The Objective2 (O2) Headphone Amp DIY Project

yeah thats why i mentioned it =) i have other balanced headamps too, but even the bal-se wire is a VERY strong contender and not at all out of place in such a high end rig.

do remember running balanced you will have 3v point to point output from the sabre
 
@ethanolson, the short answer to that question is most DIYers design "blind" without proper test equipment and using only heavily biased sighted listening as their guide. They very often think they hear things that don't really exist as has been demonstrated countless times in blind listening tests. if you throw a bed sheet over the equipment, without changing anything else, the benefits they think they hear magically disappear. So it's easy to see how they think adding an op amp stage to a DAC chip improves the sound when in reality it might make the performance a bit worse.

haha please realise that this is an area where RS really doesnt know what hes talking about. to get the 120db voltage out spec quoted in the datasheet, there is also a specific voltage out schematic. just connect the output of the sabre directly to an amp input and you'll be lucky to get 100db vs 135db with a proper low impedance IV stage. now of course RS will say you cant heard the difference, but to me thats a pretty extreme difference and you may as well use an ipod if you are happy with that.

it doesnt have a 'current mode' or a 'voltage mode' the higher the impedance seen by the output the more like a voltage source it looks, the lower the impedance seen at the output (preferrably biased to AVCC/2) the more like a current source it looks

RS i have no problem with your engineering approach, this is why i like ackos products as in his day job among other things his company privately contracts to supply RF and control systems gear to the australian defence force. you'll excuse me if i dont mind trusting his skills and measurements.

its probably better to do a little research about the gear you make dismissive comments about. opc's mosfet IV i use (among a few others) as well in the development phase, did some of the most exhaustive measurements with the AP that ive seen on this dac chip and passive resistive IV (which is better even than connecting direct) was pretty ugly by comparison. the THD +N and DNR is fairly profoundly affected by the impedance and/or transconductance of the next stage
 
Last edited:
for the D1 IV, so then after those measurements over 30 different mosfets were tested to find the highest gM part it was discovered that up to a point the supply voltage used improved the gM, so thus we have an IV stage that runs at +/-50v and dissipates roughly 50W.

i tested >150 of the chosen mosfet at the circuits operating point for VDS and gM at 3vrms 1khz sine wave input (to match the sabres output) and sets of fets for the GB were selected. I built a differential input instrumentation amp to amplify the measurements taken across gate to source at as close to the same temperature as possible and used 3 x DMM and an old HP voltmeter. each DUT was placed in a replica of one phase of the D1 circuit and operating points were matched to what would be used

still sound like fumbling in the dark?
 
Last edited:
"Amped" is far too general a term though. An amplifier is, after all, not a magic black box that'll make cans sound better no matter what. That's not how things work in the real world. And whenever "better" comes up, the inevitable question arises - compared to what? Headphone outputs vary widely, too.

Ultimately it comes down to analyzing a headphone's needs (required voltage and current levels, tolerable noise and output impedance) and matching that to whatever the source has to offer. If this works out well across a wide range of headphones, we have a good amp there.
 
@sgrossklass, I believe the higher current availibility is to account for a fuller sound, though it may also be some distortion that I like. I haven't measured it to really know.

You are right about raising the question of "better than what?" Better than the same headphones from my iPod without the amp. My laptop has lots of juice so compared to straight out of the port, there isn't much change... other than the hiss. My DAC stage really helped improve the sound there.

I've had the distinct pleasure of listening to the NuForce uDac2-HP before I knew about RS's blog and I made the determination within 2 minutes that it was worse than straight from my computer... other than the hiss. So there are differences that are audible to me... but I don't know if it's because the headphones can "breath" better or if there's some distortion at play.

My current amp can supply (much more than) the needed 125mW into my headphone's 55-120 Ohm impedance range and it was with this amp that I actually had the realization that something there is helping me hear more and I really like it.
 
Oh, I forgot to mention that my interest in the O2 stems from the fact that it's well understood because it is well measured and it has a great damping factor. These are things I just can't say one way or the other about my current setup.

