"The Wire" Ultra-High Performance Headphone Amplifier - PCB's

Hi Mr.Owen,
I made this coz' i missed the GB. the pcb singleside trace for the wire SE-SE using PTH component, Any comment?
Tommy:cheers:
 

Attachments

  • 1pcb.jpg
    1pcb.jpg
    91.2 KB · Views: 733
Hi Mr.Owen,
I made this coz' i missed the GB. the pcb singleside trace for the wire SE-SE using PTH component, Any comment?
Tommy:cheers:

i'm wondering what the buffer heatsinks are there for if you wont be using them? also there is nowhere near enough ground pour, the ground and power traces i'm not sure they are thick enough either. without more info its hard to make further comment
 
Last edited:
Excuses, excuses!

>>Still waiting for it to arrive in the mail . . .

>>As soon as I receive that magical PCB in the mail. . .

Can you believe these lame excuses, qusp? I had mine built before opc even shipped it and had pix to satisfy your demands :)

I must say the BAL-BAL is STUNNINGLY good. With HD-600s it is noticeably more detailed than the double sided pcb with LME49710s at the front end. I am not allowing myself to imagine how it might be improved with better cans.

You have given us all a great gift, Owen. Thank you. And thank you, qusp, for noticing that the qmini might not be adequate. It was not. I have had time to switch it in and out, and with both preamps, and it bogs down with four buffers.
 
Still waiting for it to arrive in the mail.

I've been enjoying the first version though. :D

ditto and ditto, i think this thing is going to really turn some heads, any improvement on the BAL-SE and we have a real knockout, the opa1632 sounds pretty damn good on its own. btw everyone, as opc touched on, the opa1632 likes to run disconcertingly hot, dont worry its normal. i was worried about oscillation the first time i used it, but its like that in every build. it'll run like that al day long, but i do wish there was an option for the DGN version, as it runs a lot cooler. i'll be running little heatsinks on them too just for fun

As soon as I receive that magical PCB in the mail mine will be built faster than you can reply "cool"... and I'm sure OPC knows we're all looking forward to it... :)

cool :cool:

>>Still waiting for it to arrive in the mail . . .

>>As soon as I receive that magical PCB in the mail. . .

Can you believe these lame excuses, qusp? I had mine built before opc even shipped it and had pix to satisfy your demands :)

I must say the BAL-BAL is STUNNINGLY good. With HD-600s it is noticeably more detailed than the double sided pcb with LME49710s at the front end. I am not allowing myself to imagine how it might be improved with better cans.

You have given us all a great gift, Owen. Thank you. And thank you, qusp, for noticing that the qmini might not be adequate. It was not. I have had time to switch it in and out, and with both preamps, and it bogs down with four buffers.


excellent!! i told you man, i take it the query about properly laid out balanced operation you had has been answered? great combo with this amp the hd600 (the BAL-SE version and what i know of the combination of components), the only dynamic i prefer is my jh13 until you spend uber money. mind you i havent yet hear the LCD-3, the LCD-2 wasnt really my cup of tea though, found them slightly lacking in life compared to HD600 properly driven. very accurate, smooth and articulate in the midrange, but lacking a bit of heft/slam for my taste.

The senns have a bit more of a raw edge on the bass notes when called for that i couldnt get the LCD-2 to accomplish. apparently the 3's are an improvement in that regard, but $$$$ is a bit rude IMO, may as well get some stats at that money, though you dont beat HD600 weight with them either until you get pretty serious and give them a serious amp.


no problem on the PSU spotting, glad to be of help, for a headphone amp i reckon this one will eat enough current for a nice jelly to go with the Xmas bird:deerman:

getting a bit excited, hopefully it gets here next week
 
To bridge izHuman, to balance izDivine

>>. . . the query about properly laid out balanced operation you had has been answered?

Yes, absolutely. The clarity is amazing. I tried for years to explain to my elderly father why he should listen to something better than Bose. His response was always that his hearing was not that good, meaning that he had high frequency loss. That rendering even lower 'impure' frequency mixes requires components above his 'hearing' was something he could not grasp. The dirac delta is what I will be listening for at the instant I die :)

I am nearly ready to turn my attention to the circuitry immediately following the DAC IC output. In desperation I chose an RDC-7 to get that part of the chain operational. It is a real value on eBay imo, but it is offensively large and power hungry. The WM874x in dual mono provides an opportunity to experiment and my expectations are high for a pair of BAL-SE wires to bridged across my earz. Anybuddy tried that yet?
 
