Supravox 165 GMF in a closed box?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I currently have a 165GMF in a TQWT. It is a very nice speaker when driven with a high impedance SET.

However, even if there is amble 'bass', it is not lineair, and certainly it is colored.
I did a test with a 20 HZ sine at 1 watt. Open TQWT: this was very audible, thus it must be third harmonic. Closed off, a volume of 38L: totally silent. The speaker just moves.

Now my 'problem': using the various calculators on the net, I get a prediction of a resonance frequency of some 86-89 Hz (Speaker Fs 65, Qts 0,333) at a Q of 0,46.
On the other hand I measured a resonance peak at 72 Hz in the closed box.

On mh-audio I calculate the EBP as 1,89; it says for the Efficiency Bandwidth Product of 90 or greater = best used in ported enclosure.
On Petoin a closed box also is not recommended because the Qtc is too low (his criterion: Qtc should be >0,6).

And indeed, the speaker without compensation starts to drop at some 180 Hz or evenhigher, @-3dB. But Linkwitz come to the rescue.

When I read the Linkwitz-Riley documentation, and use the calculators, the parameters and outcomes are just fine (as long as I use a very low target pole of say 22 Hz); I can add a LPF of 25 Hz to protect the speaker.
filter gekozen supravox Monte Carlo.jpg

Am I safe to go on with a LR filter or should I abandon the closed box?
So far I use the measured Fres @70/72, not the calculated one for 38L @85/89.
Use the measured Fres?
 

Attachments

  • LINKXFRM-2 Supravox.xls 72.pdf
    59.6 KB · Views: 64
Last edited:

GM

Member
Joined 2003
1st, to run any sims you need to add the wiring's resistance and SET's output impedance to Qts, which should account for the lower measured resonance, so the short answer is "use measured", it always trumps theory! ;): mh-audio.nl - Home.

That said, using an LRT to trade BW for efficiency quickly uses up driver power handling, so combined with a high Fs, and little Xmax to start with it may be too SPL limited to please on some recordings unless you listen at typical TV levels, so typically it's better overall to use baffle step compensation [BSC] to shelve down the mids - up BW to tonally balance it out in room: General Speaker Related Articles

GM
 
Last edited:
Thanks, you are right. I used a Le Monstre amplifier and that has a damping factor of something like 4-7. I will factor it in and see what comes out of the box sim.

Right now I am enthousiastic about the capabilities of the LRT to flatten the response; a crossover is a next step.
I have a Lomo Kinap 2A9 15" for the lows, and plan also to make a closed implementation with LR transformer circuit. That is a very very low Qts driver, 'impossibilistic" so to say. Characteristic should fit the Supravox openness.
 
I redid the measurements, both with a 51 ohm resistor and a 16 ohm one.
The box is now hermetically sealed (port closed).
It exhibits a slightly lower resonance point than previously measured.
As regards the closed box predictor, the Qts can rise from 0,33 to 0,439 with a series (output) resistance of 1 ohms.

You are right regarding the risk of an overdrive. First of all of my SET @ 8 watt.
It is just I want to see if the TQWT I have emits colored lows.
 
Hi Thanks for thinking along.
I made a big mistake in thinking it was 70 Hz - the dial made max deflection there. :blush:
Entered values in Excel: no it was a dip . . the peak is at 110/120 Hz . . . :rolleyes:
165GMF closed.jpg

The box is tapered, so the returning wave is totally damped. Around the speaker i have some damping, and all internal sides are stuffed with (thick) felt.

Seeing the resonance at 210 Hz I need to damp the front of the cabinet better. Maybe that is that area (18 cm internals width and 7 cm deep is not the most suitable enclosure)
 
Interesting to see a closed box with a 65 Hz speaker have the main resonance at 44 Hz. .

- seeing this curve, what would you suggest as a starting point for a Linkwitz transform?

The 110 Hz peak? I have no idea about the Q there. The closed box calculus suggests something like 91 Hz/Q=0,46 starting from the parameters calculated by the manufacturer. So I am lost here again.

I currently have two high pass filters of 4 dB, one at 70 Hz the second at 250 Hz. Sound is rather good already, better equilibrium than with the TQWT.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.