But I have even greater interest in the ODA because I no longer listen to music from any portable devices and don't think I will again for many years.
 
haha please realise that this is an area where RS really doesnt know what hes talking about. to get the 120db voltage out spec quoted in the datasheet, there is also a specific voltage out schematic.
I was (perhaps mistakenly) answering the question in general terms not for a specific DAC. I certainly wasn't trying to criticize any specific designer. Ethanolson's question was:

"Why does everyone have I/V stages for their ESS DACs?"

Because ESS doesn't make their datasheets public I've only seen data on a few ESS DAC chips. But the ones I've seen have an output stage built into them and their DNR spec is at the output pins of the chip--not after an external IV stage as you suggest. The application circuit in the datasheet does not show an op amp stage after the DAC.

Adding an op amp onto a DAC that already has an internal output buffer stage isn't likely to improve much besides perhaps the ability to drive unusually low impedance loads or provide a higher voltage output if desired. Adding gain after such a DAC isn't likely to improve the DNR as you suggest.

Certainly if a DAC chip is designed to be followed by an IV stage, or has a high impedance output, you should use an op amp stage. And it may well be some of the ESS chips are in that category. I was assuming Ethanolson was referring to the ones with built-in output buffers.
 
Thanks for the responses everyone. These ESS DACs still have some mysteries that I'd love to see unravaled in the future.


Moving back to the O2...
Does anybody know of a source for the battery technology used by Sanyo for the Eneloop 1500 cells that can be adapted to the O2 and still fit in the enclosure?
 
I was (perhaps mistakenly) answering the question in general terms not for a specific DAC. I certainly wasn't trying to criticize any specific designer. Ethanolson's question was:

"Why does everyone have I/V stages for their ESS DACs?"

Because ESS doesn't make their datasheets public I've only seen data on a few ESS DAC chips. But the ones I've seen have an output stage built into them and their DNR spec is at the output pins of the chip--not after an external IV stage as you suggest. The application circuit in the datasheet does not show an op amp stage after the DAC.

Adding an op amp onto a DAC that already has an internal output buffer stage isn't likely to improve much besides perhaps the ability to drive unusually low impedance loads or provide a higher voltage output if desired. Adding gain after such a DAC isn't likely to improve the DNR as you suggest.

Certainly if a DAC chip is designed to be followed by an IV stage, or has a high impedance output, you should use an op amp stage. And it may well be some of the ESS chips are in that category. I was assuming Ethanolson was referring to the ones with built-in output buffers.

ok, lets just leave that then, maybe i was too sensitive there, but as it was a question directed at me specifically and you intercepted it and made that comment it didnt sit right. but lets move on

actually you can quite easily get the DS, just email the distributor its not under NDA anymore, but its still not posted publicly. i got it when it was still under NDA, not sure if i'm still bound or not, but i wouldnt think so, i certainly hope so ha.

no its a bit of a weird one, the same output can be used as current out with higher performance, but puts out varying degrees of voltage depending on the Z it sees at the output. opc/we manage to get ~115db THD +N from literally a couple of mosfets and resistors plus a cap in the signal path, of course there are some other components for setting bias and decoupling, but no magic. i wouldnt say an opamp is necessarily needed, but not being one to avoid them like the plague i also appreciate a well done opamp IV on the sabre too.

its not really that mysterious and the DS doesnt hold that many secrets, certainly nothing you cant see around the forum. some might call it a flaw that the output is so easily wavered. it certainly is a bit touchy with that, the quality of the i2s and quality of the somewhat complex power supply arrangement, but if you can get it right its superb.

anyway, apparently my O2 pcbs are actually on their way as of today!! haha
 
Last edited:
"Amped" is far too general a term though. An amplifier is, after all, not a magic black box that'll make cans sound better no matter what. That's not how things work in the real world. And whenever "better" comes up, the inevitable question arises - compared to what? Headphone outputs vary widely, too.
Too general? Not really. I was speaking of the general population, for example the proliferation of iPod/iPhone users I've been seeing with their "Dre b's" or whatever they call them, or the earbud/mp3 player crowd.
Ultimately it comes down to analyzing a headphone's needs (required voltage and current levels, tolerable noise and output impedance) and matching that to whatever the source has to offer. If this works out well across a wide range of headphones, we have a good amp there.
For audio geeks and diy crowd, sure. General population don't much care. That's the real world.
 