20V "hybrid" frankenwire

It would actually be possible to build a 20V "hybrid" Wire, something I'm thinking about doing with one of my SE-SE boards, using the LME49610 and LME49860 chips. . . The net result would definitely have reduced measured performance on opc's AP due to the . . . off-board parts, the extra wiring, the wiring lengths between the two PCBs, etc.

@agdr It might be possible to take advantage of a rework SE-SE to solder the 860 such that lead 8 gets soldered to pad 4 (i.e. the ic is rotated 180 degrees) where the polarity is reversed and using tiny flying wires to rearrange the signals to that 990 and the second 990 pads. This would eliminate much of the contribution of "off-board parts, the extra wiring, the wiring lengths between the two PCBs" and utilize at least one power trace.

It would be 'under the board' so who'd notice :)
 
@agdr It might be possible to take advantage of a rework SE-SE to solder the 860 such that lead 8 gets soldered to pad 4 (i.e. the ic is rotated 180 degrees) where the polarity is reversed and using tiny flying wires to rearrange the signals to that 990 and the second 990 pads. This would eliminate much of the contribution of "off-board parts, the extra wiring, the wiring lengths between the two PCBs" and utilize at least one power trace.

It would be 'under the board' so who'd notice :)

Interesting! Good find. I'll have to give that a try. :) I bought a few of the rework boards and just got some more 49860 chips in. I agree, your idea would certainly cut down on all the extra wire between boards. I'm not very good at 3-D visualization from the board layout diagrams. I do much better with an actual board in hand. :)

If it does come down to needing a second board for the LME49860, another idea would be to stack the boards vertically using spacers like this to keep wiring compact:

Nylon Standoff M3 Thread 14.5mm - dipmicro electronics

Nylon Standoff M3 Thread 1.02� (26mm) - dipmicro electronics

Maybe put the LME49860 board on the bottom of the stack, the Wire amp on the top, and then run power and signal wires around the edges.
 
Last edited:
pretty sure you could sandwich the 2 boards together using vias to transfer power with home made 20awg wire 'pinheaders' and kill 2 birds with one stone. me i'd probably be mounting the 860 on a small bit of perf so you could have the decoupling directly on the top of the pins
 
Wolfsin has the blues

On closer examination it appears that severing the two thinnest traces, adding one blue wire under the 860 and a tiny jumper to the negative rail, all but one of the legs could be soldered to their pads and the bypass caps would be right where they are needed. Three half-inch blue wires to the otherwise unaltered pads of the second 990 and then one short blue wire from the lifted output to the via on the tiny severed trace and the dastardly deed would be done.

Trust none of the above, but it might be close. My blue wire daze were confined to digital repairs long ago. What would be required for those half inch signal lines in the sound path?
 
qusp - that is a good idea! Make it a daughter board right above the op amp area with wires sticking up from the main board. No need for a full sized second board.

BTW - I first learned about that higher voltage LME49610 from you in a "Wire" post a few months back. I didn't know that higher voltage variant existed.

+/-20V rails should produce at least a 17Vpeak swing, or 12Vrms. I've been sort of using the AKG K-1000 as the "gold standard" for low efficiency at 74dB/mW. A 12Vrms swing would even power those to full volume. I wonder what is left out there in headphone land that a 12Vrms swing wouldn't be enough for? Surely there is something. :D

Wolfsin - sounds promising on the direct-chip substitution! I'm going to give it a try when I get my boards. There is a SOIC version of the LME49860 too, but probably wouldn't be any help.
 
yeah thats it, just grab a bit of perf and cover the whole bottom layer with copper foil for a ground plane, then stick pins through the holes and solder to the copper for vias for the decoupling, mount it mid-way between the 2 opamp positions just above the board, or you might be able to get closer if you mount it below. dont go lifting the BW pin on the 610 like some numbskulls, thinking WB mode is going to provide an improvement at audio frequencies, because its even wider band than the 600. i wouldnt figure you for someone to do that, but just a warning..dont

the 610 doesnt seem to have been very popular, or maybe they just havent specified it in enough datasheets, or advertised it enough, because its marked as 'end of line' at digikey. the part isnt being discontinued, but its probably not a good sign.

not bad cans the K1000 either, i like them, especially when augmented with a sub :devily: