs in WinISD beta, i got a enthusiast curve every-time.

But when i try with WinISD Alpha, the more elaborate, i get about 3 db less efficiency than shpecsheet almost every time. Almost every recommended enclosure is different plot than promised.

Now i have to take an important decision in my project. I have a small thin enclosure and the driver selection will make me decide between one (or two) mini woofer on the back or one bigger woofer on the side. And one woofer will be order over the sea and the other will be buy at few kilometer of my home !

The woofer i'm interested is Silver Flûte W14RC25s04, his 17cm brother and the Monacor SP-60.

The lautsprechershop website have many recommended enclosure with specs and few of them satisfied my need :

490CuIn enclosure if woofer is side mount.
574CuIn enclosure if woofer is back mount. (because i don't need ¾"thick MDF on side.)

Monacor SP-60 is interesting but i got a gigantic difference between the two winISD version. In elaborate version, i need to use two of them to recover the efficiency loss. But the lower bass gone...

Help ! Thank's !

 

[chris661]

Sounds like you're going to have to ignore efficiency for now.

Use winISD to make a nice response curve, then make sure your amp is powerful enough - you can simply turn it up a bit to counteract the -3dB efficiency.

Chris

 

[Ron E]

Now i have to take an important decision in my project.

WinISD beta basically uses simple equations like you might find at www.diysubwoofers.org WinISD alpha appears to use more complex equations, but I found it slightly different from my reference.

Usually there shouldn't be much difference between the two. Two things that can make a drastic difference are series resistance and VC inductance, because alpha models their effect and beta doesn't. Alpha also actually checks the parameters for consistency where Beta accepts whatever you throw at it.

I'd say try Unibox, but I feel the UI would be somewhat daunting to most newbies. The trouble with all black boxes is that if you don't understand all the knobs and buttons you can get garbage output.

If you want someone to run models for you, you should post parameters or links to them. Preferred parameters would be Fs, Qes, Qms, Vas, Sd, Re, Xmax, Pe, Le

 

[infinia]

WinISD beta uses what sensitivity you give it ie from the data sheet, so if you use vendor data from 2.83 V/m and then compare it to 1W/m it will be a bit optimistic at low frequencies. I would compare the calculated sens. from the alpha version to Unibox and use those. They will be lower using 1W/m at frequencies where the driver impedance is at minima.

 

[oaristys]

Sounds like you're going to have to ignore efficiency for now.

Use winISD to make a nice response curve, then make sure your amp is powerful enough - you can simply turn it up a bit to counteract the -3dB efficiency.

Chris

I understand and thank you for the trick but i have two reason to search another solution.

1. My tube amp is only 35Watt (workin about 20 watts for good sound ?)
2. This driver should be same efficiency as the other in the speaker box. I use a fullrange and search for a woofer to extend the low frequencies in a small 8L stand box.

 

[oaristys]

Re: Re: WinISD reduce my SPL every time

...

If you want someone to run models for you, you should post parameters or links to them. Preferred parameters would be Fs, Qes, Qms, Vas, Sd, Re, Xmax, Pe, Le

Hô ! you are generous 🙂

This is my "finalist" drivers (but probably lot of other can do the job but disqualified because my problem with winISD) I'm open to suggestion here

This speaker is woofer stand enclosure to use with little monitor with Fostex Fe107e (90db)
spec sheet at http://www.solen.ca/pdf/fostex/fe107e.pdf

I will cross over somewhere between 150 and 225hz. The higher is better for power handlinkg but the lower is better for coherency and flatness (i think).


Silver Flûte W14RC25-04 5.5" spec-sheet at http://www.solen.ca/pdf/silver/w14rc2504.pdf

Fs 42.5 Hz
Qes 0.31
Qms 3.70
Vas 19 L
Sd 0.0087 M²
Re (i just have Znom 4 Ohm)
Xmax 3.5 mm
Pe (i just have Power) 60W
Le .52 mH (@ 1khz)

Silver Flûte W17RC38-04 6.5" spec-sheet at http://www.solen.ca/pdf/silver/w17rc3804.pdf

Fs 37.5 Hz
Qes 0.28
Qms 1.52
Vas 25.6 L
Sd 0.0132 M²
Re (i just have Znom 4 Ohm)
Xmax 5 mm
Pe (i just have Power) 60W
Le .53 mH (@ 1khz)

Monacor SP-60/8 ou 4 10cm spec-sheet at http://www.audiophonics.fr/monacor-sp608-p-1197.html

The 8Ohm and 4 Ohm version are identical in specs. I don't know why... And note the SPL is 1w 1m in specsheet instead of 2.88v but WinISD reduce it anyway.