Hello, I'm about to build my O2 but I have several questions :
- Is there any schematics for the front panel ? The FPE software keeps crashing so I would like to just have some measures...
- RS recommends using "pigtail" style adapter for connecting 1/4" jack. But I can't seem to find these anywhere. Do you know where I can get one ?

Thank you :)
 
Hello, I'm about to build my O2 but I have several questions :
- Is there any schematics for the front panel ? The FPE software keeps crashing so I would like to just have some measures...
- RS recommends using "pigtail" style adapter for connecting 1/4" jack. But I can't seem to find these anywhere. Do you know where I can get one ?

Thank you :)

Hi Shamharoth,

You can see a photo of the front panel I had made for my O2 by looking at the pic in message #784 in this thread.

There is a link to download the FPD file in message #794. You could send this file directly to Front Panel Express without you having to open it (assuming you don't want to make any changes).

BTW, I was very impressed with the accuracy of machining of the panels by FPE... the fit and finish was perfect. I look forward to seeing their facilities during a Seattle area Head-Fi meeting in January.
 
Last edited:
Hi Shamharoth,

You can see a photo of the front panel I had made for my O2 by looking at the pic in message #784 in this thread.

There is a link to download the FPD file in message #794. You could send this file directly to Front Panel Express without you having to open it (assuming you don't want to make any changes).

BTW, I was very impressed with the accuracy of machining of the panels by FPE... the fit and finish was perfect. I look forward to seeing their facilities during a Seattle area Head-Fi meeting in January.
I can't open any FPD file, the software keeps crashing. I tried to uninstall and reinstall it, but to no avail.
I would like to build the front panel myself as I have the required tools (and would like to save some money too :D), but I need to have some dimensions to work with. Or maybe someone can export the FPD file at the correct scale to PDF (A4 paper) ?

BTW, your panel looks great!
 
Hello, I'm about to build my O2 but I have several questions :
- Is there any schematics for the front panel ? The FPE software keeps crashing so I would like to just have some measures...
- RS recommends using "pigtail" style adapter for connecting 1/4" jack. But I can't seem to find these anywhere. Do you know where I can get one ?
You might try a different PC for the FPE software as I've never had it crash or heard of it crashing. Unfortunately the software doesn't have dimensioning otherwise I could just produce a PDF from FPE.

The coordinates for the openings are in the O2 PDF on the board drawing as measured from the center of the board and from the top surface of the board.

I also know JDS Labs is working on selling a pre-machined panel, and you probably saw the link to the thread on the laser cut acrylic panels.

The 3.5mm to 1/4 inch "pigtail" adapter can by anything similar to the HOSA YMP233 (which happens to have dual 1/4" jacks):

http://www.amazon.com/HOSA-STEREO-3-5mm-PHONE-STER/dp/B000068O5I/
 
You might try a different PC for the FPE software as I've never had it crash or heard of it crashing. Unfortunately the software doesn't have dimensioning otherwise I could just produce a PDF from FPE.

The coordinates for the openings are in the O2 PDF on the board drawing as measured from the center of the board and from the top surface of the board.

I also know JDS Labs is working on selling a pre-machined panel, and you probably saw the link to the thread on the laser cut acrylic panels.

The 3.5mm to 1/4 inch "pigtail" adapter can by anything similar to the HOSA YMP233 (which happens to have dual 1/4" jacks):

http://www.amazon.com/HOSA-STEREO-3-5mm-PHONE-STER/dp/B000068O5I/

Thank you. I'll try using using the coordinates on the board schematics.
edit: the software worked on another computer, so everything's fine. :)

For the adapter I didn't even think to look on amazon, silly me. :D
 
Last edited: