whats going on?

[lbstyling]

hello all,
in my continuing diy quest, i recently got my troels gavesen tjl3w's (seas excel drivers) out for a play- having been frustrated with them for some time due to a slight lack of drive and instrument separation, i decided -as the surround kit was all disconnected, i would tri-amp them just to see what 750w rms would do- if anything.

i wont go on about it too much but it shocked me just how different it was and in many respects its extremely close to the best sound ive ever heared. its mostly the instrument separation thats changed.

i cant see any way that the extra power is being used- the improvement is at low volume and high - so i looked into what makes this so much better and ive read several different ideas on whats happening to produce the change- i just wanted to know if you could shed any light on this?

is it that the crossovers and drivers arnt effecting each other via IM??
or maby that the crossovers are 4th order and it reduced the load the amp sees?

id just like to know- its flipped my understanding of what speaker design trade offs might be on it head at the moment!

[lbstyling]

on another note, im planning to cross my speakers over to my sub with a passive line level xo,

http://www.t-linespeakers.org/tech/filters/passiveHLxo.html

im looking for a 80hz crossover point but im unsure of my maths- i was hoping someone could confirm my answer- for c1- (using a 5k risistor for r1) ive got 3.98nf?? would a 4nf do? (assuming thats right)

[ThomasW]

im looking for a 80hz crossover point but im unsure of my maths- i was hoping someone could confirm my answer- for c1- (using a 5k risistor for r1) ive got 3.98nf?? would a 4nf do? (assuming thats right)

Calculating the correct components requires knowing the input impedance of the amp being used.

[lbstyling]

high pass is 47kw, i dont know the low pass- how do i measure it?

[lbstyling]

got it- should be 20kw for low pass

[lbstyling]

ok for the 2nd order crossover i have the following values-

80hz crossover,

amp low pass imp- 20kw

amp high pass imp- 47kw.

lowpass--
r1=5kw, r2=50kw, c1=3.98nf, c2=13.93uf.

high pass--
c1=3.98nf, c2=0.398nf, r1=5kw, r2=783.6kw.

im not confident on the values being right here- sooo many oooooo's!!

how would this change if the speakers were tri-amped, and crossed over to the sub?- ie still using the internal crossovers for the speakers just with a channel per driver, so the output from the pre amp would be split between the 3 channels and spured to the sub aswell- would i just devide the imp (47kw) by 3 for the calculation?

[lbstyling]

i dont find the guide to pllxo very clear- for example , it says that r2 = 10 x r1, and then says below this that r2 must be resolved to compensate for amplifier impedance -but is that for the low pass or high pass or both??

[lbstyling]

anyone?

[JonMarsh]

Both.

[lbstyling]

thanks jon,
ive been folowing the speaker camp thread- nice work-real nice.
im left with the question on what effect splitting the signal over 3 channels has to the equasion- i was hoping it didnt affect it?

[Dave Bullet]

Is the impedance curve on the tjl3w's pretty flat? I don't know much about impedance swings, but could that cause some form of non-linear distortion?

By separating the crossover sections (assuming it was a parallel xo) - you are going to get impedance swings anyway (you won't have separate impedance flattening on each leg of the parallel circuit), so maybe impedance swings would have nothing to do with it?

What amp were you using? Are you still using it as one of the 3? and if so - which section is it amping? (woofer, mid, tweeter)? If you are still using the original amp as one of the 3 and it isn't driving the mids - put it in the mid section and see if things go back to the occluded state as before.... if so - you just have a poorer quality amp that is being exposed by the Excels....

[Face]

The impedance curve doesn't look too bad.


I would blame something other than the speakers.

What are you using as a source? What's the rest of your rig consist of(receiver type, etc...)? Is it rated for 4ohm operation, etc...

[lbstyling]

ive learn allot kicking about on this forum-so hears my thoughts now-

the amp is a parasound halo a51 5 x 250rms @8ohms, 5 x 500rms @4 -im fairly confident in the amps ability to drive HEAVY loads. its a 5 channel power amp- so i have used 4 of the channels for the 2 speakers and a halo a23 for the tweeters.
its a significant delivery for a speaker that gets to xmax with only 25w.

i think the problem is just that the speaker is in far too large a room for its design.

i notice the w18 is moving a fairly long way even at what i would consider low volumes- increasing im distortion drastically- but i suppose this is to be expected when driving a speaker thats optimised for low end extention-as apposed to max output at a higher freq-when using only a single 7 inch bass unit- theres no missing the fact that the speaker seams to go very low for its sise when you hear it.
its a mistake on my part -going for this design given my room.

so the plan is to try and improve what i can to optimise more for my situation


-ill post the rest later-got to get back to work!!!!

[ThomasW]

i think the problem is just that the speaker is in far too large a room for its design.

Hard to know......

At no time have you posted any specifics about the room. Nor have you posted any measurements of existing speakers operating in that room.

[lbstyling]

the room is approx = 224m'3.
i have no measuring kit yet- im planning to get a mic next month and use rew.

-if i can get the pllxo sorted, (emailed the creator or this page
http://www.t-linespeakers.org/tech/f...ssiveHLxo.html - Peter Rettweiler for advice on how running a tri-amp setup will effect the calculations, no reply as yet) i will see if reducing the work the w18 needs to do will help. its a cheap enough solution to try out whether its a contributing problem or not.

the rest of the kit is just a squeezebox running into a cambridge audio azur dac magic- serious value for money that thing.-infact, i had a benchmark before this, -i prefer the magic!!
i use the squeezebox digital volume control.

[ThomasW]

the room is approx = 224m'3.

Ok so it's 1m X 1m X 224m, interesting .....

The room is ? long
The room is ? wide
The ceiling is ? high

The listening position is ? distance from the speakers?

[lbstyling]

Ok so it's 1m X 1m X 224m, interesting .....

how did you guess?? ;x(

ok its not quite so simple but to give you a idea- the room is a irregular -not far off regular a pentagon shape.
the celing is a kind of 5 sided apex- all sides are triangles connecting to a single (highest) point.

if i was to average the walls to a square box shape it would be 7m x 8 m with the celing arround 4 m high.

the speakers are say a meter or so away from the side walls and around 2m from the back wall with a 4.5 or so gap between them. they are towed in towards a central position (where i sit) approx 5 m infront of them , on a large sofa with a 1 meter gap between me and the wall behind me.

what ya think?