So, when a look to the graph, this driver seems to lost SPL just where i need to cross over it.

Fs 65 Hz
Qes 0.64
Qms 2.10
Vas 6 L
Sd 55cm²
Re 6.5Ohm
Xmax 2 mm
Pe 30W
Le .3 mH

Enclosure
490CuIn enclosure if woofer is side mount.
574CuIn enclosure if woofer is back mount. (because i don't need ¾"thick MDF on side.)

Thank you very much !

 

[infinia]

Don't forget using BSC, the (combined) woofer response will need to be at least 3~5 dB more sensitive unless you want to pad down the FR driver. Compare the sensitivities using the same method near the crossover frequency. Using normal small bass drivers plan on doubling up and the 4 ohms tap on your amp.

 

[infinia]

Re: Re: Re: WinISD reduce my SPL every time

The 8Ohm and 4 Ohm version are identical in specs. I don't know why... And note the SPL is 1w 1m in specsheet instead of 2.88v but WinISD reduce it anyway.

You need to wrap your head around this concept using ohms law.

 

[oaristys]

Re: Re: Re: Re: WinISD reduce my SPL every time

You need to wrap your head around this concept using ohms law.

Sory, i don't understand. I'm french person and i think your phrase use expression we don't use in french

Your precedent message too.

Talk to me like you talk to a 2 years old children and everything will be fine lol 😀


Thank's for your help.

 

[infinia]

Sorry I don't know your experience
do you know Ohm's law ? http://www.the12volt.com/ohm/ohmslaw.asp


Work through some examples converting 1W/m to voltage when the speakers impedance goes from 8 (Re~6.6) to 4 (Re~ 3.3) ohms.
The constances of (Volt/m<=>Watt/m) as a function in change of resistance.

Using constant voltage (ie.).(2.83V/m=1Watt@ 8 ohms) the sensitivity goes to 2X (double) when speaker goes down to 4 ohms. But using constant power 1W/m the sensitivity stays the same (the voltage is reduced to keep the same power).

You really need to get this idea in your head=mind (brain).

 

[oaristys]

Why BSC ?

(Baffle Step Compensation ?)

I talk with Solen guy about this phenomenon and he said doesn't matter in my case. (woofer + fullrange) The woofer will work only below 150hz and the sound is non directional at these frequency range. All the waves are like sphere. He said i can trust the Silver Flûte curve response in my condition.

But i'm scared about buying something for now. I want bass. I just want to buy the more efficient drivers with "60hz f3 if possible" for my speaker enclosure/stand.

For the rest, if i can't go as efficient as my Fostex Full range, i will attenuate it. Just to add the right resistor or L-pad potentiometer before it .. (?)

 

[infinia]

I think the guy at Solen wants to sell a silver flute😉 BSC still applies in your case. Let him propose a solution with crossover diagram and rationale for crossover frequency. and then start a new thread based on that. http://www.solen.ca/import/xvrdesign.htm

 

[oaristys]

I think the guy at Solen want to sell a silver flute😉 BSC still applies in any case. Let him propose a solution with crossover diagram and rational for crossover frequency. and then start a new thread based on that.

Anyway, BSC or not.. what can i do ? I have Fostex 90db effiient full range and just 7-8L enclosure for bass. Nobody here can make a subwoofer efficient enough for the Fostex here. The only solution is to attenuate the Fostex or to use a subwoofer plate amp.

Like another members here, i have to said the Hoffman law ... Seems to be sometime more strict in DIY than market speaker but the only way i have is DIY.

I have choice. Less efficiency - more power - more bass or more efficiency - less power - less bass.

The easy way is to go somhere in the middle. Little less efficiency - same power - easy and affordable crossover - less max spl.

This project is supposed to be simple, pure, light, zen, easy listening, small.

This is why i choose a clear noble cream-white Fostex cone an put two this mat varnish in the white wood cabinet.

And about silver flute... i ask im for it. (the solen guy) I don,t think sell 26$ woofer is a super deal for him lol. Better to sold my an scanspeak or anything over 300$ !

I hear lot of "correct" sounding speaker with just little iron core inductor and polarized capacitor. So i think a small 20gauge air core inductor, a cheap NP capacitor and a potentiometer to adjust at listener taste will be good. And i will be able to place other minimonitor model over it with simple "volume" control

12$ of wood, 50$ woofers. I will not invest 150$ in crossover ! lol not for a 75hz low wattage speaker. ho no. My wife will kill me !

Keep in mind that the sound is already good. Just lack of bass between 150hz.
Thank's.

 

[infinia]

Hoffmans law is real im afraid. Small drivers means low eff. at bass freq. you could double up there or look at some 8" Pioneers or Eminence.

 

[oaristys]

I will stay with Silver Flute and my actual enclosure because the first priority is there.

But this system will be very convenient because i can simply unplug the mini monitor for top quality, use in small space environnement or plus it trough active subwoofer.

I just don,t understand how ProAc can reach as base and as efficiency in a so small enclosure but manufacturer specs is nebulous.

Thank's for you help. I will buy the stock today.

Many resistor to try different L pad (cheapet than potentiometer), zobel and first order cross over with two configuration because the web calculator calculate different capacitor than Solen guy. So i by both value lol.

 

[infinia]

I hear lot of "correct" sounding speaker with just little iron core inductor and polarized capacitor. So i think a small 20gauge air core inductor, a cheap NP capacitor and a potentiometer to adjust at listener taste will be good. And i will be able to place other minimonitor model over it with simple "volume" control

Hmmm I don't think you have good eff, and low crossover frequency with small air core inductors. Inductor Rdc reduces eff. see Unibox with SPICE simulations. Series caps with FW drivers is problematic near Fb.
IMO a small SS/class D amp for an active crossover approach might be the most ZEN thing after all options have been looked at in greater detail.


see http://www.diyaudio.com/forums/showthread.php?s=&threadid=144099

 

[oaristys]

Hmmm I don't think you have good eff, and low crossover frequency with small air core inductors. Inductor Rdc reduces eff. see Unibox with SPICE simulations. Series caps with FW drivers is problematic near Fb.
IMO a small SS/class D amp for an active crossover approach might be the most ZEN thing after all options have been looked at in greater detail.


see http://www.diyaudio.com/forums/showthread.php?s=&threadid=144099

damn.. i have to go and my order is already sent.

I read on net 20gaug air core is capable of 180watt ... i only need between 25 and 50 watts ...

I own Mission 701 good sounding speaker (not as good as 2 000$ hifi speaker of course) and small, verry small parts on it.

i know is better to buy 800$ parts to do a speaker than 200$ parts. I need to stop somewhere !

The amplifier for woofer is minimum 145$CAD (with smal NHT + 10$ account + shipping + taxes) or a lot more locally.

I calculate all the fees. A true budget is not like some website where you have the total of the drivers and crossover. I need to calculate all the fees, tools usage (like router bit), gas, shipping, glue, screw, soldering, plugs, wire, taxes and taxes over shipping.

And my wife and me don't want to place 2 others peace in sound system. (crossover and amplifier). Or don't want to have 2 speaker cable + 2 power cable. No the only active solution i can see is active subwoofer hidden somewhere and leaner stand. Just one amp, one sub, one boxe. But sound like computer speaker. I don't like 2.1 system.

 

[oaristys]

Hmmm I don't think you have good eff, and low crossover frequency with small air core inductors. Inductor Rdc reduces eff. see Unibox with SPICE simulations. Series caps with FW drivers is problematic near Fb.

see http://www.diyaudio.com/forums/showthread.php?s=&threadid=144099

for the inductor, is it not the mH value decide for the result and the gauge for the resistance ?

If yes, the mH is the one of web calculator and Solen guy. For the resistance, it's at some detail the same as bigger and my wattage is low ...

I think i have a lot to learn ... but after many days long in this computer i have to do something else. 9hrs a day on computer is not a life ! I want to build some speaker, not waste all my adult live in computer and internet. I think i'm turnin garround with this project.

 

[Helmuth]

I used winsid and made boxes and did measure them.

I can say trust winisd and be critical about the claims off speaker manufacturers.

When the TS-parameter are correct then the calculated spl by winisd is very accurate.

And also be keen on the X-max cone extrusion by high levels.

 

[oaristys]

i went to Solen yesterday and finally i bug higher gauge for inductor. I talk about efficiency and inductor resistance with him and an inductor with no efficiency reduction will cost me a legs and will be gigantic (compared to speaker size). So i go for middle solution. I will have between 1 and 1.5 db drop instead of 2.5-3db..


About WinISD, except rare case, in cannot save the new drivers settings because winISD math don't fit with all drivers specs. How can i get precise simulation if i have to delete randomly few variable to let it guest the real value ?

The software said one or more value are wrong but not which one so maybe i delete the only good one and let it calculate from others wrong.