Tuning a sub with test tones...

Discussion in 'Home Theater Projects' started by Rob Formica, Oct 20, 2003.

  1. Rob Formica

    Rob Formica Stunt Coordinator

    Joined:
    Feb 20, 2003
    Messages:
    225
    Likes Received:
    0
    Ok, I've got a simple and dumb question about tuning a sub. I decided early on that I'd tune my sub using test tones, so that it would be as close to the truth as possible. I'm running test tones through it and it's not obvious at which frequency the driver moves less at, so it's got me wondering if I'm doing it properly or if it's out of the range of my test tones?

    The subwoofer is a Stryke AV15 in a 6ft3 box (pretty close once I remove the volumes of the driver and port) with about [email protected] of polyfiber. The port is currently 23" long including the two ¾" flares (3/4 + 21.5 + 3/4). The test tones I've produced go from 12.5Hz to 50Hz in 1/2Hz steps most of the way.

    I noticed a small change in excursion around 18Hz... but this seems to be present regardless of my port length. I tried the 23"(above) as well as 24" and 29" ports. The 29" port seems to have a small excursion change around 16Hz and 18Hz ... while 24" and 23' only had the 18Hz change.

    How small change in excursion is it supposed to be?

    Thanks...
    Rob
     
  2. ThomasW

    ThomasW Cinematographer

    Joined:
    Nov 6, 1999
    Messages:
    2,282
    Likes Received:
    0
    To check the port tuning you need a digital multi-meter, an amp, a tone generator, and a ~100 ohm 20 watt power resistor.

    Put the resistor in series with the + lead going from the amp to the woofer. Wire the negative lead as usual.

    Now send 50Hz sine wave into the speaker at low volume. Then with the meter set on "AC" measure the +/- terminals of the woofer. Set the output from the amp so the meter reads between 1&2 volts.

    Now lower the frequency in steps 5Hz at a time. Each time write down the reading from the meter. As you go down in frequency there will be a point where the voltage bottoms, then starts back up again with the next lower frequency. The lowest meter reading is the post tuning frequency.

    For example:

    50Hz-1.2V
    45Hz 1.3V
    40HZ 1.2V
    35Hz-1.1V
    30Hz-1.0V
    25Hz-.75V
    20Hz-.5V
    15Hz-.75V
    10Hz-1.1V

    In this example the port tuning is 20Hz. That's where the meter reading is the lowest voltage.

    Obviously you can use steps smaller than 5Hz for greater accuracy.

    If you need a tone generator and have a computer close to the amp. You can use the output of the soundcard to make the tones. Do a google search for "NCH tone generator" without the quotes. This is a great little freware program.
     
  3. JamesFarina

    JamesFarina Agent

    Joined:
    Dec 8, 2001
    Messages:
    30
    Likes Received:
    0
    Hi Rob,

    "How small change in excursion is it supposed to be?"
    This all depends on how loud you are running your tests
    If you are running the sub at modest levels you should be able to notice where the tuning point is located. It will be harder to tell at lower levels.

    Here is how I do it.

    At the tuning point the driver will not move very much, but a tremendous amount of air will be coming from the port. What I usually do is have someone help me when trying to find the tuning point. I gently touch the surround and have someone else adjust the frequency until the driver moves the least. I usually start at about 15Hz and slowly work my way up, then when I feel minimum movement from the driver I have my helper increase the frequency a few more steps, then backdown again just to be sure. It is also helpful to have some type of software program to give you a general idea of the tuning point.
    Hope this helps.

    jim F
     
  4. Rob Formica

    Rob Formica Stunt Coordinator

    Joined:
    Feb 20, 2003
    Messages:
    225
    Likes Received:
    0
    Thanks guys... that is just the info I needed. [​IMG]

    I think the excursion wasn't very obvious because I am doing the tests at a relatively low level... using a low powered sansui receiver in my workshop. The sub weighs about 120lbs, and I wasn't up to bringing it into the living room alone. So I generated my tones using NCH and burnt them onto a CD and played them through the old (easy to move around) receiver.

    I have a digital meter... and I will get myself a high-power resistor to have a measured test, hopefully tomorrow. [​IMG]

    Thanks again...
    Rob


    PS: Correct me if I'm wrong, but adding the series resistor is to increase the voltage (or decrease the current)? Would placing the digital meter in series and reading amperage without the resistor work to given the amperage is probably under 5A?
     
