LM3875 and speaker impedance questions

[whiz115]

no 1. the thing that u didnt perceived a difference in sound level makes me think that is ur amps fault for that noise and that it was in clipping. try at low levels does the sound still exist?

it's the first time i heard that sound with that amp...it's the first time even in my life that i have heard that sound! it's an alien sound! my speakers receive alien frequencies! (kidding ok?...it's exactly what i said before)

no 2. try to increase the resistor value from the zobel until the mid range seems ok for u, after u have tried with the 4ohm tw resistor.

something tells me that if i'll do that i will not see any difference with and without zobel network....the difference that i heard was only moderate...
but it's not my imagination that there was difference and not a good one! if you can accept it....

Two more things i want to know, where do u have this xover from? The speakers are store products or custom made?

the speakers are not 100% commercial the original drivers has been changed with one simular in size and wattage...ok you got me! now you're probably going to say that the quality problem is from the combination what if i say that few years ago with the other drivers the sound was also "bad"?

the 2nd order crossover are bought when i burn my first drivers.

 

[hentai]

no1 again i ask: at low levels does the sound still exist?

no2 the changes that u should get with a zobel connected is a loss in high freq reproduction from the woofer whitch u described as a loss in voice detail. The role of the zobel isnt to shape the freq response but the impedance of the woofer. As i said before because of the Le of the voice coil the woofer doesnt have a linear impedance caracteristic therefore it really messes up the xover whitch is designed for a specific impedance. Not only the woofer's impedance changes with freq it also changes with the power it's receiving because the voice coil gets hot at high levels. The zobel keeps this impedance at a steady value allowing the xover to do its job. That loss in voice detail u described should be compensated by the tw. But i recommended using a 8 or 10 ohm not lower in ur zobel. Try it with 10ohm this way u increase the xover freq cut point to eliminate the gap. OR if ur happy with the sound without it dont bother. just put the xover connect the drivers and enjoy the music.

hehe as has been stated before the quality problem is the drivers. The unpleasent bass u descried before its from the enclousure.

 

[whiz115]

1) no...that scratching noice (if you mean that) it was at high volume level
and i can hear it only from the tweeter. (probably the frequency was too high 10KHz to hear something from the woofer) The problem while i'm using the ~8ohm resistor disappears.

2)ok...i'll try it again in case i don't remember correct and i had connected the zobel network while the tw resistor was 8ohms...

i'm going to use the ~8ohm resistors with the 33uF capacitor and i'll try 2.2ohm for the tw and 4.7ohm

 

[hentai]

the noise disappears when u use the 8 ohms because the amp will no longer see a load thats smaller than 8ohm so the amp functions properly. When u used 2.2 total impedance whould be around 6 ohms and that could cause problems at full output

 

[whiz115]

the noise disappears when u use the 8 ohms because the amp will no longer see a load thats smaller than 8ohm so the amp functions properly. When u used 2.2 total impedance whould be around 6 ohms and that could cause problems at full output

ok... i used the ~8ohm resistors with the 33uF capacitor and 2.2ohm for the tw. The result was not noticeable to my ears...i had the speaker in frond of my face and i didn't heard any difference while the zobel was connected or disconnected. unlike the last time that the difference was obvious with the zobel connected.

This means that the equalization of the impedance was a success?

P.S the tw that did that scratching noise, is dead, up to yesterday the sound from the tw was very low and today i finished it while i was experimenting to see what's wrong. (no problem i found it in a local store very cheap...about 15E)

 

[hentai]

Hi whiz, i told u not to use the amp at full output...
when testing the sound of a speaker u must test the on-axis response as well as the off-axis response. Room acoustics play an important role so u must also find a good position. First listen them staying in their center at about 1 m and note ur impresions. Then move left or right, therefore changing the angle to about 60degrees from the center and note the impresion. The sound shouldnt change much in color.
So lets look back a little: When u connected the zobel and the tw had 8 resistor the speakers had a loss in mids. when u connected the zobel and the tw had 2.2 resistor u havent noticed a change. Correct?
The egalisation of the woofer's impedance for that xover is achieved when the sound between drivers is as balanced as posible.
The next thing u should try if ur willing to continue is to change the woofers zobel resistor to 4.7. Leave the tw's 2.2 for its necesary for the higher SPL.
Using the 4.7 ohms should lower the woofers xo point so mb a better match can be achieved.

 

[whiz115]

So lets look back a little: When u connected the zobel and the tw had 8 resistor the speakers had a loss in mids. when u connected the zobel and the tw had 2.2 resistor u havent noticed a change. Correct?

correct.

The next thing u should try if ur willing to continue is to change the woofers zobel resistor to 4.7. Leave the tw's 2.2 for its necesary for the higher SPL.
Using the 4.7 ohms should lower the woofers xo point so mb a better match can be achieved.

ok i'll try that.

