finding xmax

You can't really get Xmax from physically moving the cone of a 'speaker, plus the frequency it occurs at for a given power input changes when it is in a cab.....

There is a method of measuring Xmax using the distortion of the 'speaker as this is one method used by AES and some companies to measure it.

Use a sinewave and look for the distortion to rise to 10%.

Now this appears complicated but with 10% distortion it's fairly easy to see on an oscilloscope; so if you can get hold of a half decent dual trace one and a low value resistor to put in series with your 'speaker to measure current; then you look at both the voltage and current waveforms and see where the distortion rises at each frequency you test it at.

Two things here, with the driver in it's cab you can find where excursion rises to set your High Pass filter and what your max power is to not exceed Xmax to set your limiter.



This is a pic of a 'scope displaying this, a bit more than 10% distortion perhaps but you should get the idea.....The current is the center waveform with the distortion and the larger sinewave is the Amplifier voltage output.

If you are interested in this method and have access to a 'scope and a sinewave signal generator, I can give you more detailed info on how to set up and run these tests.

Hope this assists
Mik

(Edit: correct spilling)



Edited by audiomik - 17 March 2012 at 10:59am

Tom
with so many different drivers quoting different figures for similar specs - I wouldn't like to hazard a guess. However it certainly looks like you've done your research pretty thoroughly.
Low cost drivers in PPSL or isobaric loading can give some really impressive results as well.

As to PC based 'scopes, there are a number that are dual trace using L & R inputs so that won't be any real problem. The frequencies you will be looking at are fairly low - below 200Hz so wide bandwidth isn't an issue.
Various free software programs will do this.
A separate sinewave signal source is easy enough - single channel, even one of the CD's with a series of tones feeding your test Amplifier with short tracks of fixed frequencies run as a 'loop' - don't forget if you use this method to list the track numbers against the frequencies.
You'll need some external attenuators to not overload your soundcard input with an Amplifier output - see below.
To do the current sensing, a 0.1ohm 10 watt resistor used as a 'current shunt' will work OK, so you look at the voltage across it and multiply this voltage by 10 for the actual current in Amps.
Also you can then get an impedance curve from your measurements to compare with sim software results plus some idea of what is happening to the phase relationship between current and voltage into the 'speaker - Power Factor.

My recommendation is to use this test method with your low cost 'speakers so when (if) you get some expensive ones you will not harm them running the same tests.

Test setup pic:

(apology for poor quality - but a quick sketch for a simple diagram!)
The actual value of resistor R1 you'll need to find a bit by trial and error; so have given a 'starting value' which you simply add another of the same value in series until you can eventually run your test Amplifier into clip with no load without overloading your soundcard input.
Anyway, that's the 'easy' bit; interpreting your results is a bit more difficult but straightforward once you have run a few tests. Just look for a change in the current waveform shape which is the distortion.
It's directly related to the coil leaving the magnet gap, Xmax, which then changes it's impedance so this is what you see as the distortion in the current waveform on the 'scope/PC display.
PC software with more functions can show this better, for example using FFT analysis of the current waveform to see better when the distorton rises by looking at the increase in harmonic content.

Don't forget that excursion is very dependent on cab design and loading of cones, so even with 'known' types things will change.

Anyway, looks like you've an interesting project, hope this above assists
Mik

Edited by audiomik - 17 March 2012 at 12:43pm

Tom
Yup, fairly basic Ohms law stuff as you say for getting the Voltage and current measurements into the soundcard - can be calibrated using a DMM if you really want to but the results we're looking for are relative.....

OK, the Current trace does look the same as the Voltage trace until distortion starts but it varies in Phase depending on Frequency above and below resonance, plus tends towards very low Phase difference at higher frequencies. You also get a drop in current to a minimum at resonance - for some types of cab you will get two minimums in current - notably Bass Reflex - which is an old method of testing for port tuning.....
Some actual results on this variously here in this thread:
http://forum.speakerplans.com/power-factors-for-speakers_topic52128_page1.html
for 'speakers not in boxes...... same principles apply but the box will affect the results as the cone is loaded in different ways. Also a somewhat different test setup.
This thread link also shows what phase differences to expect.

Now "guestimating" 10% distortion:
This is around the minimum which is easily visible on a sinewave so any errors won't be too far off hopefully. Increase your Amplifier output voltage a bit if you think you have reached this point and distortion rises dramatically. Be careful doing this for obvious reasons.

Didn't actually measure the distortion of the Current waveform in the 'scope image posted earlier as it was originally intended to demonstrate the need of squarewave testing for Amplifiers into resistive loads to best simulate real conditions. Anyway the 'speaker was well exceeding its Xmax

Do measurements for both loaded or unloaded drivers as the mood takes you! My tests are all unloaded as haven't the space here for various different boxes.
For loaded boxes they are useful in checking that you won't run into over-excursion so also useful for setting HP Filters/EQ, max input voltage (don't really like to refer to power as it needs impedance taken into account), etc.

As to being 'dedicated' - well these days doing things like this and other sorts of 'non-published' result tests are my hobby so keeps me sort of 'sane' - also interested at present with isobaric loading.
There is a thread somewhere on here about trying it with a pair of 21" drivers, also to try using one driver to 'load' the other to see what happens to currents in voice coils and perhaps simulate different cabs,  and a post on FSP about using this loading in a tapped horn.....
With loads of time on my hands, some of these 'off the wall' experiments can be fun - often just don't know what will happen till I try it

Not sure my responses are that detailed though - haven't written up a full Research Paper for some years
Mik