Distortion problem?

[BossBunos]

Yesterday I reviewed a rew graph I made quite some time ago (1y+) and for 'fun' checked the distortion graph of the sweeps. I noticed there was quite a large bump at the right speaker in a particular region (1.6 - 1.7 Khz). Since these sweeps were taken so long ago and in the far field (3.5 meters away) I tought it wouldn't be much of a real problem but I decided to verify anyways. First I checked the screws of the driver. I noticed 1 screw had some scratches on it, I bought them used. I decided to take the drive unit out and check if maybe something inside was touching the driver. Nothing special to see there besides alot of stuffing and surprisingly thick wireing. Nothing touching the driver. So I screwed it back on tight (but making sure not to over tighten).
Then I started a sweep measurement from approximately 80cm away from the driver for both left and right speaker. Results posted below (Right speaker, left speaker). And it's basically the same result as a year ago.

 

[Sokel]

Did you check the x-over,1.7K seems low for the tweeter.

 

[BossBunos]

Yeah the crossover is at 2.3kh to the tweeter. So I checked the driver below the tweeter

 

[Sokel]

Yeah the crossover is at 2.3kh to the tweeter. So I checked the driver below the tweeter

Ok,now check the x-over for anything weird.
Could be a loose cable,etc.
If everything is ok it's probably the driver,try to push it in a little gently and evenly by it's sides,it should go in smouthly without anything feeling like scratching.

 

[BossBunos]

The crossover makes sense doesn't it. Since the distortion continues in the tweeter (2.3 Khz up) and it's unlikely that both drivers are damaged?

 

[BossBunos]

I took out the driver again to inspect the crossover. Happens to be mounted in the area below the driver .
Then I took out the driver of the well performing speaker and noticed 2 things. The well performing had some sort of filling under the gasket as shown in the first two pictures.
Going back to the distorting driver I noticed that the screw mounts were coming loose in their fitting.
So before I investigate further I will fix the screw holes with some glue and even try to add some padding at the same locations as the non distorting speaker. Will be after a couple of days though because tomorrow it will be kingsday here and probebly gonna be unable to do something useful the day after.

 

[FeddyLost]

So before I investigate further I will fix the screw holes with some glue and even try to add some padding at the same locations as the non distorting speaker.

If this will not help, you can check separate harmonics in range of interest, not a sum of them.
Sweep from 1K to 5K with max resolution and most probably you'll see some prominent peaks.
After that you can go even narrower and then check enclosure for vibration or noises with e.g. stethoscope.

But loosened screws can vibrate without doubt.

 

[DonH56]

Definitely fix the screws and see if there is a gasket or sealant missing on the bad speaker. Another thing you might try is physically swapping the position of the two speakers to make sure it is not affected by room placement or the electronics chain up to the speaker. If the problem stays with the speaker in its new position, you can power everything off, swap speaker connections at the amp, and see if the problem is still with the same speaker to rule out the electronics.

HTH - Don

 

[Mario Sanchez]

Might I suggest taking the tweeter out to test if the driver itself is complaining?

 

[BossBunos]

Definitely fix the screws and see if there is a gasket or sealant missing on the bad speaker. Another thing you might try is physically swapping the position of the two speakers to make sure it is not affected by room placement or the electronics chain up to the speaker. If the problem stays with the speaker in its new position, you can power everything off, swap speaker connections at the amp, and see if the problem is still with the same speaker to rule out the electronics.

HTH - Don

Good suggestions, those will be my next moves. I will first try the electronics swap and then the actual moving of the speakers since te salon2 is quite heavy (66kg+).

@Mario Sanchez removing the tweeter is a good idea as well but that will be one of my last options since the drivers are soldered. And my soldering skills are very limited.

 

[DonH56]

Good suggestions, those will be my next moves. I will first try the electronics swap and then the actual moving of the speakers since te salon2 is quite heavy (66kg+).

@Mario Sanchez removing the tweeter is a good idea as well but that will be one of my last options since the drivers are soldered. And my soldering skills are very limited.

Did not realize you had Salon2's. Nice speakers, but I am biased. In lieu of a swap, you could pull the bad speaker a couple of feet toward the middle of the room to get it further away from any boundaries and see if that changes the response. I strongly suspect the problem really is the speaker, but these are simple things to try first (sanity checks). The idea for moving the speaker is to see if there is some boundary reinforcement (SBIR, reflections) increasing the amplitude of certain frequencies.

I assume you do not have any sort of room correction or EQ that might also influence the response...

The Salon2's are EOL but still available so if there is a problem they should still have replacement parts.

HTH - Don

 

[BossBunos]

Will probably check these things on Sunday and will report back . Im running a minidsp SHD with Dirac but switched off everything (EQ+Dirac) before I made the sweeps. No corrections applied above 500hz anyway.

 

[BossBunos]

Update:
So I fixed the loose parts and let it set overnight. This morning I took some measurements again. I moved the speaker about 1 meter extra into the room and rotated it 90 degrees, making it face another direction.

New full sweep:

And to make sure it was not the electronics I swapped the speaker cables at the power amp:

Not sure what I want to get out of the next two charts but I applied a crossover in the minidsp at the original crossover point of the salon2 (2300hz):

As you can see the distortion is dominated by the 3th harmonic. Not sure where to move on from here.
It seems that fixing the screw cleaned up the distortion that went passed 3khz. The remaining distortion is in the crossover region.
I tried taking out the midrange driver so I could inspect the crossover but wasn't able to get it out. After unscrewing it's still pretty stuck. I used some plastic tools (to remove car interior stuff) but couldn't get it to come out. I'm terrified to damage the paint or speaker so didn't move on and re-inserted the screws.

 

[DonH56]

Good work!

I'd probably (try to) engage Revel at this point, starting through your dealer. They may have experience to point to a driver vs. crossover problem and help with replacement.

 

[BossBunos]

Yes I think that's the best thing because I'm not too confortable going any deeper.
As some final checks I actually moved the speaker next to the other one and swapped the speaker cables as well, no difference.
Also I tought I change the tweeter and bass setting on the back just to rule that out. But no change either.
Worst thing is that it's actually audible to me now after running alot of sweeps on both speakers haha. At least I know what 1~2% distortion at 2khz sounds like now.
Thanks for the help so far and I'll post a follow up when I know more

 

[Mario Sanchez]

You can examine the drivers in relative isolation without taking them out by making a measurement very close to the driver itself, maybe you can try that to see if the issue is in individual driver or is something more extensive?

 

[YSC]

I guess the tweeter could be scratching or something in the internals, it’s weird that one side have been tiered previously which likely means they had troubles that original owner want to take them apart

 

[BossBunos]

Contacted the distributor for my region and they suggested to play a sinewave at the distortion frequency ~1860 and investigate if its the midrange or tweeter.
Playing the test tone I could easyly hear the distortion, keeping my hand right before the tweeter removed the distortion. Decided to do an RTA measurement of that sine very close to the tweeter and midrange (1cm/0.4inch). Left and right are overlayed. Obvious distortion in right tweeter compared to left. But also quite interestingly the fundamental tone is 6db higher on the broken side. Could this be a faulty crossover as well?

The midrange driver seems to be intact:

Playing a 6khz tone (eliminating crossover range) I got the result below. Seems like the level is equalized at this point but there is a strong third harmonic distortion in the right tweeter:

Tweeter looks pretty broken to me But the crossover might also be malfunctioning I gues.
I must say that during sweeps and tones it's quite audible. But during music playback I can't hear it... Shared my findings with the distributor and waiting for a response... to be continued