• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Strange noise from midrange

ozzy9832001

Senior Member
Joined
Feb 19, 2023
Messages
460
Likes
301
Hi all,

Having a strange issue with my Edifier 2850DBs. For the most part, I love the speakers, but after my latest round of room treatments one issue really, really stands out.

There is an awful sound that emits from the midrange on certain songs. The issue seems to center around 250hz, but extends down to about 175hz and up to about 350hz. Regardless of volume the noise is there and it's typically the same volume. It sounds like a chair being drug across a floor. It always happens during some songs. I think the issue was always there but I thought it would be correctly when I treated the room. Turns out its definitely something with the speakers, but maybe someone can give me a clue what I'm experiencing.

What I've tried:

  1. I removed the midrange from the speaker cabinet and held it in my hand and played offending song(s) -- issue still present. Slightly reduced and not quite as offensive.
  2. Inspected driver. Nothing seems out of place or damaged -- looking at the surround and cone under a microscope doesn't seem to show any mini holes are malformations.
  3. Plugged port.
  4. Moved speaker outside where the only boundary was the ground for 100ft.
  5. Reversed speakers and the right speaker also does it.
  6. Lowered gain on MINIDSP and increased speakers and vise versa (increasing speaker volume does make it more offensive, but increasing gain on MINIDSP does not).
  7. Changed power cable (worth a shot)
Investigating frequency range:
  • If I use my minidsp and cross over at 400hz, 24db/butterworth, the issue goes away completely. Obviously I've gutted the frequency range, so that make sense.
  • Placing a notch filter in PEACE at 250hz, quality 2, the issue goes away, but again the response is gutted.
  • Placing a peak filter in MINIDSP @ 250hz, -8, 2.5Q softens it significantly.
  • The odd thing is that if I do a sweep nothing shows out of the ordinary. Everything is very, very tight and the only EQ filters I use are the peak @ 250 and a few shelf filters.
  • The weirdest part is it's like the 250hz region is out of whack, like it's louder than it should be but it doesn't show on any sweep or test tone. Lowering it by 8dB actually seems to put it "in line", perceived loudness with everything else.
The only way I can seem to get it to go away is with extremely specific positioning.

If I stand up and walk about 8' away from the speakers and stand off axis opposite the speaker it seems to dissipate. So this would be the equivalent of having the left speaker on the right side of me, 8' toed away from me.

I contacted Edifier but I'm SOL since it's been more than a year since I purchased them.

Any ideas?
 
I'm thinking it's an electronic problem. Have you tried another sound source? It could be anything... Your phone or your TV. Plug it directly into the speakers (or use the Bluetooth) and bypass the miniDSP, etc., so only the speaker is the same. If the problem remains, it has to be the speaker and I'm guessing it's the amplifier inside the speaker.
 
It could be a VC rubbing, but I would think it would be as bad a low volumes.
Probably worthwhile to use an iphone app and push tones or sweeps through.
Also swapping the left and right would be worthwhile to rule out the source.
 
It could be a VC rubbing, but I would think it would be as bad a low volumes.
Depends, if it's deformed only at one end it might not show up with minimal excursion. VC problem was my first thought also.

@ozzy9832001 have you tried the speakers in a different room? If it's some kind of weird resonance it might not actually be the speakers at all.
 
I'm thinking it's an electronic problem. Have you tried another sound source? It could be anything... Your phone or your TV. Plug it directly into the speakers (or use the Bluetooth) and bypass the miniDSP, etc., so only the speaker is the same. If the problem remains, it has to be the speaker and I'm guessing it's the amplifier inside the speaker.
Yes. Happens when hooked up through bluetooth as well.

I've even taken the speaker outside in my yard where I have a massive expanse of open space.

It also happens with the other speaker if I reverse the stereo feed.

My original guess was the amp.
 
@ozzy9832001 What songs in particular are affected by this issue?
One of the more egregious songs is by Lovebites - Nameless Warrior. Towards the end of the first verse the guitars do a small gallop and during that you can hear it really bad.

It happens during both the studio and live versions, but worse on the Live track.

It's like omni present but does vary depending on the recording. Some tracks I think have the dynamic range where it's not so much of a problem, but others not so much.
 
Could you get a recording? I had a similar issue and it appeared to just be something inherent to the speaker: https://www.audiosciencereview.com/...s/is-this-distortion-or-something-else.55409/
I listened to your sound file that you posted in the thread and I don't think it's a similar issue. Though, I'd have to hear it with music or something playing. Mine is more like clipping.

However, I may haver stumbled across something interesting.

I can't seem to find any spinorama data for the 2850's, but I could the 2800, which is basically the same speaker with some minor tweaks.

The FR has a -6 dB dip from 250hz to about 800 and then a massive 6 dB increase at 4k -ish.

I'm wondering if they tried to fix the dip in the previous model by adding some sort of peak filter to it rather than just shelving the treble because more volume sells.

Under normal circumstances, it probably isn't a problem, but on more modern music where there is far less headroom, minor annoyances happen. Easy to just write it off as a bad recording.

The 2800 and 2850 are nearly identical, and it would make sense that a small tweak like that would net only a half numbered increase.

