Castles_Saloon

You're probably thinking...'what the hell is the remixed version?'

Well, like most of you, I was fed up with the performance of the TSX audio performance. You can add new speakers and ipod interfaces (which i did), but unless the amp is changed out for the balanced JL 300/4, the stock amp is the bottleneck in sound quality.

I went with the cheaper and maybe more laborious solution. After 20 or so hours of crafty handywork, the stock amp was opened up and the top pcb was dissected and frankenstiened until the circuit changed to the following schematic:

This changes the crossover points, remains a flat frequency response per band, and added attenuators to the tweeter and rear (sw only) stages.
http://picasaweb.google.com/castles....16266190242914
The original response was:
http://picasaweb.google.com/castles....16270485210226

Now, the response looks like:
http://picasaweb.google.com/castles....16270485210242

Also, LC filters were added between the amp IC and the output connector removing lots of unnecessary high frequency noise.

SQ is much improved. BTW, its amazing how much bass you can get when the amp isn't just driving a 50Hz peak, but full lower frequency extension.

Thanks for all your help on this forum. I couldn't have done this without all your help!

mercman

The response looks really good. I would like to here it! You are truly a die-hard rocket scientist! I see you chose to roll off the rear door speakers as well, how is the “sound stage”? Did you build a new daughter card or just change the components on the active filters?

Great work!

jeff

Castles_Saloon

Thanks for the props! The rear door speakers were rolled off too. I installed Dayton RS150-4 6"
http://www.partsexpress.com/pe/pshow...number=295-372
Above ~2kHz, it gets a bit ugly so i figured i should kill upper frequencies on the mid drivers (acoustics aren't as good sitting in the back, but i'm fine . All mids were rolled off below 100Hz mainly to crossover well with the rear deck, but also to avoid the playing close to the resonant frequency and creating distortion. These mids are certainly more clear and detailed than the stock drivers.

The sound staging isn't perfect because i do fade only to F1 so the rear door speakers pull the stage more towards the cabin's center instead of far in front of me. Although, the rear fill do help the mid/upper bass. I haven't tried disconnecting the rear doors to have a farther in front soundstage. Might me fun!

I contemplated building a daughter board, but there's circuitry on there that i didn't want to mess with too much. Too lazy to identify all of it. From recall, i think theres 8 dual 'new japan radio' op amps, one thats dedicated to creating pseudo ground references of 4V, one for the left channel and one for the right. The supplies are 0V and 8V. I lifted the rest of the op amp I/O pins and hand wired the circuit. These op amps had pretty good specs so i didn't change them. I had to disconnect all installed components that led to/from the connectors pins (between PCBs) and FR, FL, RR, RL, TWR, TWL, SWL, and SWR.

Attenuation stages on the tweeter and rear deck came in really handy for tuning since i wasn't sure how acoustically efficient the speaker locations were.

It was a long process, but fun and worth it!

Reach

Wow, brilliant and I'm gust going to say it: one of a kind! I don't imagine anyone else going through this. Thanks for bringing a bit of science class to the forums

Knifey

Any objection to posting a pic of the amp internals post-op? I'd like to see it, the more frankenstein the better.

Castles_Saloon

I'll update with post-op pics in a few days. my tools were stolen so i need to buy another 10mm socket before a quick dissection.

feuss2

iTrader: (1)

Can't wait to see the post-op pics! Can I send you my amp for a makeover?

Castles_Saloon

We'll see if you still want an amp makeover after the post op pics

jlukja

I'm curious. What are you running in the front and tweet locations?

Castles_Saloon

Front and Rear door speakers: Dayton RS150-4 6"
http://www.partsexpress.com/pe/pshow...number=295-372

Tweets: Seas 27TFFNC/G (H1396) 1" textile dome
http://www.madisound.com/catalog/PDF/seas/h1396.pdf

PICTURES
Amplifier
Original stock amp:
http://picasaweb.google.com/castles....85459815985506
http://picasaweb.google.com/castles....85459815985538
Original stock frequency response:
http://picasaweb.google.com/castles....16270485210226
After changing caps to LC filter (tweeter output, bottom pcb):
http://picasaweb.google.com/castles....39084895291762
After crossover modification (top pcb):
http://picasaweb.google.com/castles....39033355684178
New crossover schematic
http://picasaweb.google.com/castles....16266190242914
Stock's modified frequency response: i wish this was the cabin's resonse!
http://picasaweb.google.com/castles....16270485210242

Speakers
6" dayton pic:
http://picasaweb.google.com/castles....39677600778626
Dayton to stock driver magnet comparison:
http://picasaweb.google.com/castles....39759205157266

Sound Insulation
Doors: this helped reduce road noise from the side
http://picasaweb.google.com/castles....39892349143458
http://picasaweb.google.com/castles....39922413914546
Cut into sound insulation to allow window movement:
http://picasaweb.google.com/castles....39995428358594
Floor: a lot of effort for very little improvement
http://picasaweb.google.com/castles....40192996854242
http://picasaweb.google.com/castles....40394860317186
http://picasaweb.google.com/castles....40515119401490
Tricky Pedal fastener (image shows how to remove easily):
http://picasaweb.google.com/castles....40042672998866
Better than new: (,minus a broken clip or two)
http://picasaweb.google.com/castles....40626788551202

capnxtreme

Awesome work.

Could you possibly provide a list of the additional parts that you used?

If we wanted to reproduce your work from the schematic, how would we find the corresponding pins on the PCB?

Are the stock speakers 4ohm???

Any additional info would be awesome, thanks!

CCColtsicehockey

would you possibly consider doing this for a fee.

Castles_Saloon

Thank you!

As far as parts go, all values you can see on the linked schematic. Resistors are metal film (0.125W), Caps were teflon & polyester mostly (avoid ceramics since they have microphonic based noise) and should probably be rated for a min. of 20V. The adjustable resistors were pretty generic (they can handle many turn cycles so not the cheap stuff).

The LC filter (replacing output tweeter caps... L=15uH (type? used in switch mode power supplies) C=470nF(teflon).

Mapping pcb pins to the schematic is somewhat arbitrary. Two critical things...
1-all discrete components (resistors & capacitors) directly connected to the output connector (labeled solder pads - swl, swr, twl,fr, fl, etc.) & the input pins (FR, FL, RR,RL) should be removed. Don't forget to check both sides of the board.
2-There is one IC (i'll later let you know which one) that buffers the 4VL and 4VR signals. Leave this one alone.

Otherwise, i lifted the opamp +,- & output pins (I think 1,2,3,5,6,7). These can be verified if you look up the New Japan Radio datasheet. Each input to rc network to opamp to network to output pins was handwired. Be careful to ensure components are fairly low because their is some height restriction.


All speakers are ~4ohm.

Hope that helps! Feel free to ask more questions.

Castles_Saloon

The install labor would be too consuming to hand wire, but I'd consider making a generic pcb. Note the PCB would be bare and you would need to purchase R's, C's, opamps seperately and solder them. Most would be larger components and not too difficult to solder.

If there was a very large interest, then maybe i could generate PCAs (printed circuit assembly - pcb with components).

The connectors could be salvaged from the old board.

If there's interest, i'll give it more thought.

Egotist

Wow!