  5. JamesFarina

    JamesFarina Agent

    Joined:
    Dec 8, 2001
    Messages:
    30
    Likes Received:
    0
    hi Rob,
     
  6. JamesFarina

    JamesFarina Agent

    Joined:
    Dec 8, 2001
    Messages:
    30
    Likes Received:
    0
    Rob,

    Sorry about the last reply.(accidentally hit enter)
    Adding a series resistor will lower the voltage but the current will stay the same.

    Voltage divider law Vx=(Rx/Rt)*Vt
    Rx=component resistance
    Rt= total resistance of circuit
    Vt=total source voltage
    Vx=voltage drop over component.

    Jim F.
     
  7. Rob Formica

    Rob Formica Stunt Coordinator

    Joined:
    Feb 20, 2003
    Messages:
    225
    Likes Received:
    0
    Ok, I picked up two 200ohm 20W resistors (connected in parallel to give me 100ohms), and placed this in series with the driver and receiver.

    Obviously this little receiver had a hard time putting out a signal in 104ohms [​IMG] so the volume was quite low... but these are the numbers it gave in my temporary mockup:

    12Hz --> 1.2 V
    13Hz --> 1.1 V
    14Hz --> 1.0 V
    15Hz --> 0.9 V
    16Hz --> 0.8 V
    17Hz --> 0.8 V
    17.5Hz -> 0.8 V
    18Hz --> 0.8 V
    19Hz --> 0.8 V
    20Hz --> 0.9 V
    21Hz --> 0.9 V
    22Hz --> 1.0 V
    23Hz --> 1.1 V
    24Hz --> 1.2 V
    25Hz --> 1.3 V
    26Hz --> 1.5 V

    but no obvious drop at tuning... my guestimate would be between 17 and 18Hz. Since I'm aiming around 20Hz, I can safely trim off 1" and retest it. Just funny that it doesn't drop below the 0.8V but and my multimeter doesn't offer greater precision in AC volts.

    I wondering if this may be due to the amplifier being pushed a little hard? Maybe drop the resistor to 50ohms [​IMG] ?

    Thanks...
    Rob
     
  8. JamesFarina

    JamesFarina Agent

    Joined:
    Dec 8, 2001
    Messages:
    30
    Likes Received:
    0
    Rob,

    The amplifier isn't being pushed very hard at all. 0.8V across the driver, thats about 0.16 watts, the amp is coasting. The less resistance causes the amp to be "pushed" harder

    I would try the method I previously discussed with feeling the surround. Make sure to play the driver at a higher level. Just be careful not to leave the driver play too long while testing for the tuning frequency! Here's what you do; set your generator at 10Hz and make sure the driver has about a total excursion of 0.5in, slowly increase the frequency until the driver moves the least. There is your tuning frequency. Very simple, no need for high price voltmeters.
    Hope this helps.

    Jim F
     
  9. Rob Formica

    Rob Formica Stunt Coordinator

    Joined:
    Feb 20, 2003
    Messages:
    225
    Likes Received:
    0
    I was afraid you were going to say that... [​IMG] as my workshop setup must put out a maximum of about 25W to 30W/chn. When pushed hard would give me the 1/2" excursion at12Hz (once the enclosure is unloaded)... and nowhere close to that at the other frequencies. My original, relatively low level, testing did seem to have a slight excursion dip around 18Hz (see my first post) but not as obvious as your recommended higher power testing... as it looked pretty similar at 17Hz and 19Hz.

    If your wondering why I'm not using my subwoofer amp, it is that it's missing in action somewhere in St-Louis. I do have a 100W power amp in my music setup, but I was hoping on not to have to move the wall unit to get it out. It should get me closer to the 1/2" excursion.

    I know... I know... lazy...
    Thanks for your insight...
    Rob
     
  10. JamesFarina

    JamesFarina Agent

    Joined:
    Dec 8, 2001
    Messages:
    30
    Likes Received:
    0
    Hi Rob,

    If you want to give me all the dimensions of your enclosure and port I can plug the info into winspeakerz and we should know the tuning frequency within 1-2 HZ.

