Passive monitor design update project

[macka]

Hi Guys.

I want to try and draw on your practical experience with a project I am working on .

The project is a review of the bass reflex tuning and passive woofer / midrange crossover of a studio monitor the JBL 4345.

The monitor is a 4 way that can setup for passive and active operation @290 hz. Historically the system has always performed subjectively far better in active mode than passive mode.

The project is to explore how the existing system and the full passive crossover mode can be improved to enable the sole use of my Passlabs X250.5 power amp. To me at least, its a crying shame not to be able to fully explore the full potential of this fantastic amplifer.


The JBL camp have long said that you can only really sucessfully bi amp such systems because of the low crossover point and yet I note there are numerous hi end systems commercially available with low bass/mid crossovers. While I have no doubt that biamping is the easier and to a point the ultimate solution I don't accept that a full passive network if designed correctly cannot be made to work very well.

For some background I will post some technical data on the system provided by JBL over the next few days and some other data I have obtain through measurement.

I built the monitor as diy clone a few years ago and have been quite pleased with it even though its a 20+ year old design. In recent times I have improved the crossover with parts upgrades and used an improved diaphragm in the horn compression driver. In case you are wondering, no it does not sound like a horn system. The system's merits are hi sensitivity (94-95 db), wide bandwidth -3 db 32 hz - 20 khertz, low intermodulation distortion and dynamic range.

Modelling the project is a bit of a problem in that I only have Bassbox 5.1 by Harris and AudioTester. I have thought about obtaining Bassbox Pro and ther Xover package or alternatively Soundeasy but the latter is a rather complex package and it would take a while to up skill. Perhaps some of you can offer a few pointers.

Macka

[macka]

The inital project design problem is to review the current system bass reflex tuning.

According to data on hand the system was tuned to about 28 hz in net enclosure volume of 9 cu ft3. The original designer advises the gross volume was about 10 cu ft3. Take away the internal mid range enclosure, bracing, ports and rear volume of the woofer and add the effect of 3/4 inch fibreglass results in the net volume of 9 cu ft 3. The system has 3 round 4 inch ports with a length of 8.25 inches located close to each other on from front baffle near the side wall.

Subjectively the system sounds crisp and tight in active mode using a 18 db slope @290 hz.

In passive mode the sound is decidedly woolly with a lack of lower midrange clarity.

I think that part of the problem is the stock choke is 0.5 ohms (5.4 mH) and partly the way the mid cone filter interacts with the motional impediance of the woofer.

The passive crossover filter appears to have an effect on the tuning of the woofer and the voltage drives to both the woofer and the mid cone.

As an initial attempt at reducing the dcr of the woofer filter choke I obtained a 0.2 ohm choke of the same value. This actually made matters worse. The sound taking on a heavy woofy character and the lower midrange sounding chesty.

On further exploration I discovered the LC values for the woofer filter had been optimised for the 0.5 ohm dcr choke. The lack of suitable software makes to difficult to compute updated values for optimised voltage drives.

I will post below data on the factory voltage drives (into 8 ohm loads) and a model of the mid range tuning on 0.5 cu ft enclsure which is sealed.

I have explored some T/L tunings and even the Bullock on Boxes T/L tables but I find the models offered by various packages appear to offer differing tuning combinations refered to as optimium.

[sreten]

Hi,

Regarding tuning options :

Experiment with closing off ports, 2 only, 1 only.

I prefer reducing diameters with foam lining rather than closing.

/sreten.

[macka]

Thanks for the tip.

I decided to remove all the fibre glass fill and start with a basic volume and as you suggested seal one port to allow straight forward adjustment. The remaining two ports only needed a slight adjustment so I taped a short length of duct on to make the correct Lv.

The stock design as mentioned above is 9 cu ft net and has fibreglass on all walls except the front baffle.

For this test I removed all the fibre glass to arrive at a net volume of about 8.2 cu ft 3 as provided by the sumulation for a driver Qts' of .282 allowing for crossover choke dcr and cable and amplifer damping. I made the education assumption of a fairly air tight and stiff enclosure with Ql losses equal = 10

It seems to work fairly well. I was surprised how well the simulation determined the new tuning. Subjectively its a bit of a harder but firmer bass quality and I suspect this is due in part to the removal of the lining.

According to Bullock a properly tuned enclsoure should not need damping material on the walls if used only for bass. While I agree in principle, the internal walls of mdf are reflective and the woofer is still operating to a degree above 500 hz (@12 db slope at 320 hz). There is also the issue of standing waves.

So I will attempt to line the walls during the week with some different types of damping to test the effect including pvc backed rubber underfelt and bubble pack which is apparently very good for reducing wall pressure induced standing waves . The idea being not to effect the enclosure volume like fibreglass but offer a reduction of internal noise which would otherwise find its way out via the large woofer cone.

I also did a simulation of axial room modes and I have significant pressure modes at 43 hz, 86hz and multiples of this.

If anyone has any ideas on attempting to deal with this please comment.

I plan to buy SoundEasy for some proper modelling and simulations shortly.

Macka