Setting up 3 way JBL horns for correct polarity

[Earl K]

Do you think a 6db XO at 5kHz for the 2404 is OK, or should I use 12db as others have suggested?

If it was my project ( & because I like large format drivers on round horns ) ,
- I would run the 2441s all the up to a fairly linear point ( say 10 to 12K ) on that horn . I would do this even if it meant the creation of a HF compensation circuit ( hopefully not needed ) and the insertion of a series LCR notch filter strapped across the driver.
- I'd also first try out a gentler slope in lowpass section of the bandpass on the midrange horn ( I'd attempt to add the natural 6 db acoustic rolloff in the HF to the 6 db electrical rolloff ).
- Note: To be able to use a ( Speaker Level ) LCR you would need to add a 12 to 16 ohm inline resistor so that the LCR notch could actually resonate & then attenuate .

- Then I would cross the 2404 in with a 3 or 4 pole ( speaker level ) circuit . I would create either a classic 3 or 4 pole filter. Or maybe 3 pole "bump filter" which has a bit of EQ in it ( resonant bump filter ). JBL has shown the way in how they used to employ these resonant filters on UHF devices within their "Legacy" 43xx ( big box ) products.

[Doogster]

If it was my project ( & because I like large format drivers on round horns ) ,
- I would run the 2441s all the up to a fairly linear point ( say 10 to 12K ) on that horn . I would do this even if it meant the creation of a HF compensation circuit ( hopefully not needed ) and the insertion of a series LCR notch filter strapped across the driver.
- I'd also first try out a gentler slope in lowpass section of the bandpass on the midrange horn ( I'd attempt to add the natural 6 db acoustic rolloff in the HF to the 6 db electrical rolloff ).
- Note: To be able to use a ( Speaker Level ) LCR you would need to add a 12 to 16 ohm inline resistor so that the LCR notch could actually resonate & then attenuate .

- Then I would cross the 2404 in with a 3 or 4 pole ( speaker level ) circuit . I would create either a classic 3 or 4 pole filter. Or maybe 3 pole "bump filter" which has a bit of EQ in it ( resonant bump filter ). JBL has shown the way in how they used to employ these resonant filters on UHF devices within their "Legacy" 43xx ( big box ) products.

Hey Earl

Just thought I'd touch base with you. I finally have the 3 way PLLXO set up (I even made up a proper chassis for it). Everything sounds very balanced - certainly a lot better than it was. After some tweaking I will try out your suggestions.

However, there is one nagging thing - the 2404H tweeter. It is very hissy. So much so that it is distracting. It doesn't matter what amp I use. Is this normal for this tweeter? Any suggestions for how to tame the hiss?

Cheers. Doug

[Earl K]

Hi Doug,

Nice Work !

As far as I know , you're the first poster here at LHF, to design & construct his own full blown PLLXO setup ( 2 or 3 way ) . So, Congratulations !

However, there is one nagging thing - the 2404H tweeter. It is very hissy. So much so that it is distracting. It doesn't matter what amp I use. Is this normal for this tweeter? Any suggestions for how to tame the hiss?

- The above text is a little vague for a proper response !

- Do you mean the tweeter circuit is annoyingly hissy / without program ( music ) playing ? If so, look to your electronics ( gain of preamp / amps / as the culprit(s) )

------------------------------------------/ or /

- Do you mean that the tweeter is hissy when music is playing through it ?

--------------------------------------/ If so then : /

- Just what are the LCR values in this tweeters' PLLXO ?

- What gain setting are your amps set to ?

- A Hunch ( based on your other PLLXO values ) : You may have made this ring radiator device ( which is IMO, already overly resonant below 10K ) , even more resonant by building a PLLXO that could be ringing ( if its' multipole filters are mathematically to close to each other ) .

- IME ; all JBLs' ring radiator type tweeters suffer from this characteristic of being overly resonant below 10K. That is the most significant reason I can think of, to severely limit their response curve to above 10K.
- ( IME: 5K to 10K is an extremely sensitive area for the ear when it encounters undue resonance ; ie , one wants a device with well behaved damping reproducing this range . )

- The solution ? Cross above 9 or 10K. Or to maintain a 5 K point,( & it's conjecture only ) eBay a pair of 2407 drivers / then mount them into some tiny wavequides ( or even just parallel slots, a la Smith Horn tweeter ) and then enjoy the sound of damped Mylar .



EDIT: ( added sometime later )


- The hiss problem could also be a bad FR curve .
- What does the response curve look like ( for just your tweeter setup ) ?
- Here's one FR plot ( done by Mr. Widget ) of both the 2404 and 2404-1 tweets. IME ; this is not a listenable curve .

- Depending on what sort of response you're now getting, you may want to try some extra, HP filtering on those 2404H(s) . I just did that with the single 2404H that's here. The filtering was at speaker level & only single pole .
- It wasn't until I got the cap size down to about a .1uF cap that I thought what was coming out of the tweeter was worth keeping around ( & blending into the overall ) .
- This was done by simply listening to only the tweeter .
- This simplistic approach was also done with a Fostex N945 ( a bullet style tweet ).
- I do believe that each range should be listenable in it's own right before being allowed to contribute to the overall .
- Whether or not there's enough level to work with, is beside the point . It's the EQ contour ( at this point ) that's most important at the starting point .
- That's why I've tried to get across a multipole ( 3 or more ) filter keeping the most audible portion of the passband, to over 9 or 10K , is the best way to use these things.
- Yes, with 2 uF inline with either of these 2 tweets ( my starting point in cap size ), everything coming out sounded like some form of distorted noise . And yes, lots of hiss .
- RTAing either of these tweets ( with the .1 uF ) inline cap showed an EQ contour that was : @ 3 db down at 5K, with a gentle upwards slope to 12K, & then a plus 4 db spike at 16K. Overall a pretty decent looking ( & sounding ) curve.
- Do the same exercise. Get a computer based, RTA measurement setup , if you don't already have one...

[louped garouv]

However, there is one nagging thing - the 2404H tweeter. It is very hissy. So much so that it is distracting. It doesn't matter what amp I use. Is this normal for this tweeter? Any suggestions for how to tame the hiss?

Cheers. Doug

how much power are you feeding the 2404H in relation to the other drivers?

just out of curiosity....

[Robh3606]

However, there is one nagging thing - the 2404H tweeter. It is very hissy. So much so that it is distracting. It doesn't matter what amp I use. Is this normal for this tweeter? Any suggestions for how to tame the hiss?

How much attenuation do you have in line?? They are something like 110db @ 1 meter. If you don't have them padded down a bit I can see the hiss level from them being distracting.

Rob

[louped garouv]

How much attenuation do you have in line?? ...

If you don't have them padded down a bit I can see the hiss level from them being distracting.

Rob

generally and theoretically speaking, is it better to pad down via a circuit or reduce power being fed?

[Robh3606]

generally and theoretically speaking, is it better to pad down via a circuit or reduce power being fed?

That’s a really good question and I am not sure I can answer it. Any of you Professional Sound guys want to tackle this one??

Every thing has a residual noise floor. With no signal that is what you are hearing. The lower this is the better for obvious reasons. I try to use the multiple gain stages I have available to get the lowest noise floor I can buy manipulating the gain levels in the amps, crossovers and Eq’s. For us the S/N is probably the most important issue than maximum driver output. With professional SR they need Max Output with the best S/N they can manage.

One of the advantages of an active system is you don’t waste power in the passive networks. You use the gain structure in you crossover and amps to set the driver levels. The problem with this is any residual noise in the chain is not attenuated like it would be in a passive set up. You don’t loose power but you pick up noise for the trade. If you had a noisy 2404 you could drop in a 10Db passive attenuator providing you have the gain available to make it back up. But more gain equals more noise

You have to balance this all out and see what works the best for you. Chances are you will end up better off with the passive attenuator in place but it all depends of how quiet the gain stages are. With my compression drivers I have to use passive compensation which helps knock the noise level down a bit. When seated I can’t hear the system on with a mean sensitivity of about 98-100db per watt. It was a real PITA to get it set-up this way. Without the passive compensation on the compression drivers I can hear the system.

Rob

[Doogster]

How much attenuation do you have in line?? They are something like 110db @ 1 meter. If you don't have them padded down a bit I can see the hiss level from them being distracting.

Rob

Rob

I am not padding them down at the moment. What would be a good value to try? I think I have some L-pads hanging around somewhere.....

Incidentally, today I replaced the 2404Hs with Fostex FT17H tweeters. The hiss has decreased dramatically, but then they are much less sensitive drivers. I actually prefer the sound of the Fostex drivers now that I've listened to both. The Fostex sound more mellow (yes, I have adjusted the levels for difference in sensitiviy) than the JBLs.

Cheers. Doug

[Doogster]

how much power are you feeding the 2404H in relation to the other drivers?

just out of curiosity....

Hi

I haven't measured what I'm giving them, but I have set the levels by ear to what I think sounds natural. I am using a 20 watt tube receiver at the moment (the other drivers are being driven with a 1 watt SET amp and 500W SS amp for the bass).

Cheers. Doug

[Doogster]

Hi Doug,

Nice Work !

As far as I know , you're the first poster here at LHF, to design & construct his own full blown PLLXO setup ( 2 or 3 way ) . So, Congratulations !

- Do you mean the tweeter circuit is annoyingly hissy / without program ( music ) playing ? If so, look to your electronics ( gain of preamp / amps / as the culprit(s) )

------------------------------------------/ or /

- Do you mean that the tweeter is hissy when music is playing through it ?

--------------------------------------/ If so then : /

- Just what are the LCR values in this tweeters' PLLXO ?

- What gain setting are your amps set to ?

Hi Earl

Thanks for your reply.

I might take a pic and post it if anyone's interested what it looks like.

The 2404 is hissy without anything playing. I think it's the whole replay chain that's less than pristine. I use a homebrew tube linestage, phono stage and power amp, all of which use AC on the heaters. It's the monstrously high sensitivity of the 2404s that's causing the problems.

See my post below to Rob - I have now replaced the 2404s with Fostex FT17H tweets, which are about 12 db less sensitive, and the problem has decreased dramatically (there is still some residual hiss, however). I would still like to experiment with the 2404s, but after attenuating them.

These JBL drivers are so sensitive I will have to restructure the gain of my preamp. I currently have all the amps set to very low gain because anything higher prevents me using my linestage properly (as it stands I can't turn the volume past about 8!).

Cheers. Doug

[louped garouv]

my master volume control pot rarely goes above "4"

[Robh3606]

I am not padding them down at the moment. What would be a good value to try? I think I have some L-pads hanging around somewhere.....

I would drop them at least 10db, Try them attenuated to match the Fostex's.

Rob