IMHO, best thing is to start simple, no correction circuit or EQ. That's what I do with new speakers. After a while, could be a few songs or a few days, I get a better feel of what I like or don't like about the speaker system and act accordingly.
Hope the driver and cabinet works out well for you. Let us know how it turns out.
IG
Hope the driver and cabinet works out well for you. Let us know how it turns out.
IG
Hi there: My approach to a two way speaker set-up is to biamp a TB 4-inch bamboo cone driver with a bass unit. The TB's slightly forward sound can be tammed by adjusting the tone control. The TB is set in an open back with 60 degree wings. I tried it in various experimental boxes, however I prefer it in the OB. The biamp crossover is set at 135HZ to the TB. I like the set-up enough that I'm constructing OB for TB 5-inch. I listen to classical and 50/60's coljazz music. ...regards, Michael
Hi Michael,
I am trying to have a compact full range system instead of an elaborated open baffle two way system. I am sure it sounds wonderful in open baffle, sure it will. I know the tone control can help, but I would like to bypass tone controls on receiver/integrated amps as I found that they can add some "cloudiness" to the sound. Direct output sounds a lot better IMHO.
I am trying to have a compact full range system instead of an elaborated open baffle two way system. I am sure it sounds wonderful in open baffle, sure it will. I know the tone control can help, but I would like to bypass tone controls on receiver/integrated amps as I found that they can add some "cloudiness" to the sound. Direct output sounds a lot better IMHO.
Hi borispm, you might like to take a look at the frequency response leveller (notch filter) calculator here:
Strassacker: Lautsprecher - Boxen - Selbstbau
I found it very useful for flattening a 6dB peak at 8.5K on the TB W4-657D.
With the calculator you have to enter the impedance of the speaker at the frequency to be suppressed.
Cheers,
Steve
Strassacker: Lautsprecher - Boxen - Selbstbau
I found it very useful for flattening a 6dB peak at 8.5K on the TB W4-657D.
With the calculator you have to enter the impedance of the speaker at the frequency to be suppressed.
Cheers,
Steve
Thanks a lot, Steve! It seems like I will end up with an .15mH inductor, an 8ohm resistor and an 1.66uF cap.
Still very hard to choose between a single inductor or the notch filter. Everyone says I should make it as simple as possible i.e. no filter at all, but the rising frequency response and broad and tall peak at 10kHz-15kHz looks terrible to me and I can already imagine what a 10dB peak in that region be like...
Still very hard to choose between a single inductor or the notch filter. Everyone says I should make it as simple as possible i.e. no filter at all, but the rising frequency response and broad and tall peak at 10kHz-15kHz looks terrible to me and I can already imagine what a 10dB peak in that region be like...
Borispm,
"Seems like everyone advices me to not to use correction circuits of any kind, maybe I should compromise between filterless, lively sound and filtered, polite sound by using a single inductor instead of a more elaborated circuit. After all it's better to have lively sound than dull and lifeless sound. 0.2mH seems like a decent start, thanks for your suggestion, IG."
I also suggest not using a correction circuit (at least high level). Very often peaks in the frequency range and magnitude you are concerned with are prominent on-axis (where measurements are usually made), but are not present 10-20 degrees off-axis. Therefore, the only compensation you will need is a little toe in - have the speakers cross in front of you instead of aimed straight at you. Provided that there is not a highly (acoustically) reflective surface behind you or directly to your sides (like glass windows or glass doored cabinets) a little toe in will largely resolve the HF peaking in this driver.
My experience with series, high level correction ciruits is similar to that reported by GM. They seem to take some 'life' from the sound.
OTOH, Correction networks can be used to advantage with drivers for which toe in does not work satisfactorily and that have otherwise exceptional performance. Lastly, there is also a personal preference/tolerance issue here. They are also useful for people who can't tolerate the residual peaking audible off-axis, but otherwise like the sound of the driver.
My preference is to select drivers that won't need correction circuitry.
Regards,
Bob
"Seems like everyone advices me to not to use correction circuits of any kind, maybe I should compromise between filterless, lively sound and filtered, polite sound by using a single inductor instead of a more elaborated circuit. After all it's better to have lively sound than dull and lifeless sound. 0.2mH seems like a decent start, thanks for your suggestion, IG."
I also suggest not using a correction circuit (at least high level). Very often peaks in the frequency range and magnitude you are concerned with are prominent on-axis (where measurements are usually made), but are not present 10-20 degrees off-axis. Therefore, the only compensation you will need is a little toe in - have the speakers cross in front of you instead of aimed straight at you. Provided that there is not a highly (acoustically) reflective surface behind you or directly to your sides (like glass windows or glass doored cabinets) a little toe in will largely resolve the HF peaking in this driver.
My experience with series, high level correction ciruits is similar to that reported by GM. They seem to take some 'life' from the sound.
OTOH, Correction networks can be used to advantage with drivers for which toe in does not work satisfactorily and that have otherwise exceptional performance. Lastly, there is also a personal preference/tolerance issue here. They are also useful for people who can't tolerate the residual peaking audible off-axis, but otherwise like the sound of the driver.
My preference is to select drivers that won't need correction circuitry.
Regards,
Bob
Can't speak for the others, but what I am saying is to at least START simple, with no correction circuit. Nothing wrong with using them afterwards if that's what it takes to make them sound better to you. One cannot judge from a FR graph exaclty how a speaker system will sound.
Actual example
: I have some Philips AD9710, which have a notorious peak around 3kHz, but I find that, with a lot of source material, there is no need to do anything about it. Although on some stuff, I will notch it down a bit with my preamp's parametric tone control.I just don't want you to turn this into a headache before actually knowing if the peak will cause you one.
IG
FYI, Check out what other designers have done with this driver.
The 3CRTi
Parts Express DIY Project
and the Lineup series, with and without tweater
HTGuide Forum - Lineup Series Build Thread (F4, D4, D44, R4, R44)
From a little research the filter on Wolf's 3CRTi works the best and has been used on a couple of other DIY efforts to good effect.
Good luck
Mark
The 3CRTi
Parts Express DIY Project
and the Lineup series, with and without tweater
HTGuide Forum - Lineup Series Build Thread (F4, D4, D44, R4, R44)
From a little research the filter on Wolf's 3CRTi works the best and has been used on a couple of other DIY efforts to good effect.
Good luck
Mark
Hi guys,
Sorry I have been busy for a while (and will be busy again in late May, thanks to my uni exams—duh!).
I realized that the rising response starting from 6kHz may be on axis and a bit of toe in may fix it, after referring to the similar if not identical rising of the bamboo cone sibling. All I have to do is to apply a simple notch to iron out some excessive cone breakup at 10kHz and everything should be fine. Still waiting for my long summer holiday to order a pair of these and play around...
Sorry I have been busy for a while (and will be busy again in late May, thanks to my uni exams—duh!).
I realized that the rising response starting from 6kHz may be on axis and a bit of toe in may fix it, after referring to the similar if not identical rising of the bamboo cone sibling. All I have to do is to apply a simple notch to iron out some excessive cone breakup at 10kHz and everything should be fine. Still waiting for my long summer holiday to order a pair of these and play around...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.