JBL 708i Monitor Review (DSP: Part 2)

[beatelund]

Thanks for the measurement. Have to ask, is this really the measurement of the BSS output? Not the 708i with the BSS doing the processing?

It is The 708i diy with Roon doing The processing.
It has the same as filter as the bss file But without The sidechain limiter

 

[Tangband]

Guess that 1uF capacitor is used as a linearisation of the horn/waveguide response (and to drop its sensitivity), here I simulated a typical 1" tweeter with waveguide (unfortunately not a similar horn in the database), black curve is the amplitude response of the driver and waveguide in an infinite baffle, blue curve in a typical enclosure and red curve with just a 1 uF filter:

View attachment 256017

Since it is starting quite high it is also sufficient as a high pass and protection despite being first order.

The secret of jbl compression driver sound/crossover revealed…. Good information for the DIY:er.
This crossover is probably usable with other compression drivers/horns to.

 

[changer]

Yes, it is a very useful solution. If the low response of the waveguide is smooth, some PEQs and filters can be saved for linearization.

 

[Robbo99999]

Guess that 1uF capacitor is used as a linearisation of the horn/waveguide response (and to drop its sensitivity), here I simulated a typical 1" tweeter with waveguide (unfortunately not a similar horn in the database), black curve is the amplitude response of the driver and waveguide in an infinite baffle, blue curve in a typical enclosure and red curve with just a 1 uF filter:

View attachment 256017

Since it is starting quite high it is also sufficient as a high pass and protection despite being first order.

This is what I had in my mind (in terms of the competing interrelationships) when Tangband described it earlier:

JBL 708i Monitor Review (DSP: Part 2)

For the tweeter in 708i the 2409H 8ohm they are using a condensator with 1uF. It is always active regardless of single wire or bi amp mode. In bi amp mode the passive crossover for the woofer is bypassed and it is using a 1,5khz LR24db crossover both in the BSS presets and Crown DCi N A 1...

www.audiosciencereview.com

 

[RobL]

Just reading about KEF’s development of the LS-50. They reduced unwanted port output from longitudinal port resonance by incorporating a flexible centre section in their port design. It reduced unwanted output from the resonance by 15 dB. Made me think of this thread. I know they applied for a patent on their approach, don’t know if it was granted.

 

[MAB]

Just reading about KEF’s development of the LS-50. They reduced unwanted port output from longitudinal port resonance by incorporating a flexible centre section in their port design. It reduced unwanted output from the resonance by 15 dB. Made me think of this thread. I know they applied for a patent on their approach, don’t know if it was granted.

Yeah. Or Technics:

Port Resonance Suppression: Inspiration from Technics

You may have seen the measurements I posted here of the Technics SB-C700, a coaxial 2-way standmount. Despite being an innovative design with superior measurements to the LS50 (especially wrt directivity

www.audiosciencereview.com

I am still unclear on the relative audibility, but the 708 seems to be on the high side for this resonance.

 

[RobL]

Yeah. Or Technics:

Port Resonance Suppression: Inspiration from Technics

You may have seen the measurements I posted here of the Technics SB-C700, a coaxial 2-way standmount. Despite being an innovative design with superior measurements to the LS50 (especially wrt directivity

www.audiosciencereview.com

I am still unclear on the relative audibility, but the 708 seems to be on the high side for this resonance.

KEF’s method got this result:


Pretty impressive really. Moving the port to the back enabled a further 15dB of suppression.

 

[Naturlyd]

Here are the measurements of my Crown CDi amp with the JBL 708i settings, both single and biwire. You can see that the curves are identical for the single wire setting (you can't see the green curves as they are overlaid perfectly by the HP measurement) and for the high pass in the biwire setting. I checked this twice. This should be exactly the same transfer function as the one used in the tested unit.

 

[Naturlyd]

Thats my first measurement of an amplifier and it was quick and fun. I used the Howto from @xrk971 - thank you! The gear is a RME UFX II and a 100W 8Ohm resistor from RS-Online. The reported noise floor was around -137dB, XRK971 mentions around -130 with his interface. Is this really possible?

 

[mdsimon2]

Thats my first measurement of an amplifier and it was quick and fun. I used the Howto from @xrk971 - thank you! The gear is a RME UFX II and a 100W 8Ohm resistor from RS-Online. The reported noise floor was around -137dB, XRK971 mentions around -130 with his interface. Is this really possible?

Those aren't real noise floors as they are lowered due to FFT gain. From the RME UFX II specs noise floor should be about -113 dBFS unweighted (presumably 22K bandwidth but they don't say). A SOTA ADC like a grade Cosmos ADC is still only capable of about -126 dBFS unweighted in mono mode.

Michael

 

[jhaider]

Here are the measurements of my Crown CDi amp with the JBL 708i settings, both single and biwire. You can see that the curves are identical for the single wire setting (you can't see the green curves as they are overlaid perfectly by the HP measurement) and for the high pass in the biwire setting. I checked this twice. This should be exactly the same transfer function as the one used in the tested unit.

View attachment 259878

Interesting - I’d never measured biwire. Your data do show a difference between the single- and bi- wire tuning files: there’s a little more baffle step built into the woofer filter in biwire.

Notably, pretty much every instrumented review of 708P going back to Manny LaCarrubba’s (including has noted the tweeter level is just a smidge high. Assuming they ported the biwire tuning files to the plate amp DSP - reasonable assumption IMO given they don’t provide different spins - that is explained well by your measurements. The steeper BSC on the woofer lowers the woofer level at crossover, but the tweeter is, oddly, not padded to match the new level. It’s the same as on the single wire mode, which has more woofer sensitivity in the blend region. So the tweeter is a little hot in biwire mode and on P models. Weird oversight by JBL.