P.S i already playing with the angle when i'm testing them etc...when i did the tone test i've notice that the speaker behaves better in some specific low frequencies when i block the bass reflex hole. Is there any way to calculate the correct dimensions of the cabinet for my speakers? and the exact bass reflex tube i need? i know that few improvements are possible but at least i want to know how much is that difference.

 

[audioguru]

Half-decent woofers have TS detailed spec's which you plug into software or a graph that shows the best enclosure size, whether a port (reflex tube) should be used and its diameter and length.

If the enclosure doesn't match the spec's of the woofer then the bass will be missing or will be boomy.

 

[whiz115]

Half-decent woofers have TS detailed spec's which you plug into software or a graph that shows the best enclosure size, whether a port (reflex tube) should be used and its diameter and length.

yes they have...check first pages

If the enclosure doesn't match the spec's of the woofer then the bass will be missing or will be boomy.

probably they don't...
but one day i'll buy new ones...up to that day i believe my knowledge will be
in higher level so i could make a better choice. All this story about my
speakers not only helps me improve them...but it also helps me understand electronics!

 

[audioguru]

A few important Thiele-Small parameters are missing with your el-cheapo woofer: Vas is the compliance volume, Qm is the mechanical Q and Qe is the electrical Q. It stated only the resonant frequency. If all the parameters are known (the woofer can be tested to find out the values but it is time-consuming) then software can be used to determine a suitable enclosure, or to tune the vent or cover it on the enclosure you have.

 

[whiz115]

i have these parameters too... nothing is missing..

Qts 0,52
Qes 0,65
Qms 2,63
Vas 90
B*1 4,8

overall diameter 195mm

bla bla bla bla
bla blaaaaaaaaaaaaaaaaaaa

tell me which program to download so i can enter these parameters...

 

[hentai]

Hi,
the Vas seems so big for a 8 incher... this will need high internal volume for your speaker.
Also the Efficency bandwidth product which is Fs/Qes has a value of about 52 which shows that the woofer will probably work better in a closed box. In a bass reflex of small dimensions will sound boomy.
What's the internal volume of ur speakers?
As for programs go for WinISD pro or LspCad.

 

[audioguru]

I used Boxplot program on my old computer.
I picked up some thrown away finished speakers in nice looking enclosures that were way too small for the garbage 8" drivers they had. Then I bought pretty good woofers and tweeters for them, used Boxplot to make the vent tune to a frequency too low then used tuned bass-boost to make them sound perfect.

 

[hentai]

Interesting, but if u tune the vent too low wouldnt that give u a high group delay? especially in small boxes... which will make ur bass very loooong?
For the eqalization u used an active eq?

 

[audioguru]

My tuned bass-boost circuits use a 2nd-order equal-values Sallen and Key highpass filter with too much positive feedback. The boost is 10dB at 38Hz and the small speakers sound like a much bigger speakers.
Years ago I did the same circuit for a 4" sub-woofer that I played at the beach.
Recently, Silicon chip magazine has a very similar Bass Extender project.

 

[hentai]

Neat circuit thx. In IB speakers this is a good improvement, but in bass reflex i belive u lose the tightness in bass because of the lag. I reckon that is more or less noticeable depending on the type of program u feed it with.

 

[audioguru]

Because the enclosure is too small, the low frequencies are reduced in level. When the bass boost increases the level of the low frequencies to normal then the delay caused by the tuned circuit is also boosted to normal. If the bass level is increased to more than normal then the delay would be very noticeable.

The circuit also adds low frequencies to my sealed speakers that have an fc of 60Hz without the circuit and 38Hz with the circuit.

 

[hentai]

Hmm... it is my opinion that GD doesnt reduce with signal boost on the contrary. But this depends on many things as well as the woofer itself... on very small boxes the freq where GD has its peak, may be outside the speakers range so it wont be noticeable. Mb i'll give it a try in my next project and see how it sounds, but small boxes may be the important factor that makes it work.

 

[whiz115]

What's the internal volume of ur speakers?

Do you mean how many liters they are? i don't know and it is kinda difficult
because i must include material thickness...and i don't know how to calculate all these things

P.S about the Vas datasheet says Vas= 90 dm3


ooops....sorry!!! overall diameter is 206mm and bolt circle diameter is 195mm!

now the Vas is correct?

 

[hentai]

i checed the site and it seems Vas is correct if we are to trust them.
1 dm3= 1 liter.

Do you mean how many liters they are? i don't know and it is kinda difficult
because i must include material thickness...and i don't know how to calculate all these things

give the exterior dimensions and the material thickness

those are pretty much the standard dimensions for an 8 incher. That Vas i see in 10 inchers. Anyway is the cone really deep?