[dhruvmeena96]

Reliability issues again: I built several pairs with knowles GV drivers, all of them broke after a few weeks of use. Out of 6 GV (3 pairs) I bought only two are still working.

It’s always the bass that breaks. On 3 of them, the bass driver just went silent, one of them the bass started to make a snaring sound.

I use them with open vents.

What could be the problem, I‘m like burning money? Did any of you use the knowles GV without issues over a longer period of time?

Noticed the same problem with my builds. Lows starting to snare and die after a few weeks. What „glue“ do you use for the tubes? Also use it with vents open, but only tape side.

The crazy thing is, sometimes the bass starts working again...
So it must be an electronics issue?

Its better you get them as different driver, understand the crossover of GV
and then make it yourself

This way, you can use good quality caps and resistor, and debug them as you want

The only fixed driver i support is GQ, SWFK, TWFK, DWFK

the driver with single nozzle
my experience with them was way better than
GV, GK, HE versions

i too broke one GV

Bass rattling problems

it happens when the driver is pressurized due to something
it can be damper, it can be uncontrolled driver electrical damping

it can be anything

And this always happen when signal is too powerful and driver aint that tight in some factors

the pressurized air inside BA hits the diaphragm, making it bend unwanted, causing it to rattle

Compression Horn speaker also has this issue

 

[dhruvmeena96]

@Furco @Shilohsjustice @piotrus-g

i need 4.7uF cap

is this good

Its a Nobiom oxide cap

 

[eunice]

The bass of the GV is HODVTEC (I think), so these issue (pressurizing, overpowering) should also happen with a single hodvtec?

 

[dhruvmeena96]

The bass of the GV is HODVTEC (I think), so these issue (pressurizing, overpowering) should also happen with a single hodvtec?

Thats why you can dampen it individually with L pad and Z circuit(You know, what Z circuit i am talking.....Zobel....knowles provide proper data for HODVTEC and RAB/RAF)

Then you can create proper low pass, which will take the over bursting pressure the HODVTEC is creating.

If calculation goes correct by my side, it can go deeper than planars and dynamics while playing safe but at cost of 8dB attenuation and 5dB return loss

but instead of TWFK, you have to use DWFK

it would be HODVTEC and DWFK, damper on DWFK(But if you dont want dampers, i can help you)
You would have to design the crossover

and i can eliminate the dampers from design now

 

[imsurokim]

How do these look?

The base looks perfect. The end of the ear canal on the right one still seems to not have the correct angle, you will have to manually account for that.

As a tip: You will have to round the edges where you cut the ear canal, for easier insertion. But you don’t need to do it right now, better sand the edges of your first shell after you inserted the tubing - you will see how much space you need for the tubing.

Right ine meaning right one on pic or right ear piece? Sorry I screw up pics. Right ear piece is on left on pics which still has piece of red blocker.

 

[eunice]

Right ear piece is on left on pics which still has piece of red blocker.

I meant the right earpiece, the one with the red foam at the tip.

 

[imsurokim]

Yeah. I would either cut left ear piece to match right one or build some on top of ear tip. I am still waiting for fotoplast to come. Since it has shorter ear canal it fits more comfortable in the ear

 

[Ivan TT]

i was shocked myself that a 4xRAB setup can go this low(Lower than dynamic driver, way lower.....something like LCD4)

Changed my 4 driver RAB from all series to 2x series paralleled, bass port cloth removed.
This makes it sit better with the output stage of my DAP utilising OPA1622 (9V rails may not have been providing enough voltage swing for series connection, but 1622 has really good output current and is designed with 32Ohm earphones in mind).
The bass is deep, mids are fullbodied, highs are silky (but not too hot, just as I like it).
Will try Zobel later time permitting, the plan is to use breadboard and trimpots allowing for easy adjustment of values (no offence @dhruvmeena96 but in my experience practical outcome is often not what calculations suggest it will be, but I really hope that this will not be the case this time).

 

[Lucas Taboada]

First I used dreve sk47 silicon glue. Then I switched to loctite super glue and then cover with silicon glue. The silicon glue is not stick enough.

Still, always the same problem. I‘m kind of devastated, as I started to do ciems for friends and family and they all break. First I thought it was me dropping them, or the fiio lightning cable and so on. But that can’t be, as they all break.

The GV sound so nice.

You are using the Loctite 495? As I said before, I'm experiencing "flumes" with that glue in closed ambients (gluing faceplates). And yesterday I glued tubes to some drivers and put in a ziplock , today the drivers had a lot of white stuff.. It's really weird because this glue wasn't supposed to cause this. In your case, could be the glue that is killing the driver? I don't know... I have some GVs here and I'm scared of assembling them now

 

[Bassiklee]

When I assemble "motors"(driver assemblies with tubes, dampers, etc), I use loctite 495, but VERY sparingly. One drop on a plastic bag, then use a toothpick to put a tiny bit of it on to hold the tubes on the drivers. I pre-assemble the tubes, and let that glue gas off before I introduce the drivers. I even leave the tape on the vents on the HODVTEC side of the GV until the glue has long since gassed off. Assembled 22 of them today.

 

[Lucas Taboada]

When I assemble "motors"(driver assemblies with tubes, dampers, etc), I use loctite 495, but VERY sparingly. One drop on a plastic bag, then use a toothpick to put a tiny bit of it on to hold the tubes on the drivers. I pre-assemble the tubes, and let that glue gas off before I introduce the drivers. I even leave the tape on the vents on the HODVTEC side of the GV until the glue has long since gassed off. Assembled 22 of them today.

I see, so Loctite 495 maybe cause this in closed situations only and not every time like normal superglues? If I let it gas off for some time before storing in a ziplock it will be ok I guess?

SR driver with the white flume


What type of tubes do you guys use in the GV by the way? 1 or 2mm inner diameter? I'm a bit worried about fitting two 2mm ID tubes in the canal. Using 1mm ID will cut too much HF?

 

[Bassiklee]

Gassing off, yes. What I do is put down some parchment paper, and lay the motors out on that for a few hours. Nothing sticks to parchment.

Tubes: I start with 1mm..... Secret recipe.

 

[dhruvmeena96]

Changed my 4 driver RAB from all series to 2x series paralleled, bass port cloth removed.
This makes it sit better with the output stage of my DAP utilising OPA1622 (9V rails may not have been providing enough voltage swing for series connection, but 1622 has really good output current and is designed with 32Ohm earphones in mind).
The bass is deep, mids are fullbodied, highs are silky (but not too hot, just as I like it).
Will try Zobel later time permitting, the plan is to use breadboard and trimpots allowing for easy adjustment of values (no offence @dhruvmeena96 but in my experience practical outcome is often not what calculations suggest it will be, but I really hope that this will not be the case this time).

Zobel will dampen the driver treble, but will also load the amp more properly, so amp side distortion figure drops(pure resistive and since driver is working less on treble part(which is easy to get wrong on figures due to high speed movement) clears up distortion figures in mids and treble

The 1kHz distortion noted by @Rabid86 was due to the usage of different driver and the circuit for RAB. Damping factor became to high and due to high level reduction of treble which made mids seem to look more distorted.




Add a series resistance of 18~33 ohm, to underdamp the circuit and increase voice coil resistance.

This will get the setup to 40~55ohms, but this will technically dampen the mid bass and increase treble while attenuating slightly.

Etymotic uses this method to increase treble and lower bass.

Well, I think, you should create a breadboard zobel and variable series resistor(with values) with 2x3.5mm female jack and switch.

If I explain zobel, the cap decide the low pass frequency for whereas the resistor decides the way it will slope

Values are between Re(500Hz for BA) and Re x 1.25 for the taste point of view


By the way, the bass was deep, mids were fullbodied and highs were silky was the review of series parallel or the whole series?

 

[eunice]

What type of tubes do you guys use in the GV by the way? 1 or 2mm inner diameter?

I use a very soft 1mmID 2mmOD at the driver (length approx 2mm) . Then 2mmID and 3mm OD through the canal.

I have the impression it reduces some resonances if I do it like that.

Regarding glue: what to use then? I also have UHU which is better for gluing wood but worse for gluing the tube to the drivers. UHU also creates the white flume sometimes. I tried SK47 silicone glue which doesn’t create any flumes but also is not sticky enough to attach the tubes.

 

[eunice]

Assembled 22 of them today.

Wow. Let us know if any of them fail after some days of use.