diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Full Range (http://www.diyaudio.com/forums/full-range/)
-   -   5dB dip at 120Hz in my speakers? (http://www.diyaudio.com/forums/full-range/108657-5db-dip-120hz-my-speakers.html)

cjv998 17th September 2007 08:19 PM

5dB dip at 120Hz in my speakers?
 
I am currently using a pair of Paradigm Mini Monitor v5 bookshelf speakers amped with a LM3875 gainclone from Peter Daniel's site, crossed over via my computer's soundcard (M-Audio Revolution 5.1, crossed at 80Hz) to a Hsu STF-2 10" ported sub. I noticed what sounded like a dip in frequency response from approximately 115-140 Hz. I downloaded a software-based frequency-generator/phase-shift control program, and played around with it for a while to try and deduce the cause of my problem. (A link to the program is below if anyone is interested...it's free, by the way).
Anyway, while playing around with the phase-shifts, I noticed that it sounded like my speakers were 90 degrees out of phase at those frequencies (adding in a 90 degree phase shift seemed to increase volume at those frequencies). I could be hearing things with the phase shift correcting my problems (the more I listen to it, the less confident I am about that being the problem). However, I definitely do have a 5-6dB dip around 120Hz. I have the speakers in an 11'9" by 12'2" room (on the long wall, room's resonant frequency is ~93Hz, if I remember correctly), with the back of the enclosure 1' from the rear wall (the speakers are rear-ported...maybe I could use that to boost my dip in frequency response?). The speakers are about 4' from the side walls, also (I should probably move them a bit to keep them from dividing the room into thirds like that as well). So, any suggestions as to what's going on and how to fix it?


Here's the link to that frequency generator program:
http://www.allapp.com/Two-Channels-F...ownload_30980/

Okay, I think I'm going insane. I moved the speakers around a little, and thought I put them back in the same place. Apparently not, because the 5-6dB dip shifted up 40Hz, so it's now around 160 Hz. It seems a little less severe as well, maybe only -3dB. If anything, they speakers may be 1" further from the wall, and the toe-in is probably different, but I didn't really have it measured before I moved them unfortunately.

Nelson Pass 17th September 2007 09:11 PM

Have you thought of floor-bounce?

:cool:

cjv998 17th September 2007 09:28 PM

The speakers are on 30" tall MDF stands. The sub is on spikes on the carpeted floor, obviously. Could you elaborate a little on what floor-bouncing is? I'm not quite familiar with the term.

Scottmoose 17th September 2007 11:19 PM

2 Attachment(s)
Floor-bounce is an almost inevitable occurance with any loudspeaker in a domestic environment. The frequency at which it occurs is determined by the height of the driver above the floor. This might be a bit disjointed as it's late & I've matchsticks holding up my eyelids, but here goes... The drop in SPLs is caused by the reflection of your driver's output on the floor (which is out of phase with the initial output) bouncing up and interfering with the forward output. The reason it doesn't all vanish is down to a few things. Time delay partly -the reflection is always slightly delayed compared to the initial forward output of the driver. Also, signals are rarely perfectly 180 degrees out of phase with each other, so while you may get some attenuation, it's not total.

The only way of getting rid of the effects, short of using active Eq to boost that region ~flat, is to put a huge amount of damping material on the floor in front of your speakers to attenuate the reflection down to a level where it's not going to cause any harm. The carpet helps compared to a hard, polished surface, but the LF wavelengths where it occurs are really too long to be too bothered by a piece of Wilton or shag-pile. We're taking about a foot-deep mat of fibreglass or dacron, here at least, to entirely kill it. As this isn't very practical, most of us just live with it. Unless very severe, floor bounce is fairly innocuous. Our hearing isn't great in this zone (see attached Fletcher Munson curve) so we just tend to key off the peaks, and psycho acoustic masking fills in, or causes us to ignore, the missing bits.

Scottmoose 18th September 2007 08:52 AM

I knew I should have been more careful when I wrote the above. This is classic evidence that you shouldn't do anything when you're tired and not thinking right guys. I suspect it's because I was also thinking about an idea for a dipole at the time, and the brain finally packed up. Nurse! Bring the electric apparatus and the drugs! :o

There is no phase change when sound travelling in air strikes a medium of higher acoustic impedence (although there is when sound travelling in a solid encounters a medium of lower acoustic impedence). However, the reflection of the driver still interfers with the forward output of the driver, attenuating it.

Nelson Pass 18th September 2007 08:27 PM

Quote:

Originally posted by cjv998
The speakers are on 30" tall MDF stands. The sub is on spikes on the carpeted floor, obviously. Could you elaborate a little on what floor-bouncing is? I'm not quite familiar with the term.
That's approximately the geometry which will give you the
dip you describe. Your big clue would be a peak at about twice
the frequency of the dip.

badman 19th September 2007 02:06 PM

One solution
 
Not really viable here, but!

Once solution is to use multiple bass drivers in a vertical array to disperse the floor bounce frequency.

cjv998 20th September 2007 10:39 PM

Quote:

Not really viable here, but!

Once solution is to use multiple bass drivers in a vertical array to disperse the floor bounce frequency.
So I guess that's why tower speakers have multiple midbass drivers? Makes a lot more sense now. I figured it was just to make them louder and more impressive-looking, but the floor-bounce issue explains it I guess. Anyway, I should have some time tomorrow to listen for a peak at around 250-300 Hz. If this is an issue with floor-bouncing, what can be done to fix it (if anything)?

ronc 20th September 2007 11:07 PM

If this is an issue with floor-bouncing, what can be done to fix it (if anything)?

EQ or tilting the speakers back.

ron

tinitus 20th September 2007 11:08 PM

I cant say really, but your 80hz xo point could be suspected to have some influence on this


All times are GMT. The time now is 06:05 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2