Simple Subwoofer Level Setting

Introduction
Sometimes it can be hard to find the correct setting for the crossover between subwoofers and main speakers. It was when I tried a more scientific solution the history of this page started. The first attempt to make a simple SW level setting method was based on frequency sweep and I think it worked quite well. Not everyone owns a function generator and because of the room resonances this method has it flaws. After some valuable feedback from Ulrich in Germany, here is a new method based on the simple principle of two out of phase signals canceling the sound output.

The method can be used if you have a crossover with an adjustable delay function like the MT Audio Design XO-2001. I have tried the method on my ESL-63/SW-63/XO-2001 setup, it works very well and adjustment by ear is no problem at all. Apart from the crossover with adjustable delay you also need a test tone, in this case a 100Hz sinus wave (delivered with each XO-2001 on a CD).

Description
1. Set the delay potentiometer at 2'30.
2. Connect the SW and main speakers out of phase, use only one channel.
3. Run the 100Hz test tone on your CD player on repeat.
4. Adjust the delay position to the point where the least sound is produced, don't sit too far away from the speaker (the room interacts). If the minimum is found in the maximum delay position, switch the polarity on your SW (or main speaker).
5. Adjust the SW level potentiometer to the point where the least sound is produced.
6. Repeat the delay and level adjustment until you can not get a lower level.
7. Switch the polarity on the SW (or main speaker) to get them in phase.
8. Play some music and enjoy.
9. You may need to adjust the SW level to get the kind of sound balance you want and your room needs. If you play at low levels, a higher SW level may be preferable. The delay is now perfectly adjusted and no further adjustments are necessary.

With the SW and ESL connected in opposite phase (low pass filter in XO-2001 is inverting) the correct delay is around 3 o'clock. The SW-63 is then a part of the ESL-63 delay line and the bass integration is perfect. Setting the delay electrically like this is NOT the same as adjusting the distance from the SW relative to the main speaker. When the delay is electrically adjusted to be correct, the phase is correct for frontal distribution of sound and also for the rear!

The accuracy of the method can be calculated from the graph below, kindly sent to me by Ulrich in Germany.

Ulrich's own comment:
Please see the attached JPG-file. It describes the accuracy of the phase and amplitude alignment between subwoofer and ESL. If you use my method of inverting the sub and playing around with level and delay you can achieve up to 40 dB cancellation at the crossover frequency of 100 Hz; at least for one fixed listening position. Using the graph in the file now tells you the max error in phase and amplitude with the attenuation as a parameter.

Example: you get -30 dB after adjusting the level and delay on the XO2001. Then you can see that you are somewhere on the -30 dB circle and the max phase error is ca. 1.5°, while the max amplitude error is ca. 0.2 dB. That means an almost perfect match of the sub and the ESL at the crossover freq. As you can see the method I described is very sensitive and leads easily to results, because even if you don´t measure the notch you can estimate the depth of the notch by ear if you have a calibrated level setting on the amp or the typical -20 dB muting function. Again measuring with a simple mike is the better method, using one of the analyzers you mentioned; the quality of the mike is uncritical.

For those who have ESL-63/SW-63 with original crossover
When I had adjusted the level properly with my first method (frequency sweep), I measured the 100Hz output at both the ESL-63 and the SW-63s binding posts. With the 4 ohm connection alternative the level at 100Hz is two times higher on the ESLs binding posts than on the SW-63 terminals. With the 16 ohm connection alternative the levels shall be the same. Using Ulrich's method with out of phase signals on my ESL-63/SW-63/XO-2001 setup resulted in lower level on the SW-63, I got approximately one third of the level on the SW-63 (4 ohm connection). Gradient have stated that the voltage sensitivity of the SW-63 (16 ohm connection) is the same as on the ESL-63, with 2/3 of the level (16 ohm connection) on the SW-63 and original Gradient crossover, I think the bass level will be a bit low. The final adjustments must always be made ear, but it is a good starting point to get the levels within a dB or two. You only need a multimeter and a function generator, cheap multimeters are readily available. Not everyone has a function generator, but many have computers with soundcards. These are very useful for measurements and below you can download a free function generator for use with your computer, this can be used for the level adjustment described above. A free spectrum analyzer is also available below and you can have some fun with this. If you need a more advanced measuring program you can download the WinAIRR Demo Software from AudioXpress.com, this is fully functional and the shareware program is not too expensive either. The 'Analyzer 2000' can also be used for soundcard measurement.

Download Function Generator
Download Spectrum Analyzer
Download Analyzer 2000
RadioShack Digital Sound Level meter
 

Final adjustments
When you are done with the level setting you can try moving the speakers into different positions, this can make quite a difference as this example from Chris (received in e-mail) shows:
I want to relate an experience I had yesterday regarding the setup of my Quads with the 63 sub, using the stock Gradient crossover. My room dimensions are 16' x 24'. The Quads were along one short wall about 8' into the room, and about 4' from side walls. Speaker center to center about 7', seating position 8' from speakers, about 7' into the room. Frequency response tests were done using 1/3- octave warble tone at -20db for the following frequencies, 200hz,160,125,100,80,63,50,40,31.5,25,20.

The second column in the table above shows the result after speaker repositioning.



 

This site uses frames, but if you have found this page via a search engine and do not see any frames, you can use the button below to get to the main frame of the MT Audio Design site.