Computer speakers Thought experiment help

So here I am getting tired of my current computer speakers.
Which BTW are some cheap shielded Sony woofers in a tin can filled with BAF.

Using the drivers I have on hand I am thinking of a small 3-Way
Jaycar CW2135 6 inch woofer
Which according to the Jaycar catalogue has a 6dB peak at 2k.

I also have a quad of cheap but reasonable Full-range 80mm drivers from a HTIB system for which I have no data but were used full range with a tweeter that used a 1.5uF cap and a pair of the Audax Mylar tweeters.
Initial thoughts are a small sealed box WMTM and XO points at approximately 400 and 6k

questions, will I be OK with just a second order L/R XO at 4oo or should I also use/need a notch filter at 2k?
Should I cross over the Audax tweeter lower and use a second order on that with a first order on the mids. Bearing in mind I want to keep the cost down so the cheaper and simpler the better this would be


  • 20130126_0170.JPG
    186.4 KB · Views: 200
  • 20130203_0205.JPG
    144.9 KB · Views: 198
What if you added a little Tripath chip amp from eBay? Then, if the levels between the 6s and 4s are quite different, you can adjust easily.

Be sure the 6s and 4s are physically acoustically isolated from each other.

Do you have any way to measure actual frequency response and impedance curves? If not, crossover design will be a total crapshoot while wearing a blindfold in a dark coal mine. That's not to say the results will be bad, but you can't just use textbook values for real-world crossovers.
Approximate impedance curves are close enough for this speaker.
After all most woofers are at or near their DCR around 300/400Hz and I can always use Zobels if needed, the LG have a natural roll-off that should work with a tweeter cut high enough and I know that a 2uF capacitor works fine with the little 12mm Audax.
More a question I guess on whether or not to use a small sealed box or go for a bigger box and rear port it

Quick testing tells me the 2 * 80mm in parallel should equal the woofer almost
From what I recall (i.e. I could be wrong, it is 15 years ago) of the Audax tweeter (tw10**?) I think it suits a 1st order hi pass (combining to produce a 3rd order acoustic roll off), there is probably no reason why you couldnt go 2nd order, however.

The slight peak of the woofer at 2k is more than 2 octaves away from the Fc @ 400hz. going 1st order here may just be OK, although you're correct in thinking a notch may be necessary. Try both? I have a similar woofer for a similar application, and im using (or planning to) a XO a little higher up (~600hz) for similar reasons (hoping to avoid the notch)
Last edited:
yeah data on those audax is a bit sparse, but there is some still around. All i recall is that there are several interchangable faceplates for the same 10mm dome and neo motor and formerless VC. Their Fs is 2.8k from memory, 5k is an absolute min, just for some reason i seem to recall that recommended, OR that it combined nicely to give a near perfect 3rd order roll off. 2nd order would be safer, i agree, but at low-ish SPL it may be a non issue
Last edited:
Yes, I've also recovered the 'Tw010' file from my HD :eek:
Fs is 3 KHz has ferrofluid; these are the data for suggested xover circuit
Fc ------ S -------- L ------ C ------ P

6KHz 6dB/oct ----- - --- 3 uF ---- 25 W

6KHz --- 12 ---- 0,12 ---- 3,3 ---- 45

I've also did some WG experiments, indeed I'll use a little one I've made ( from elliptic tv speakers), but the tweeter is more like an on/off device rather
than one 'rich and full of nuances' :rolleyes:
To append my previous post I should clarify that my less than 'awe inspiring' experience of the 10mm audax was formed in my youth (and my lack of knowledge) and in my using the recommended cap for 1st order filter.

The 2nd order filter, or a 2nd order filter at 5 or 6 k could sound miles better.

Ive described my impressions in another thread some time ago. I found a one note sound, like a boomy tweeter, if that makes sense. Now i look back and realise it was likely due to hearing the fs through an inadequate filter.

In short, maybe 1st order isnt as good an idea.
Last edited:

Cascaded 1st order series might work well with those x/o points
with L-pads for level adjustment. Attenuation of mid and treble
will effectively vary the amount of BSC with 1st order x/o's.

rgds, sreten.

FWIW the Audax's are far better than the shedload of cheap
lookalikes used in car audio and budget massmarket audio.
Last edited:
Moondog, I'm using 2x 2135's per side in the wife's HT system, one of Jaycar's better offers... I'm crossing them over to 2108s 1st order at ~ 500Hz with no problems. Despite that graph, these poly cone drivers drop off fairly rapidly some where around 1600Hz IIRC, the 2108s which came out at about the same time go up to 4KHz nicely, but can't handle as much power low down as the 2135s.

FWIW here's the parameters I've got in Unibox: Fs 51, Re 6.6, Qm 3.01, Qe .58, Sd 123, Vas 16.8, I'm using them in ~48L for two drivers tuned to the mid 40s
Last edited:
Mmmh...I was to say something that strongly denied the use of cardboard
When suddendly I remember that it's the base of my speakers !
There are some very thick kind of cardboard , some are used as wood substitute. I found some out of a motorbike shop, and it was used as packaging
and carrying (transport) a big and heavy motorbike .
Some PVC sheets (found in shop for architect students) to cover the uglyness
of dirty pulp ...