Speaker tweeter modification: L-pad & high frequencies

[raabje]

Speaker tweeter modification: L-pad & high frequencies

hello group,


First of all, I am an absolute beginner when it comes to DIY audio (but I am technical...)

I think my question is asked many times before, but it would be realy nice to get advice for my own little problem...


My question is: what is an easy way to reduce the level of a tweeter, and second question, what is possible to reduce high frequencies (hihat sounds, shakers, samba-balls, etc.)

My setup: NAD521i cd player, Denon 425R receiver, BNS 2-way speakers (not a real high end set...). With an other audio set I have damaged the woofers and tweeters of the BNS speakers, I have replaced (2 years ago) the woofers and tweeters with Visaton gear. But now I am getting tired of the high frequencies, I would like to bring it down. Probably the Visaton speakers don't match that good with the BNS filter. I am getting more and more critical these days...


tweeter: http://www.visaton.com/english/artikel/art_170_1_8.html

DT 94 8 OHM Art. No. 9004
20 mm (0.8") polycarbonate HiFi driver with round shaped front plate for the highrange above 3000 Hz. Balanced frequency response and good omnidirectional sound distribution. High power handling due to ferrofluid-cooled voice-coil.
Frequency response (-10 dB) 1200 - 22000 Hz

woofer: http://www.visaton.com/english/arti...t_458_1_19.html

17 cm (6.5") HiFi woofer with coated paper cone, powerful drive and extended rear pole plate. Special voice-coil and elastic rubber surround for long strokes.

Frequency response (-10 dB) fu -8000 Hz


The crossover filter of the BNS works at 5000hz. A year ago I have conencted a resistor of 3.9 ohm in serie with the tweeter. Today someonde told me this will have an effect of the crossover filter.


What I want: First of all get the level of the tweeter more balanced with the woofer. Someone told me a 15ohm potmeter would do between tweeter and filter, but this would influence the crossover filter? An 8 ohm L-pad will be probably more wise. But I have read on this forum that an L-pad will pump up the high frequencies a bit? So, the volume is better, but the effect is worse? True?

So, what would be the best solution?

- remove the 3,9 resistor for a L-pad?
- or connect a 15 ohm, 5 watt potmeter resistor?
- is there an easy way to damp the high frequencies?


Thanks and bye,

Jeroen

 

[paulspencer]

An L pad has two resistors, usually of 5 or 10w type, one in series, the other in parallel with the tweeter. They are given values such that they do not have any effect on the impedance of the tweeter, and merely reduce the power to the tweeter.

If there is an Lpad in the tweeter already with 2 resistors that perform no other function, then you can replace them with different resistors to get more attenuation. If you aren't sure, then best to leave what was originally designed, but add on an Lpad consisting of two resistors.

Do a google search for an Lpad calculator.

This is probably the simplest tweak for an xo, anyone can do it. I found I was always turning back the treble tone control, but found that after I eventually tried an Lpad, got a better result.

My preference is for at least 2db lesser level to the tweeter. In many cases, however the tweeter will be a bit more efficient so a 4db pad might be a good guess in many cases.

 

[raabje]

Okay, so a L-pad is nothing more than 2 resistors soldiered at the right places. I thought it would be something like a single part, like a potmeter.

I have found a calculator:

http://lalena.com/audio/calculator/lpad/

I understand it's a matter of taste, a 2db or 4db L-pad or even more. So, it's trial and error.

Maybe a solution would be to build a double L-pad circuit, 2bd/4bd, with a little A/B switch at the back of the cabinet?

I will first remove the single resistor, to get the crossover work at the right frequency. I can always replace it later.

thanks!

 

[wintermute]

Okay, so a L-pad is nothing more than 2 resistors soldiered at the right places. I thought it would be something like a single part, like a potmeter.

I have found a calculator:

http://lalena.com/audio/calculator/lpad/

I understand it's a matter of taste, a 2db or 4db L-pad or even more. So, it's trial and error.

Maybe a solution would be to build a double L-pad circuit, 2bd/4bd, with a little A/B switch at the back of the cabinet?

I will first remove the single resistor, to get the crossover work at the right frequency. I can always replace it later.

thanks!

You can buy a variable lpad as well which does look (but doesn't behave) like a potentiometer, but a couple of resistors will be cheaper

Tony.

 

[raabje]

Aha, so L-pads do excist as a complete unit.

I will check the electronic parts shop on friday, check the prices of spare parts against a factory L-pad unit.

I have read somewhere that the material used for resistors, inductors etc. will influence the sound (logical, everybody knows.. ), so I guess high quality spare parts will not be that cheap. And a double or triple option (2/4/6 dB) circuit with A/B/C switches will cost more than selfbuild 3 single circuits (= 6 circuits for both speakers) and pick the right one. Probably the factory build L-pad is the right one, and I like to idea to operate the L-pad "on the fly".


Not that it has to be that cheap... but the cost have to be in balance with the cost of the total speaker, old secondhand BNS cabinets with ordinary Visaton speakers..... Next christmass I will go shopping for new speakers (Quad 12L or something like that), that's for sure.

 

[paulspencer]

I have read somewhere that the material used for resistors, inductors etc. will influence the sound

You read a lot of things about audio ....

a resitor is a resistor is a resistor and I don't think you will hear any difference with different types. Other crossover parts may be another matter ... as long as the values are about right and it has sufficient power you should be fine

Looking at your drivers, they have an "interesting" response! The tweeter has a rising response:
Tweeter:

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

The woofer also has a rising response as you go up higher:
Midbass:

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

Look at the low efficiency of the midbass down low - 82 db @ 100 Hz! The tweeter is 93 db @ 15k! Hence, you have a big problem here! You have a speaker that is so inefficient it's ridiculous! ... unless you can get more of those midbass drivers ultra cheap ...

This is just beyond what an Lpad can fix, since at 82 db and below you are feeding them crazy amount of power to get any kind of output!

EDIT: images won't come out ...
so here they are:
Midbass:
http://www.visaton.com/english/bildgross/458_afrequenzgang.html
Tweeter:
http://www.visaton.com/english/bildgross/170_afrequenzgang.html

 

[raabje]

okay, got it...

My BNS/Visaton setup is totally out of balance...

The gap between the bass and tweeter is too big to fix with a L-pad.


You know what, I think am gonna send my speakers to the junk-yard... I guess it's not worth to spent any more money at.

I could try to build a 6 or 8 (or 10?) dB Lpad, for the educative reasons: I still am curious what will happen to the sound.

I have bought the What hifi? magazine today, Wharfdale diamond 9.1 could be a nice little speaker, to bridge the gap when I am ready yo buy real good ones + a new amp. that will take some time.

My problem with high frequencies started after I had bought speaker stands (Atacama nexus 6) one week ago! These stands work so well (compared to my self build ones), now I can hear what my speakers really produce.

 

[wintermute]

Hi Raabje,

don't give up that easily a lot of response curves for woofers are measured in non-optimal boxes. This will result in the bass response looking way down when it will be very different when implemented in an optimal box.

Unless you have a response measurement in your box then don't take the published response curve as gospel. For instance the spec sheet for my Vifa M26WR-09-08 10" woofer says it was tested 320L box! hardly optimal for that woofer, which is suited to vented up to about 100L.

The other thing is, that even though your speakers wont be perfect (what speakers are) you can still learn a lot from playing around with them, and as long as you get enjoyment out of them then what does it matter.... better to persist and experiment and learn than to throw them away and start from scratch, if nothing else it will teach you what can and can't be achieved and if you buy the variable LPAD you can allways use it for determining the optimal resistors for future projects

Tony.

 

[paulspencer]

Well, nothing wrong with experimenting, but if he wants to get something hifi, I think he'll be better off starting with some decent drivers which have a reasonable response. Those things need 4x the power of a regular hifi speaker! ie. 200w instead of 50w, no that they could handle 200w ...

 

[wintermute]

Hi Paul,

you may be right, but I just had a look on visatons web site and they have a 3way using two of these drivers for bass duty... assuming they are in paralell then you would (in reallity) get about an extra 3db. Now if you look at the freq response of this speaker it is at about 88db at 60Hz (the response plot for the driver is at 80db at 60Hz).... now this is what I was talking about earlier about the published response curves not necessarily showing how the driver will perform in an optimal enclosure... just by putting two drivers in parallel you aren't going to get an extra 8db of sensitivity.....

Even so assuming a single driver will be around 85db it's still pretty ineficient, but not as bad as 80db

here is the link ---> http://www.visaton.com/english/artikel/art_546_6_21.html

Note that neither the response plot for the midbass or then one for the speaker mention what testing method was used, so both IMO can be taken somewhat with a grain of salt.

My feeling is we all need to start somewhere, and in the case where someone has already built something, we should encourage them to learn from what they have done, and then they can use that knowledge when they move onto their next project

I think that saying if he wants hifi he should get some decent drivers, is not at all constructive and may make him go back to the idea of junking them..... just my oppinion

Tony.

 

[raabje]

hello there

(Wintermute --> Neuromancer! great book... )

Well, you are both right, for the learning aspect it would be interesting to give old unbalanced speakers one more shot. A good reason to dive into DIY audio!

I think I could soldier a cheap 6 or 8 dB L-pad, maybe first for one speaker, to find out what will happen with the sound.

But a pair of new speakers is also very tempting. I think I will do both, first give them a shot, and try to enjoy them for a few weeks, or maybe longer? The new speakers will be at this moment entry-level ones, I mentioned the Wharfedales 9.1's, or maybe Monitor Audio Bronze B2's. So a better sound, but not perfect at all. I will search on this forum for experiences and tips&tricks.


What do you think what would be in my case a good value for the Lpad? 4/6/8/10 dB? (if I don't get the variable Lpads, don't know yet, the guy in the electronic store told me it's 11-12 euro for one)

Thanks for your input, really appreciated!

 

[wintermute]

Just noticed that the crossover freq is 5K..... that should be fine for the tweeter, but is not really suitable for the midbass. Even if you put the LPad on the tweeter you may find that the speaker sounds too bright, especially with the response peak of the midbass at around 5K....

The fs of the tweeter is 1900Hz so the minimum crossover freq for it should really be about 4K..... the midbass on the other hand probably shouldn't be crossed over above about 3K and if the response curve of this woofer is to be trusted then you really have a problem with it's rising response.... It actually is about the same level as the tweeter at a bit under 4Khz, so even if you attenuate the tweeter, the midbass will still be putting out high levels (in it's upper range).

It's a hard call. I'd try 6db but you might end up having really prominent midrange with no bass and no high end.

I also just realised that you used an existing cabinet, and were replacing blown drivers... you may get much better results if you built a new cabinet for the woofers and got a tweeter that can be crossed over much lower (like 2000Hz), but this is really going a bit far, especially as Paul has pointed out the drivers aren't particulary efficient, and they do appear to have a rather consitently rising response curve. I'd only suggest this as an experiment on the proviso that you were buying the tweeters for another project and they happened to be suitable for experimenting with this one.

The other thing you could do is model the driver (using published T/S params) and compare to your existing box, if your existing box is to big, try adding some bricks if too small add some stuffing.... if it is a bass reflex, then chances are it is grossly mis-tuned, and getting it right will be very difficult!

Tony.

 

[paulspencer]

The tweeter after padding will have an easy ride. I'd consider 2.4k a good starting point with an Lpad with a value to match efficiency of the midrange above say 200 Hz. As suggested you might model it and see how your box models, you might tweak the box tuning a bit. And you will probably benefit from turning up the bass tone control a little since it's about the only way you will get any kind of bass.