How to make a speaker more forward sounding

andy2

Member
2004-11-17 4:46 pm
I've just finished building Troels TJL speaker yesterday. The sound is slightly laid back, and on some recordings, very distant.
Here is the crappy picture so far :
http://www.geocities.com/hai_vu1/TJLBoth.jpg

I am thinking of modifying the xover to make the sound slightly more forward which better suits low-volume listening. More treble would help too. Here is the xover of the TJL:
http://home1.stofanet.dk/troels.gravesen/index_b/TJL_loudspeaker.htm
It is near the bottom of the document.

On his website, the measurement show a slight dip (1db or so) in the 1k - 3k area. I hope to bring it up to about +1db in that region. Can someone suggest a change in the xover?

On the plus side, the bass is very good -- tight, powerful, and very controlled. The treble is too relaxed though but I am playing around with the series resistor
 

Ron E

Member
2002-06-27 10:41 pm
USA, MN
Assuming everything is wired correctly, the easiest way to add treble would be to reduce the value of the 4.7 ohm resistor in the tweeter L-Pad. This will raise the whole tweeter range. It looks to be a reasonably well designed crossover, I would try that before making any more extensive mods.
 
Not an easy feat, but in my (short) experience, lowering the order of the XO made a HUGE difference in brigning the soundstage forward. When I designed my last set of speakers, I had started with a 4th order on woofer and 2nd order on tweeter, and as I went along tweaking my filters, I remodeled them to use less and less componants, and the sound became clearer and clearer as well as more warm and forward.

But you have to be willing to model your own filters for that...
 

SY

diyAudio Moderator Emeritus
2002-10-24 10:19 pm
Chicagoland
www.SYclotron.com
Only a little more, since I don't have Calsod in front of me at the moment. Basically, I'm suggesting that you create a shelf in the frequency response by lowering the level of the top octave uniformly by 2dB with respect to the octave just below that, along with bringing up the entire tweeter level by 2dB. The net result is that you bring the trough from 5K-10K up a touch; this will make the sound more immediate and a bit brighter.

The easiest way to do that is with the network I suggested. If you think about it, at low frequencies, the C in the series RC (paralleled with the tweeter) acts as an open circuit, so the R does not contribute to the attenuation. At high frequencies, the C acts as a short, so that the R shunts the tweeter interacts with the series resistance feeding the tweeter, dropping its level. The transition occurs at f = 1/2piRC.

You'll need to do some crossover sims to get the right values, but it's do-able with a couple go-rounds with any good crossover CAD program. The one I use (Calsod) will actually do the optimization for you once reasonable starting values are plugged in.
 
SY said:
BTW, starting guesstimate values (NOT checked!) would be reducing the tweeter feed 4R7 to 3R9, increasing the shunt 5R6 to 6R8, and paralleling the tweeter with a series of 4R7 and 3uF.


SY's solution of altering both 4R7 and 5R6 is better than merely reducing the feed 4R7; because changing just that one resistance (not both) will affect impedance and hence crossover point. By changing both R values of the L-Pad, you can safely affect only SPL and not worry about altering the crossover point. To begin with, I'd change the R values to the following:

( The current Value: 4R7 and 5R6 equates to a 7.7dB reduction or attenuation for a 8 ohm driver.)

1dB accentuation: 4.3 and 6.9 (i.e. 6.7dB attenuation)
2dB accentuation: 3.9 and 8.6 (i.e. 5.7 dB attenuation etc)
3dB accentuation: 3.3 and 11 ohms.

hope this helps.
 

ralphs99

Member
2006-07-06 1:23 am
Hi Andy,

You may also wish to experiment with the speaker positioning.

Often a subdued treble is mistaken for an exageration in the lower registers. The brain is easily misled in this way as I have found out the hard way on a number of occasions.

You said the bass was very powerful in your original post, perhaps the speaker placement is exagerating the bass slightly, causing the treble to appear subdued.

Before breaking out the soldering iron, I would try experimenting with the placement, moving the speakers further away from the rear and/or side walls.

You might be pleasantly surprised at how a small movement of the speakers can radically alter the perceived voicing of your speakers.

Cheers, Ralph
 
Hi Andy,

I'm a bit of a neanderthal about these things, but I'd probably try to localise the source of my problem with an EQ before fiddling with even the Lpad. Then you can determine whether you've got problems with your entire tweeter level, just in a single area (2-4k, say) or around the crossover.

Not wanting to be insulting or anything - but are your tweeters in phase? Sorry, sorry - it must be the weekend coming up:)
 

jomor

Member
2004-02-20 1:36 pm
Athens
andy2 said:

On his website, the measurement show a slight dip (1db or so) in the 1k - 3k area. I hope to bring it up to about +1db in that region. Can someone suggest a change in the xover?

On the plus side, the bass is very good -- tight, powerful, and very controlled. The treble is too relaxed though but I am playing around with the series resistor

The 1K-3K area usually gets lowered by the designer to compensate with early wall reflections. It is done intentionally and on small rooms (not unusual in european houses) a typical dip is about -3dB. I suppose your speakers are far away from side walls.

Dont change the series resistor more than 1 Ohm because even slight changes alter the tweeter's phase (which will blur your mids)
 

jomor

Member
2004-02-20 1:36 pm
Athens
one more thing, YOU CANT change the value of one element without messing up the phase around the cross frequency. Usually if one element of the high pass is changed, to restore the phase matching of the drivers one or two elements of the low pass filter need to be changed too. Thats why calsod (or any other) simulation software is necessary for such changes.
 

Salas

diyAudio Chief Moderator
Paid Member
2002-10-08 11:31 am
Athens-Greece
The 1K-3K area usually gets lowered by the designer to compensate with early wall reflections.

I have replied to another thread before that this is a missconception. Early reflections are Haas territory and can only interfere with inteligibility and spatial perception.
Various rooms can present different echograms in the mids and give delayed energy variations in the midrange and elsewhere.
Yes the designer's room could have been livelier and/or maybe he was happier to engineer a dip.
Many people are accustomed to slighlty or even much recessed midrange. Normally those not familiar with live natural instrument acoustics. In a hobby its valid to engineer custom preferences provided it comes with a disclaimer when recommending designs.
 

jomor

Member
2004-02-20 1:36 pm
Athens
salas said:


I have replied to another thread before that this is a missconception. Early reflections are Haas territory and can only interfere with inteligibility and spatial perception.
Various rooms can present different echograms in the mids and give delayed energy variations in the midrange and elsewhere.
Yes the designer's room could have been livelier and/or maybe he was happier to engineer a dip.


Who said the opposite.. I just explained why the designer probably made it this way, obviously Andy's room is different, and obviously Troel's design,
as any other design on this world, will sound different on each room.


salas said:
Many people are accustomed to slighlty or even much recessed midrange. Normally those not familiar with live natural instrument acoustics.

Reflections comensation has nothing to do with "loudness" sound that many riceboys like, as you imply. And trust me, there are designers in our country who do have live experience and do get musicians as their advisor for the development.


salas said:



In a hobby its valid to engineer custom preferences provided it comes with a disclaimer when recommending designs.



I could agree, of course I m not authorized to stand up for Troel's designs, but then again (no offense please) I dont see any disclaimer (nor baffle step compensation, not to mention measurements) at this design either.

Speaking in general, each member speaks for himself according to his own experience and knowledge (or ignorance), in a public forum there is no other way to exchange opinions, and obviously its at the readers'judgement to risk adopting the ideas presented by anonymous co-members or not. It is quite healthy that Andy trusted Troels' design and much healthier that he decided to adjust the design to his own preferences :)