View Full Version : 2425 diaphragm installation... and sub1500s

08-18-2004, 02:43 PM
Hello everybody, I'm new here. Liking the forum a lot particularly all the sub1500 stuff. I'm an electronics with music tech student and my final year project is going to be one of these drivers in a motional feedback system - building the boxes at the moment. Will post back some updates if anyones interested.

Anyway, onto the actual question. I've recently brought some new genuine JBL diaphragms for my 2425s. They currently have clones in and i've been noticing some irritating distortions arround 1.6-2K particularly obvious with stuff like coral music. Eventually tracked the problem down to VC rubbing. Tried fiddling with them and moving things about for ages; it got better but never really went away which is why i got the JBLs. I've replaced a fair few comp driver diaphragms in my time but these JBLs and the problems iv'e been having made me think a little more about my technique. They come with a note in the box saying:

"this part is not self-centering, this product should only be installed by an authorised factory trained technician. centering of the voice coil within the gap requires specific test equipment and training and is critical to the performance of this product"

Is this mostly BS to make you go spend the money on getting a dealer to do it? Do they actually have specific test equipment or do they just do a frequency sweep or somthing? Is there any perceived wisdom for installing these things or do i just shove it in like i always have before? It seems to work ok with ladies..... :-)

cheers in advance


08-18-2004, 03:47 PM
Centering the Dia is easiest while playing a signal of sine with f around 500HZ.
Distortion is easy to hear.
You CAN center a dia with extensive measuring equipment but most of the dealers do it the easy way by ear.

Mr. Widget
08-19-2004, 10:01 AM
Welcome aboard demanding one.

"Will post back some updates if anyones interested. "

You bet. What type of motional feedback are you going to use? I find the type of servo circuit that Velodyne uses sounds pretty awful above about 60Hz.

"...with stuff like coral music."

Do you make your own underwater recordings? :smthsail:

Sorry, I just couldn't resist. :banghead:


08-19-2004, 10:38 AM
"...with stuff like coral music."

Widget you should be ashamed of yourself, that one was just too easy!:moon:

Re. motional feedback this topic has always fascinated me. You can use piezo elements from appliance buzzers/old motorola tweeters etc. attached to the dustcap to sense velocity and G's.

As long as the woofer can handle it, it's possible to get amazing performance.

08-20-2004, 06:20 AM
Aint you never heard them coral sing Widget? beautiful harmonies!

The technicallities of the project still need to be worked out, I am very much still in the research phase. Essentially what i want to try and do is create a PA sub using motional feedback which to my current knowledge no one has tried yet. I think the drivers have just about advanced far enough and power is cheap enough now to give it a try. Previously you just wouldn't have got the output levels needed. I want to try and get flat bottom octave performance from a 2ish CUfoot box at around 120dB. This may be a bit of a tall order but i'll have fun trying!

I'm almost certainly going to use an accelerometer for cone position sensing. I've spoken to a few people who have tried other approaches like back emf measurment or using dual voice coil woofers with one coil driving and the other sensing but they have said that these don't work fantastically well. Accelerometer seems to be the general consensus. I did toy with the idea of using laser range finders which would have been very cool but i dropped this fast when i discovered how much they cost! It wouldn't surprise me to fine that the Velodyne's problems are caused by a cheap accererometer, many of them start to get inaccurate as the frequency goes up. Iv'e found some analog devices ones which calain to be good to 400Hz which will be way out of passband and hopfully ok. They are still expensive though; around $100. I'm going to try for some free samples!

The general concept is to use the compliance of the air in the cab to stop woofer mechnical damage (calculate volume to prevent cone excursion going over Xmax with full power input). You present the input signal to the cone then measure its movement with the accelerometer to see if has moved far enough to reproduce the signal faithfully. if it hasn't you feed it some more power untill it has. Obviously you put some limiters in the chain to stop it all getting out of hand and hopfully you get solid bass to 20Hz at 120dB.

This should be fairly simple to impliment in analogue so i'm looking into setting myself a challenge and going for full digital control over everything. This may be an unrealistic goal in the time i have; as i say i'm still in the research phase.
I'd be very interested to hear form anyone who has any thoughts on or experience with motional feedback. The more info i can gather at this stage the better.

Got the first box going yesterday jsut to get some idea of performance and to find out how much extra bracing i might need. It leaks loads at the moment! Sounds lovely at low power - very musical an deep. At high power you cant really tell cos the leaks take over. Its a bit down on output from what i had hoped for, I guess im just used to 2242 high efficency type stuff in big boxes at the moment. There is still a lot of tweeking and leak fixing to do it may get lots better. The second box should help too. Its all designed to match my current mids and be a pretty stack. The picture shows the sub and the main drive rack. The MacroTech is driving the 1500 bridged - gives it a good steady 800w with over 1000w burst power.

08-20-2004, 07:44 AM
I think I recall seeing an article in Electronic Design magazine where the author used a piezo element inconjunction with some op-amps to derive acceleration. It was was back in the eighties sometime. You also might want to search the AES papers, there is bound to be something there.

Sorry I can't be more specific. IMHO this could be a winner concept.