Strange FR dip in the midbass - what could be causing it?

[fjhuerta]

I have this very strange problem. I don't believe I've ever seen it...



There's this dip in the midbass, centered at around 340 Hz. I don't see what could be causing it.

I'm using a vented box, 20L, tuned at around 35 Hz, with an Audax AP210ZO woofer. I padded it with 1" foam all around.

The dip is very obvious when using a wide measurement window, and when doing a nearfield measurement on the woofer.

I can't understand what's going on. I don't think it's a diffraction-related issue.

Can it be that the box is not properly damped?

EDIT: I just noticed my carpenter didn't use the right measurements for the box. Two dimensions are identical - 28 cms wide x 28 cms. deep x 39 cms. tall, using 3/4" MDF. Could this be part of the problem?

[Jed]

How long is your gate when doing the nearfield measurement?

I was first thinking it could be a floor bounce cancelation, but since it is there in the nearfield, it might rule that out.

My suggestion is to run an impedance sweep and see if there may be some energy storage issues, or surround to cone mismatch. It's not unusual for "floppy" cones to exhibit a dip like you have pictured.

[jbateman]

The width and depth are approx 1/4 wavelength of the frequency of your cancellation.
To see if this is the cause, remove the driver and make a nearfield measurement.

[fjhuerta]

How long is your gate when doing the nearfield measurement?

I was first thinking it could be a floor bounce cancelation, but since it is there in the nearfield, it might rule that out.

My suggestion is to run an impedance sweep and see if there may be some energy storage issues, or surround to cone mismatch. It's not unusual for "floppy" cones to exhibit a dip like you have pictured.

Hi Jed!

I believe I used a 60 ms gate. I also thought it was floor bounce cancellation, but when I saw this issue in the nearfield, I knew I had an issue. The impedance sweep is identical to the Parts Express one, and there aren't any issues I can see.

Jbateman:

My next step will be to do as you say - make a NF measurement of the driver only.

I just noticed one thing, though:



This is not the first time something like this has happened to me. This is the graph of an RS125 / RS28a MTM I built. It also has that weird dip, although higher up in frequency.

Now I'm really wondering what I'm doing wrong, so as not to do it again!

[Jed]

Is that a spliced FR of the woofer with nearfield and farfield data? If so you need to add delay to the phase and baffle step to the nearfield portion of the response. Looks like you are not aligning them properly. I can tell because of the knee in the phase where you would likely splice the 2 responses.

[fjhuerta]

Is that a spliced FR of the woofer with nearfield and farfield data? If so you need to add delay to the phase and baffle step to the nearfield portion of the response. Looks like you are not aligning them properly. I can tell because of the knee in the phase where you would likely splice the 2 responses.

Hi Jed!

Yes, I didn't align the phase on the spliced FR. Reason being, I first tried to understand where the dip was coming from.

Here's the nearfield measurement, with the dip clearly visible.



At this point I'm suspicious of the box dimensions. What I'd love to understand is - why? In order to avoid it next time!

[Jed]

We'll know better when you do a nearfield of the woofer free air. Then, do another of a woofer you know does not have the dip in the response.

[fjhuerta]

Yep. Will do such a thing tonight, Jed. I'll do a free air measurement of the Audax, and then I'll place a Dayton RS225 in the enclosure.

[Jed]

Yep. Will do such a thing tonight, Jed. I'll do a free air measurement of the Audax, and then I'll place a Dayton RS225 in the enclosure.

Do a free air of the RS225 as well. That will clarify if the dip is a measurement artifact in your gear.

[augerpro]

I'm betting that is just the normal response of the Audax. Those aerogels have a dip just before the cone breakup-which is very low. After using an inductor for BSC that's exactly how I would expect the response to look.

[fjhuerta]

Hi Jed!

I don't think it's my measurement system. I had my mic calibrated, and this is a NF plot of a Seas P21 I took in-box a couple of months ago.



Nice. I love SEAS! :T

Brandon, it's possible you are right. I know the Audax aren't the most precise of all drivers. But I'll know for sure tonight.

They sure do nice, though. I'll post pictures of the cabinets tomorrow.

Thanks a lot for your help!

[jbateman]

From Audax archives, for what it's worth:

[fjhuerta]

I see a dip... about the same size as mine. And there's something strange going on at 400 Hz on the impedance plot, too.

Is it safe to say this isn't one of the best performing drivers out there?

[Hdale85]

Very safe to say They were cheap and was popular for a cheap design.

[fjhuerta]

I took the driver out of the box, and...



There's that dip. The original one is a bit higher up in frequency, and it's a bit more noticeable.

The grey line is a damaged Dayton RS180 (slight dent on the cone), which performs flawlessly.

What do you think? IMHO, the problem is not the box... it's the driver!

[norcad]

Is this a problem? Can you hear it? Without the measurements, would you still say that the lower midrange is weak? I understand that you want to find out why, but now that you know maybe you could try live with it?

[fjhuerta]

I'm not completely sure about its audibility, since I haven't yet built the crossover (parts are on the way!).

This was a project that used leftover pieces from other ones - the RS28a I'm using is, in fact, the first driver I ever bought! I'm waiting for the big inductor to arrive, and a couple of caps. I wanted to try several things (a wide baffle vs. the narrow ones I commonly use, a second order acoustical crossover (impossible to do with the Audax, apparently), a wool felt mask around the tweeter (great results!)).

I *suppose* the Audax will sound "warm" "tube-y", etc., since that's how a friend who got me into tubes describes their sound. Accurate? Not so much He'll probably love it, though (I'm beginning to believe, after measuring some of his pre-amps, that he's addicted to 2nd order distortion), and I'm thinking that maybe he may like to swap the finished speakers for some of his gear.

I had a gut feeling the Audax wasn't the best of speakers, but I wasn't expecting this kind of performance. I didn't take any harmonic distortion sweeps, but I hope a 1.4 KHz crossover is low enough for them. According to my friend, he used to have a pair of 8" Aerogels, and they sounded "wonderful" with a "first order" (ie - an inductor, and who knows what acoustical order) crossover at 2.5 KHz... yikes.

[Curt C]

I took the driver out of the box, and...
There's that dip. The original one is a bit higher up in frequency, and it's a bit more noticeable.

The grey line is a damaged Dayton RS180 (slight dent on the cone), which performs flawlessly.

What do you think? IMHO, the problem is not the box... it's the driver!

I would guess that male vocals will sound a bit 'thin' with that dip left in. If it were me, I'd be tempted to reduce the value of the woofer inductor and shelve down the woofer response above the dip to try to minimize the abberation. IMO, a little 'extra' bass below the dip won't sound as bad as the vocals voiced incorrectly.

C

[fjhuerta]

Hi Curt!

Do you mean "increase the value of the woofer inductor"? Decreasing it would shelve up the woofer response above the dip!

[Curt C]

Hi Curt!

Do you mean "increase the value of the woofer inductor"? Decreasing it would shelve up the woofer response above the dip!

-To compensate for the shelving filter you add. :B

To be honest, I may be full of applebutter, here. :E

Rather than conjecture, I'd need to model it to see what combination of components would actually work best. If you'd like, email me the .frd and .zma files, I should be able to look at it tonight. My address is on my website.

C

[John_E_Janowitz]

Do you have an impedance curve of the driver? that will tell you if it's the driver or the enclosure. Could very well be a spider resonance if there is a dump in the impedance curve at the same spot. we are dealing with this in our 6.5" driver now, but it accounts to only about 1-2dB of dip now that we properly damped the resonance. Looking at the published curve from the audax driver, I do see a bump at 400-500hz which could likely be it. If so, you might be able to lessen it by properly damping the spider more.

John

[dlr]

Do you have an impedance curve of the driver? that will tell you if it's the driver or the enclosure. Could very well be a spider resonance if there is a dump in the impedance curve at the same spot. we are dealing with this in our 6.5" driver now, but it accounts to only about 1-2dB of dip now that we properly damped the resonance. Looking at the published curve from the audax driver, I do see a bump at 400-500hz which could likely be it. If so, you might be able to lessen it by properly damping the spider more.

John

I'd be interested to learn what methods are recommended for this, including damping compound, application method and location, etc.

Dave