    Jim F
     
  11. ThomasW

    ThomasW Cinematographer

    Joined:
    Nov 6, 1999
    Messages:
    2,282
    Likes Received:
    0
    The testing method I posted gives an accurate measurement of the tuning frequency of a ported box. And it's accuracy isn't contingent on using high dollar test gear.

    From the results of your first test I'd say you have a box tuned to ~17.5Hz.

    You may want to check the range settings on your meter, most will give a 3 digit readout. So you may have it set to measure too high a voltage. You should be able to get readings like 0.81, 0.82, etc
     
  12. Rob Formica

    Rob Formica Stunt Coordinator

    Joined:
    Feb 20, 2003
    Messages:
    225
    Likes Received:
    0
    Ok, I pulled my main power amp out, and tried testing with it... 100W visually helps with a little more excursion. Definitely at 18Hz with a 22" port, and it remains at 18Hz with a 21" port, and 20" port (I'm still using the 1/2hz increments).

    I also compared that with the voltage and an SPL meter with similar results with each method.

    I tried the SPL meter at about 1" from the cone as someone else suggested, since at the least excursion and that close, the air pressure should noticeably drop... it does by about 6db to 8db.

    WinISD estimates a 22" port for a 20Hz tune for my 6.03ft3 box (11778in3 – 350in3 for driver – 500in3 for braces – 500in3 aprox for my port)... but i stuffed 8lbs of polyfill into it (7.75lbs on the scale).

    Just seems odd that the measured tuning remains pretty constant with the small port changes...

    What I believe is happening is one of two things:

    - that the polyfill is increasing the apparent "acoustic" volume of the box... and if I estimate a 30% increase... the port length would have to change substantially for each 1Hz increment... which would explain why it remains pretty constant with each 1" I cut off.

    - the box is not air tight (ex: the port isn't glued in) and the fs of the actual driver is becoming the dominating factor.

    What do you guys think?... I may leave it at 20" and just finish the box and re-test for curiosity when I'm done. [​IMG]

    Thanks again...
    Rob

    PS: Thomas: my digital meter actually does read 4digits... but the AC voltage scale offered is based on either 200V (one decimal point) or 2000V (no decimal points)
     
  13. ThomasW

    ThomasW Cinematographer

    Joined:
    Nov 6, 1999
    Messages:
    2,282
    Likes Received:
    0
    No wonder your tuning is so goofy. IMO you've got WAY WAY too much poly in the box, and it's without doubt messing up the port tuning. Sorry I missed this info in your first post!

    The large amount of poly is the reason the port length adjustments aren't changing the tuning frequency.

    It's not a good idea to 'stuff' a ported box to increase the 'apparent' volume. One 'stuffs' sealed boxes. One lines the insides of ported boxes with a layer of damping material, start with a layer 1"-2" thick.

    Now recheck the tuning frquency......
     
  14. Rob Formica

    Rob Formica Stunt Coordinator

    Joined:
    Feb 20, 2003
    Messages:
    225
    Likes Received:
    0
    I just wanted to post to thank you guys for your help... [​IMG] as I think I've got this tuning figured out.

    I removed most of the stuffing, and re-ran tests with drastically different port lengths... from 6" to 22". It showed me that theoretical and practical results could vary quite a bit.

    What I originally forgot to take into account was the "non-linear" relationship between port length and tuning. :b If I do the calculations for different port lengths in any given box... we see that for each single change in tuning frequency requires substantially longer and longer ports the lower the frequency goes. Or in other words, a 1" change at 17hz range may represent 1/6Hz while it'll represent over 1Hz in the 22Hz range. :b I posted some of my results in this table and note that the voltage values are arbitrary and only relevant to the one port length.

    http://www3.sympatico.ca/bonaz/Klips...-tune-test.htm

    Well WinISD was a little farther of from the correct port length that I expected... as it's recommended 22" port should been about 16" long. After playing with the numbers, I've decided to go with 17.5" port for approximately a 19.5Hz tune... as this should give me a pretty flat response. [​IMG]

    Thanks again for helping me out, I honesty appreciated it...
    Rob.

    PS: I'd recommend anyone who wants to tune a box with test tones to try the resistor/voltage meter method... very accurate, and doesn't upset the "significant other" with loud test tones. I ran several tests at each port length with different voltages and averaged them out to help identify the minimum impedance. easy
     

Share This Page