It also makes sense that I need to drop about -8 ish dB on the EQ for it to sound right.

As for the actual inside of the speaker, both look intact and like I said it's common to both speakers if I reverse stereo or change the input on the minidsp.

So the only expanation I can come up with is a signal chain problem, and gain staging is fine, something with the way or designed possibly with the crossover since it's at about 200hz, the amp or the power going into it.
 
@ozzy9832001 Maybe a dumb question, but anyway: how exactly you performed the stereo channels swapping test? Did you just switch channels at the input, or you internally re-wired the output of the amplifier?
 
@ozzy9832001 Maybe a dumb question, but anyway: how exactly you performed the stereo channels swapping test? Did you just switch channels at the input, or you internally re-wired the output of the amplifier?
Just the inputs on the minidsp -- I'm not sophisticated to do anything internally.

They are better though when I saw the channels at the mini dsp AND move the speakers. So, that's something. I haven't done any measurements, but it's possible some reflections have changed to make it less aggressive.

I can set the EQ to about -4.5 now instead of -8.

The weirdest part is: I can't even hear an audible difference with the lower EQ. Everything does sound like it "belongs" more though.
 
Just the inputs on the minidsp -- I'm not sophisticated to do anything internally.
Ok. This looks definitely strange to me. If I understand it right, affected is always only ONE side - left "by default", right if the input cables are reversed. This behavior would clearly point at the input device or data (not the amp) as the cause. BUT - you said that the issue manifests itself even with BT input and that says exactly the opposite - the amp. Or maybe the data itself..? This whole thing interests me because I have exactly the same speakers and so far I didn't notice any abnormalities in the output signal, even during some experiments with "loudness" (based on Fletcher-Munson contours) eq profiles etc...
 
Ok. This looks definitely strange to me. If I understand it right, affected is always only ONE side - left "by default", right if the input cables are reversed. This behavior would clearly point at the input device or data (not the amp) as the cause. BUT - you said that the issue manifests itself even with BT input and that says exactly the opposite - the amp. Or maybe the data itself..? This whole thing interests me because I have exactly the same speakers and so far I didn't notice any abnormalities in the output signal, even during some experiments with "loudness" (based on Fletcher-Munson contours) eq profiles etc...
I'm just very confused by the problem and maybe it really is something simple that I'm overlooking or a concept that I've misunderstood.

The issue I'm having trouble wrapping my head around is that it goes away with EQ and I can increase the speaker and/or windows volume (usually I stage everything @ 86dB with windows volume 100% maxed and then lower it to taste) and it doesn't come back. What I mean by that is: I can further increase the speaker volume past my configured 86dB and it won't become audible again.

Regardless of input source: I've tried BT, Toslink, and RCA. I've tried 2 different DACs the PCs internal soundcard and using the MINIDSP.

It really sounds like a clipping or compression issues, but I've tried lower both input and output gains on MINIDSP to no avail.

Currently, my chain is as follows: MINIDSP (USB SOURCE) to speakers. MINIDSP input and outputs both set to -0dbfs, windows volume 100%.

However, something I noticed today while doing some testing with BT because I was trying to just see what the speakers "real" volume was, is that BT is significantly louder than when it's hooked up to the PC or MINIDSP. Louder by nearly 10dB. Something along the path is attenuating the signal.

Maybe it's a windows 10 issue I don't know but it's strange for sure.
 
Is there an internal issue with the speaker? Is the mid-range driver properly isolated from the woofer (I.e; separate internal volume for each)?
Is there a really bad resonance in these boxes? They're fairly cheap so I'm wondering about internal bracing.
 
Is there an internal issue with the speaker? Is the mid-range driver properly isolated from the woofer (I.e; separate internal volume for each)?
Is there a really bad resonance in these boxes? They're fairly cheap so I'm wondering about internal bracing.
Regarding the driver isolation and resonances: midrange driver is isolated and additionally damped in its small box. Speakers themselves are quite heavy - weight of one box is approx. 12 kg. I have them too and I dare to say that they are very immune to resonances, even when blasting some "fat" heavy metal recordings..
 
Last edited:
Regarding the driver isolation and resonances: midrange driver is isolated and additionally damped in its small box. Speakers themselves are quite heavy - weight of one box is approx. 12 kg. I have them too and I dare to say that they are very immune to resonances, even when blasting some "fat" heavy metal recordings..
I don't mean to be snarky, but 12kg is pretty light for a box like this. It sounds like there are some serious resonance problems and this might be made worse by a poorly designed crossover. Does the box have any bracing? Does anyone have information on the crossover? Maybe the mid range has no high pass capacitors in the circuit or they are not properly filtering it and it's just being overdriven.
 
Bracing probably not, but as I already said, build is very solid and really far from vibrating like a cardboard box - you barely feel vibrations when touching the box even if blasting really bass heavy metal music. Plus as OP said, the issue is present even with midrange driver pulled from the box, prejudices aside, please. Crossover is purely electronic, DSP based. An amateurish try (not mine) to capture drivers in nearfield see below.
 

Attachments

  • 20230327_132059.jpg
    20230327_132059.jpg
    216.2 KB · Views: 88
Back
Top Bottom