Instructive and sad story about LCD-3 modification attempt
Jun 6, 2015 at 2:26 PM Thread Starter Post #1 of 54

cleg

Headphoneus Supremus
Joined
Oct 14, 2011
Posts
1,572
Likes
4,791
Preface. On russian forum I've met excellent DIY maker with nickname Snorry. He makes really excellent headphones, using old soviet TDS-7 and TDS-15 as a source of drivers, but also he is a big fan of headphones modding. He's approach based on objective measurements, not only subjective listening. I think I should tell more about his works later.

Recently, he wrote his opinion about Audeze's flagship LCD-3 and his modification attempt. I've thought that it's an interesting experience, and I've tried to reinterpret it (as much as my poor english allows me to do it) and post on Head-Fi, as I think it's extremely interesting. So, all direct speech below, it's not me, but Snorry.

I've recently got LCD-3 Fazor for modification. It was provided by other person, he understand all risks of losing warranty and possible damage of such expensive headphones. Also I want to thank to Audeze engineers for interesting model.

Technically, headphones are interesting, but there some things in their sounding that I don't like. I'll try to use my experience and tune headphones. I won't mind, if Audeze engineers will take some of my notes for consideration.

So, what should be improved.


  1. Owner asked me to do something with high frequencies, being "too harsh" for him on some records, but he wanted to keep bass intact. On my first listening, they seemed to be really fatiguing on highs, especially on female vocals. I've made measurements that showed pretty noticable channel imbalance on highs.



    Without smoothing


  2. On my pretty big head, headpads became too squeezed, and Fazors starts to push my ears with their edges, it's not comfortable.
  3. Appearence. I don't like this black grill esthetically and acoustically (more on this below)

I'll post more information during the mod desription.



To be continued.
 
Jun 6, 2015 at 2:46 PM Post #2 of 54
First inspection showed some possible reasons of highs imbalance.

1. Earpads
They've appeared to be different in height and softness.



On photo above — measurements in thickest place. 5 mm of difference looks too much.

If you push cups towards each other, you can see that they have different hardness



Earpads should be definitely redone. With stock ones, channel imbalance is inevitable.

Also, I'd like to notice, that only outer part is made from natural leather. Inner parts are made from average quality leatherette.

Even accurate and slow attempt to pull them off, causes material stratification.



2. Grill
Unfortunately, grill's gaps aren't aligned with magnets and fazors.



Why is it bad?

Flat diaphragms of planars are very sensitive to multipath sound reflections. If you place your palms opposite grills closer then 30 cm, you'll hear sound changes. And outer grill and fazors gives much more reflections, and this causes some interference on high frequencies.

In this implementation, fazors on outer side are useless, moreover, they make sound worth.

So, grill should be replaced.

3. Damping
Old LCD-3 had damping done with some kind of felt. Now, it's some king of white paper.



This makes assembly simpler, but also gives parasite reflections to membrane. So, to tame highs a bit, this should be also replaced.

Important notice. This paper is glued with two-side duct tape, so remove it without leaving traces — impossible. Here you'll lose your warranty. Headphones owner gave me permission to remove it, so I'm moving further.

To be continued...
 
Jun 7, 2015 at 8:32 AM Post #3 of 54
So, first step of modification is grill. For a long time, in my mods I'm using grille made from stainless steel, fluted across the magnets. It's acoustically transparent and lets air flow freely. So, it drastically reduce parasite reflections, as there is no parts, parallel to membrane. Everything is angled, no direct reflections. I'm working on receiving the patent for this element of my mods.

This grille should be attached in some way. I want to thank Audeze engineers, screws, they're using made grille attachment easy.

I've tried to make new part in common style with zebrano, used for cups. It's hard to get zebrano in our country, so I've used composite, made of 16 layers of red wood and oak.

That's how it looks.

 
Jun 7, 2015 at 8:39 AM Post #4 of 54
I've looked to torn earpads, to rests of duct tape...



... and thought that it would be a masochism to try different earpads models. So I've made two simple details.



They attach to the same holes that used for driver attach, just take screws 4-5 mm longer.



Now, it takes just few moments to replace earpads.



I got no idea, why Audeze doesn't make it in stock headphones. Use two sided duct tape for headphones that cost about $2000, is strange.

Additional benefit, this element adds about 5 mm to earpads depth, so fazor doesn't touch my ears anymore.
 
Jun 7, 2015 at 8:52 AM Post #5 of 54
Now about fazors

Fazors are cheap plastic details, painter from one side, to make it looks like metal



They are mounted, using the same two-sided scotch, and on one ear, fazor's grill was about 5 mm displaced, compared to magnet system of headphones. It's one of the reasons of highs imbalance.

Then I've measured width of fazor and magnets.



And width differs. This gives additional source of parasite reflections.



And on the outer side, it's really hard to align fazor properly, as magnet system grill is closed by damper. So, it's pretty easy to get something like this.



And it can be not 100% parallel...

Inner fazors can ba controlled by sight, so they were aligned correctly.

So, I've decided to use needle file to make fazor's bars same width as magnets.
 
Jun 7, 2015 at 9:09 AM Post #6 of 54
Before tearing fazors, I've make a closer look to drivers.

1. Nope, front fazors are also aligned not perfect.



2. Membrane surface is dirty. I've found some hairs stick to it (green arrows), and some rests of black sponge (red arrows)



I think sponge was used to remove leftovers of glue, but some parts of sponge struck.

As for membrane construction.

You can't see wirings itself, they're under magnets. Membrane is reinforced with foil squares, gaps between them are translucent



3. Second driver is even more interesting. You can see same problem with fazor alignment, here it's even more noticeable, and mismatch grows from left to right



4. You can see dust between membrane layers.



5. Edges of membrane aren't perfectly glued



6. One of drivers have problems with parallels



Some conclusions.

- hairs and sponge will be removed
- most sound impact cames from bad fazor and membrane alignment.
- dust and bag edges glue can't be fixed, but they didn't impact on sound much. They can cause some durability problems, here membrane can exfoliate, but most probably it will be in distant future.

To be continued...
 
Jun 7, 2015 at 10:55 AM Post #8 of 54
I've obtained the consent for drivers disassembly from headphones owner. Here are the photos of membrane



And now, the sad part.

Don't ever try to disassembly drivers. Also, don't try to unscrew the contacts, for example for rewiring.

Contact spacers and screws are glued, so any attempt to unscrew them causes irreversible damage of contact pads. I've disassembled old versions of LCD-2, and they didn't have any glue, so I've assembled them back without problems. New LCD models have disposable drivers.



Scratches near the contacts pads are from my attempts to get to aluminium and fix contacts. In this case it's impossible. Aluminium is under the polymer layer, so it's impossible to attach to wire.

This was the end of this modification attempt. LCD-3 surprised me a lot, so I've decided to share my observations, hope it will be useful.

Translator's note: Snorry will appreciate any feedback and can answer your questions, I'll translate them to him.
 
Jun 7, 2015 at 8:14 PM Post #9 of 54
This is awful, it makes sense why I had to replace the left driver on my LCD-2 four times. I babied the heck out of them and they would always go out on me ~4-5 months after getting them back. 
 
Jun 7, 2015 at 10:53 PM Post #10 of 54
This makes me sick to the stomach I've had to send mine back 2 times for failed drivers. Is this a recent thing or have they been doing this for a while?
 
Jun 7, 2015 at 11:03 PM Post #11 of 54
Not very good manufacturing practices.
 
I don't find any distortion in the higher frequencies on my classic (non fazor) LCD-3. So I'm really hoping I don't get the dreaded driver failure. I don't think I would like the new fazor version.
 
Jun 8, 2015 at 1:16 AM Post #12 of 54
My main curiosity is whether the "beat up" condition of the diaphragm in the two exposed diaphragm photos, are due to the aggressive method of ripping apart the magnet arrays, or are they actually like that (all scratched up) even under the magnets?
 
Jun 8, 2015 at 2:26 AM Post #13 of 54
I've PM'd the Audeze HeadFi account to invite them to respond to this thread. As the owner of two LCD-XCs and a new LCD-3 I can say I've zero problems aside from a couple fitment issues that were trivially remedied. Audeze is famous/infamous for continuous improvement so I have no doubt that if these issues represent ongoing flaws rather than a one-off QA miss they will be fixed.
 
Jun 8, 2015 at 3:39 AM Post #14 of 54
My main curiosity is whether the "beat up" condition of the diaphragm in the two exposed diaphragm photos, are due to the aggressive method of ripping apart the magnet arrays, or are they actually like that (all scratched up) even under the magnets?


I'll clarify with Snorry, but most probably, there was no any "aggressive" attempts to rip apart magnet arrays, everything was made very gently and accurate (but it doesn't helped to avoid contact pads destruction).
 

Users who are viewing this thread

Back
Top