JBL LSR305 tweaking

I have just purchased a pair of modern legendary active monitors JBL LSR305. They were on sale at thomann.de obviously because a MkII model has been released recently.

This little gem has been reviewed eg. here JBL LSR305 Studio Monitors

I ran some measurements and listened to a single speaker today. Measurements line up with noaudiophile well. Sound is very impressive but there is a strong coloration making them sound boxy, also knucle test result is poor. I put the mic almost in contact with panel and every measurement showed strong distortion around 240Hz. It is obviously a 1/1 wall resonance based on height of the cabinet. --> http://semimac.org/wp-content/uploa...-Vibration-Analysis-Loudspeaker-Enclosure.pdf

Internal cross-bracing should kill that well, I'll try that during next days and report my findings!
 

Attachments

  • LSR305 panel resonance disto.jpg
    LSR305 panel resonance disto.jpg
    140.2 KB · Views: 1,966
Founder of XSA-Labs
Joined 2012
Paid Member
I got these a few years ago to try them out. The polar dispersion is very nice measurement - but the mid bass was very weak and distorted. Just lacked midrange power and presence. I had to return them because could not sound anywhere near as good as my FAST with 10F/RS225. But what bothered me was that despite being internally DSP’d - the response was not flat.

http://www.diyaudio.com/forums/full...-rs225-8-fast-ref-monitor-45.html#post4347377

It’s the red curve:
lsr305-tg9fd-10f-fast-copare-4ms.png
 
Last edited:
Interesting !
Which mic are you using- And how close to the cabinet?

Wondering if for instance MiniDSP umik1 can be used to test for distortion.

I use UMIK-1 and REW. For this resonance test I held the mic's foam-covered head in contact with the panel. The metal body didn't touch panel. I played a normal sweep at same spl level as in other measurements (roughly 85-90dB/1m)

The amps are not terribly effective and like XRK said, bass gets stressed easily. Normal mesurements show remarkable distortion below 400Hz. Some of it may be resonance-based, but it is not as peaky as the panel measurement.

I have been thinking of cutting off the waveguide to be used in a semiactive 3-way that I am planning now. We will see... Other option is to make copy molds to clone the waveguide shape...
 
I have just purchased a pair of modern legendary active monitors JBL LSR305....

“Legendary”?!? That right there is hilarious to me.

They are a $90. internally amplified speaker and yes they perform pretty well for the price but c’mon, what did you really expect?

Wrap them in a towel around them and ratchet strap a couple of bricks to the sides of them and see what they sound like.

The 708’s might acheive legendary status as the M2 is moving toward.

Barry.
 
I have a pair of white LSR305s for my garage and a single LSR308 for my Etherwave theremin. Very interested in watching this thread.

I have considered removing the baffle and filling the backside with sikaflex sealant (self-leveling?) to deaden it a bit. An internal cross-brace is not a bad idea either.
 
Legendary humour? But they really are extraordinary for the price!

An internal cross-brace made -25dB reduction in resonant sound, measured with backplate opened and knockin the side panel. Distortion peak measured at 60m (2') was reduced too. Me happy for solid 1 hour "work" on these. I hope that my daughter willl be happy too for her new desktop speakers!
 

Attachments

  • lsr305 center brace knock spectr.jpg
    lsr305 center brace knock spectr.jpg
    387.4 KB · Views: 1,948
  • lsr305 center brace inside.jpg
    lsr305 center brace inside.jpg
    343.9 KB · Views: 1,949
  • mancave lab.jpg
    mancave lab.jpg
    276.7 KB · Views: 1,921
  • LSR305fix panel resonance disto.jpg
    LSR305fix panel resonance disto.jpg
    137.8 KB · Views: 1,798
  • Like
Reactions: 1 user
Legendary humour? But they really are extraordinary for the price!

An internal cross-brace made -25dB reduction in resonant sound, measured with backplate opened and knockin the side panel. Distortion peak measured at 60m (2') was reduced too. Me happy for solid 1 hour "work" on these. I hope that my daughter willl be happy too for her new desktop speakers!

Great work with the brace and love your workspace. :) And it's not like hot glue is even that intimate of a bond. (maybe playing a damping role?)

Anyone sealed the vent (aperiodic, aka, a sock or 3, or completely closed off) and then run them with a sub? 6L6 is that what you're doing? Seems like they'd do a whole load better.

Been thinking of getting a pair for my older brother as they've got nothing at home, but need to make sure I get something they can use a laptop/cell phone to drive as well.
 
Daniel, they have adjustable gain, any cell phone can drive them. You just need to get adapters to XLR.

Roky, dsp can not do anything good for distortion. These obviously have so called baffle step compensation in dsp, which means that frequencies below 500Hz are boosted or higher frequencies attenuated. Either way distortion in bass is higher, with same spl at 1000Hz.

So these cheap boxes suffer from various rattles and resonances in the cabinet, plastic front plate, metallic back-plate and wires wibrating etc. The loudest noise came from the box panels ringing like a bell with signal of 240Hz and around. 2nd harmonic means buzzing at 480Hz, 3rd ringing at 960Hz etc.

Resonances from cabinet walls and other mechanical parts are usually easy to fix. But if I wanted to make these hi-fi, I would construct new heavy cabinets, fill the backside of plastic front plate with something heavy and puttery and place the amplifier board outside the cabinet or in an isolated chamber. I am sure that the wise engineers at JBL know this too, bet they had to fit a certain retail price and still make profit with manufacturing! Nothing wrong with that!

These same tricks can be used for any cheap loudspeaker. This was not first time for me either.
 
So I used Audacity to analyze sound of knocking the cabinet. My calibrated UMIK-1 microphone was the source and recording was set to mono 16-bit. The opened cabinet was laying on the carpet, I held the mic's head on level with the open back and knocked the sidewall of cabinet with my midfinger knuckle several times.

Then I painted several knocks in the soundstream window and clicked "Analyze" and "Spectrum" to get spl/f graphics. You can adjust parameters of analyzer if you want, more samples means more low end and resolution.

The problem is that knocks aren't equal, but taking several knocks in analysis will make sort of averaging.

I have learned this myself and I use this or modified methods for several other purposes too. The challenge is to make it reliable/repeateble. Nice to check the sound of different guitars this way! Some luthiers use this kind of method too and there are several published articles.
 
...The problem is that knocks aren't equal...

Bruell and Kjaer actually make a mechanical knocker to standardize this sort of job.
It's mainly for structure-borne noise in architecture and drops half kilo masses about 40 mm.
Maybe you could just manually release a small hammer to fall from a measured distance?
Fairly repeatable and basically zero cost.

Nice result of your simple modification, well done.

Best wishes
David
 
^Actually I have tried a pendulum system like that. However just knocking works. Knocking with a plastic hammer or other object would be better, but knuckles are always handy! Listening to knocks is in line with measurements!

My -25dB is "worst scenario", the cabinet's back wall was opened, cavity working as an amplifier in the measurement. But still, with music the fixed speaker sounds much better than original!