The JBL 4345 Club

[Ian Mackenzie]

Quote”Not my experience with them. I had 2122's in my active set-up and swapped in 2123's. Did one driver at a time so I could compare them. Once the sensitivity was adjusted and the EQ tweeked so they measured the same in their passband I know I could not tell them apart. As far as a risk of hearing damage I don't understand that at all as the mids are always attenuated to the woofer sensitivity. Once that's done then the whole system is what's driving the SPL.

Rob”

That post took a while to find.

Knowing Rob of course he would dial them in. He dials everything in.....Lol

I will comment on the drivers shortly.

[Earl K]

Andrew, that's a beautiful 4345!

Nice work!

[Robh3606]

Hello Ian

That post took a while to find.

LOL yeah I bet that is a while ago. Here are some old 2122 measurements on my 4344 baffle no network. Check out the step response!

Rob

[Ian Mackenzie]

Great.

It’s all about the measurement technique.

My nearfield measurement had some artifacts above 2 khertz. So on a fine day l will drag it outside and shoot ground plane measurements.

It’s not simple or easy to isolate each driver and compare them as there are what you might call a lot of moving parts in these systems. They both have different passive networks, they have different sensitivities and so you have to reset the Lpad levels. I spent a few hours calibrating the Lpads to bang on for the 2123. I first checked the voltage drives and compared both channels. I then set each Lpad to bang on flat within 0.2 dbm with LMS and matched both channels.

With four drivers it’s a bit like an RIAA phono stage variation specification-+0.5 dB. As a specification -+ 0.5 is not very good. A good phono stage is -+0.25 dB. A small difference here and there and you have a different tonal balance and you can make out each driver. You might think that’s crazy but with four drivers there is a lot of latitude for tonal differences in level. It’s actually quite logical but in reality what happens is far from flat.

Not wanting to appear an elitist old fart my intent is to unlock the real potential of these systems and share how to do that.

The presentation goes from good to being there when it’s bangon. The sounds stage really opens right up and the drivers completely disappear.

My subjective impression so far is the 2123 is really smooth and there is a nice nimble or agile quality about the vocals particularly at the top of its range.

[Robh3606]

My subjective impression so far is the 2123 is really smooth and there is a nice nimble or agile quality about the vocals particularly at the top of its range.

Yeah I really like them especially on vocals. I follow a lot of all girl bands, love the harmonies and they are really clear and easy to follow on them. Speaking of bands you check out Milk Records Melborne??


Rob

[Ian Mackenzie]

Hi Rob

I will check it out.

I put the 2122 in today on one channel and will post details soon.

Okay objectively the 2123 is the more efficient driver. The consequence is it’s low end response is hurt. This is shown is the normalised the response. Both drivers in this measurement are very smooth and beyond that l would not read anything into these curves. They are more entertainment value and fun just to see what’s going on.

If you are thinking of using these curves to design your project they are unsuitable. They are more entertainment value and fun just to see what’s going on. The impact of the baffle is not visible and the upper mid band is hidden by artifacts. The scale is 5 dB per division (unsmoothed)

I need to listen more closely to the 2122 before commenting further below.

As a Rob hinted at earlier the bigger impact on the system is how well the midrange driver is integrated and that’s the skill of the designer in these systems.

[ompdiburi]

Just a heads up l will post a discussion comparing the 2122H and the 2123 soon.

Btw l auditioned the much talk about Dynaudio Focus 60XD today.

It’s a nice active loudspeaker. It has 4x 150 Watt amps per box and some serious dsp horse power.
RRP $15000+

I later played the same cds on my 4345 clones with a custom hi-end analogue active crossover and Pass Labs amps. The 4345’s completely blew away the 60XD on all fronts. Effortlessness, soundstage depth, instrumental clarity and vocal definition was in a completely different league. As they say nothing even comes close.

So don’t even think about trading your 4345’s anytime soon!

I built a 4344 "like" clones and I drive them active using two pass aleph 2 mono blocks on the top, can you tell us more about your xover design? Thanks!

[Ian Mackenzie]

I bet that sounds absolutely marvellous.

I still have my Aleph 2 mono blocks.

The original diy crossover is here.
https://www.passdiy.com/gallery/proj...-low-crossover

That was a while ago now.
I have since simplified the design but my application of Biamping has become more of a science. The end goal is the benefit of Biamping while the hardware itself becomes invisible.

I am still working on this and hope to have more news soon.

[ompdiburi]

I bet that sounds absolutely marvellous.

I still have my Aleph 2 mono blocks.

The original diy crossover is here.
https://www.passdiy.com/gallery/proj...-low-crossover

That was a while ago now.
I have since simplified the design but my application of Biamping has become more of a science. The end goal is the benefit of Biamping while the hardware itself becomes invisible.

I am still working on this and hope to have more news soon.

THX for the info! Did you try any tube xover?

[Ian Mackenzie]

The quest for transparency ends up being about a lower noise floor. To that end l have found Jfets work exceptionally well with high efficiency loudspeakers.

[Dr.db]

To me it looks like the baffle supports the 2122 output above 300hz by about ~5db compared to the lower frequencies, am I right?
In that case a crossover somewhere close to this seems logical, as it makes other network-compensations redundant... The 4344 crossed at 340hz and this might be the reason for that!?

Hello Ian



LOL yeah I bet that is a while ago. Here are some old 2122 measurements on my 4344 baffle no network. Check out the step response!

Rob

[Ian Mackenzie]

I did listen to 2122H but not long enough to draw any conclusions.

There are more inconsistencies in setting up the system than the differences between these two drivers.

What l do want to do evaluate the 2122 after market recone kit

[JBLCarousel]

Hello ,

I keep getting requests to re-post pictures of my JBL 4345s from Kenrick Sound.

Here you go:

[JBLCarousel]

Wishing you all the best.

- Tom

[DES-1]

Thank you, Tom