[adydula]

Yes he has disappeared for some time and I have seen where his domain is getting renewed by someone...AGDR had many conversations with him on and offline about the O2 and its design and parts etc...for the most part Nwavguy did post this list of stuff that would most likely be in an ODA:

For the most part the AGDR version of his amp has all these incorporated in it. The only real issue is AGDR using SMD parts and yes for some this a hard stop for DIY, especially if you have no experience soldering!



I listen to my ODA and have many amps here to compare to and its just as good as any of them of not better considering the versality of its design. This is one amp I will never part with, its that special to me.

The super Cmoys of AGDR is a neat experiment when TI came out with their headphone exclusive design for an IC, the OP1688, AGDR jumped on it and saw the chance to make the Cmoy amp even that much better, better specs, more power etc. This on AGDR will sell at cost the smaller parts mounted on a small daughter card to help make it easier to build.

I dont know what he is going to do in the future but I hope to be around to build one of whatever it he decides to do next!

Alex

[agdr]

Please do post your O2  Booster Board listening impressions guys!  Each headphone and IEM model has its own sound signature.  I'm always curious how various models pair up with the Booster Board vs. the other headmps you have tried - and vs. just the stock O2 itself.   

Remember that just by installing the Booster Board (which has an inherent slew of 28V/uS) your slew rate will double, via the NJM2068 gain chip at 6V/uS now being the slowest thing in the chain rather than the 3V/uS NJM4556A stock output buffers.  NwAvGuy made a pretty soild case in his blog writings that slews faster than the 3V/uS don't matter because the digital processing chain in the recording studio won't be more than 2 or so.  But I know a lot of audiophiles - and their ears! - debate that.  See what you think.

Something pretty neat, the image below is dB Cooper's Booster Board being tested for DC output offset.  Those meters are on the mV scale, so that is 11 microvolts and 22 microvolts (= 0.011mV and 0.022mV, of course).      Once the chips warm up that will often drift by 50uV or so, but that beats the pants off the O2's stock 3.3mV = 3300uV offset.  The net result is to allow your heapdhone and IEM drivers to stay much closer to their natural mechanical idle position at idle and at zero signal crossings in the music.  Rocket - sorry I don't have a picture of yours or I would post, it had been tested about a month ago when it was built.  But the numbers were similarly low.

Adydula - hey thanks for all your efforts in testing and A/B'ing this stuff! You built those Super CMOYs before there was an updated build instructions or updated BOM, just off the schematic.  Pretty darn good DIY.

[waynelee72]

I want to buy your Super CMoy.

[agdr]

I want to buy your Super CMoy.

Hi, PM sent. 

[adydula]

I must admit that I have built and O2, AGDR ODA, Three AGDR Super CMoys but never an O2 Booster board. I elected to build the more complex and much more versatile ODA, AGDR's desktop O2.

But I have decided to build a O2 Booster board!! So I will be able to comment on the differences between it and the original O2.

Stay Tuned!
Alex

[adydula]

Ok,

I just ordered my parts for the O2 Booster Board from Mouser.com.....

Should be here by the end of the week and it will make a nice weekend project.

I have two 02 amps, and have to decide which one to open up.

Here is a short list of the O2 Booster Board benefits:


Nearly twice the current output capability on peaks, 250mA vs. 140mA per channel on a stock O2 Headphone Amplifier. Why this matters
Nearly zero (50 microvolt) DC output offset voltage, vs. typically 3000uV (60x more) for the stock O2.
Zero turn-on or turn-off thumps due to an included headphone relay circuit
Doubles the slew rate to 6V/uS, and 20V/uS if the NJM2068 gain chip is replaced by 2x LME49990 adapter
Power rail clamp diodes.  This is an important O2 ommision - some versions of the 7912 voltage regulator have been known to fail to start up into some load conditions (in general, not just the O2) without a rail clamp diode.
Lower input-resistance induced distortion.  Many op-amps have a little-known problem of distortion being caused by higher levels of series resistance on their inputs.  The NJM4556A is believed to have this problem.  The OPA827 and OPA140 FET-input op-amps used in the O2  Booster Board have been desgined by TI to specifically reduce this problem.
Adds 2 LEDs that show when the power management mosfets in the O2 are on.  Helps if any O2 troubleshooting is ever required, to see immediately if both power supply rails are OK.
Allows for true zero ohm output vs. 0.5 ohm for the stock O2 Headphone Amplifier.  Lower amplifier output resistance means a better damping factor.
THD and noise figures of the chips, per the datasheets, slightly beat (better) the O2's stock NJM4556A




Alex

[dB Cooper]

I just had my O2 upgraded with the booster board. I was going to DIY it, but I wanted to make some changes to the gain settings on the mainboard and I quickly discovered that these old eyes were not up to the task of working on a board with a layout as tight as that of the 02. (I envy adydula being able to take on a surface mount board.) So I contacted AGDR and asked him about sending both the O2 and the booster board back to him to complete. He quoted me a price which was completely unfair (to him!) and completed the work in next to no time and before I knew it, my modified O2 was back in its spot.

Mini review: I do notice differences, which largely seem to be in inverse proportion to the impedance of the headphones. First ones I plugged in were my 16 ohm Etymotic hf's. These have always seemed just a wee bit on the 'soft' end of the definition scale, but playing some personal reference recordings, I noticed right away that I was hearing further into the mix than before, and a 'crispness' with transients that had been slightly lacking. What I noticed with the Ety's was also largely true for my other in ear monitors. (Innerfidelity has an interesting article on how balanced armature drivers work; apparently they are prone to resonances if not implemented and driven properly. It's worth a look.) So it seems that for low impedance headphones, the difference between 0.54 ohms and ~0.04 ohms may actually make a useful difference in performance. (The only possible small downside is that the Ety's may have 'leaned out' a bit in the mid-to-upper bass, which they don't need; I'm still making up my mind about that.) The bass is fine with my Apple In-Ear BA IEMs (don't laugh; they are great performers, especially for the money and if you spring a few bucks for the Comply tips) and SE535s.

My 300 ohm Sennheisers sounded clearer but the difference was much more subtle. I'm sure that to a 300 ohm headphone like the HD650, the difference between 0.5 ohms and ~0 ohms output impedance is probably negligible. But it does seem to play with a little more 'authority' and clarity with the booster board. (I'm now using a little EQ with the HD 650 but that is a discussion for another time). The O2 is notable for it's excellent low noise floor and, although I don't have any way to measure it, I hear no noise attributable to it at any volume setting that I would actually use, so the mod does not exact an audible penalty in noise performance. The other benefits adydula quoted from agdr's web page are worthwhile too (no more turn-on/turn-off noises etc).

I just wanted to say that not only does the product seem to perform very well, but Martin at AGDR was very helpful, answered all of my numerous dopey email questions, corrected a number of assembly errors on the original O2 board (my amp was purchased assembled), suggested some modifications to the gain parameters which resulted in an amplifier which is nearly perfectly 'tuned' to both my low impedance IEMs and my high impedance full-sized cans (thanks to the gain switch) - no small feat, and a tribute to both his work and the original design work of NWAVguy, the Jack Kerouac of audio.

Very satisfied customer.

[agdr]

Here is a short list of the O2 Booster Board benefits:

Hey Alex!  Here are a couple of updates to that list, for the latest V3.5 O2 Booster Board:

* I'm now listing the two green power-rail LEDs and their series resistors as "optional" by default on the BOM.  I put those LEDs on the board purely for diagnostic purposes.  I have helped so many folks debug their NwAvGuy O2 builds over the years and it always starts with "measure your power supply rail voltages for me".    With the LEDs you now for sure both rails are up.  But they do pull a small amount of current, which reduces battery life a tiny amount, and the big thing is new holes have to be cut in the O2's front panel to see them.  I have a CAD design for a new O2 front panel with the holes posted, but Front Panel Express charges $25 or so to make a panel.  So the best thing to do is just leave the LEDs off - or covered up, they won't interfere with a front panel that doesn't have the holes - unless you want to add the panel holes.

* I'm no longer recommending that LME49990 swap-out for the O2's NJM2068 chip.  National is discontinuing that chip this year for one thing.  I've been able to confirm that some chips in their latest batch will oscillate if used in the O2.  When I first posted that modification years ago I tried several LME49990s and didn't get oscillation with any of them.  Seems that something in TI's process has changed lately.  The issue here is that NwAvGuy's board was only designed for the low-bandwidth (27mHz) NJM2068 chip.  The LME49990 is 110mHz which can require more detail to power supply rail bypassing with capacitors, more than just the 0.22uF film caps that NwAvGuy has on each rail for each op-amp's power supply pins.  The LME49990s are single op-amps too, whereas the NJM2068 is a dual, so an adapter board for "two single SOIC-8 to a dual DIP-8" has to be used, which adds additional opportunities for oscillation, especially if the bypass capacitors are not right there on the adaptor board.

Even with the original NJM2068 (which is really a pretty good chip), adding the Booster Board will increase the slew from 3V/uS to 6V/uS.  The slowest thing in the chain was the NJM4556A output chips.  With those replaced by the 20V/uS Booster Board the new slowest thing in the chain in that NJM2068 gain chip.

* I've updated the power rail reverse protection diodes to SB140s from 1N5818s. Just  better diode all around for the same money, IF you get the Vishay version.  The Fairchild version of the same SB140 is nearly identical in specs to the 1N5818.

* I've updated the relay reverse clamp diode to a SMD S1G, which is a smaller 1A diode.  The one I had in there before was a 3A diode, physically larger than it needed to be.

* Lots of cosmetic stuff with part number markings on the board to make things as legible as possible.

[agdr]

Mini review:

Thanks for the review!  That makes a lot of sense, that the low impedance cans would show up the most audible difference.  High impedance cans are often a lot more tollerant of amplifier output problems.

Hey one slick we did with dB Cooper's O2 is add the input attenuation modification.     Changed the 274 ohm input resistors to 5K, which then forms a 1/3  attenuator with the 10K O2 volume control.  So the input signal gets knocked down by 1/3, which helps with "hot" sources.  The RF filter capacitors have to be changed from the standard 220pF to 12pF to maintain the same 2.7mHz-or-so corner freuqency NwAvGuy had set for the input RF filter.  The 5K resistor will have more Johnson (thermal) noise than the original 274 ohm, but probably won't be audible with standard gain levels.

The 1/3 attentuation (2/3 pass-through) applies to the gain settings too!  If the gain is set up for 1x / 2.5x, then after the mod you get 0.67x and 1.7x.  If you want to maintain 2.5x in the top position the "high" gain has to be increased to 3.7x so the product with the 0.67 input attenuation is still 2.5x.

[adydula]

Hey Martin!

I go the parts you listed above, the BOM had these listed and I am skipping the LEDS.....Drilling holes is always a pain!! LOL.

dbCooper, congrats on your O2 Booster board, glad you like it!!.

I would like to say there are few people I have run into in this audio DIY hobby that I respect and consider a friend and AGDR is one of them!! :>)

He is a really nice person, has a great insight into the circuitry and is always trying to make things better!!!

I have never had an email that went unresponded to and his work is top notch, he has gone out of his way to help people get there stuff up and running....

With the advent of the missing nwavguy, its nice to have AGDR around and a few others over at DIY to help make the O2 a bit "better".

My parts are on the way, guess I have to get out my 3X walmart readers once again!!

Alex

[dB Cooper]

The biggest "flaw" of the O2, in my opinion, is the concessions made to the design goal of "portability". The amp isn't particularly portable anyway, really. About a third of the circuit board space is eaten up by the batteries- space that could have been used for an improved power supply, or opening up the board layout so old farts like me (or first-time kit builders) could do their own assembly work (I've built many Dynaco, Hafler, and Van Alstine kits in my younger days, but the O2 board is just too 'miniaturized' for my eyes), or both. Accommodating large 9V batteries and the extra parts count required for associated functions wastes a lot of board space, but some reclaim that space for the ODAC of course. I've read that the vast majority of O2's are run as 'desktops' and never have batteries installed anyway.

[adydula]

I dont know anyone that uses and O2 for a portable amp. I agree its just too big, too heavy etc..

That said, its a great performer and many of us waited and waited for the designer to come out with the desktop version.

AGDR's ODA is a great desktop type of and O2 amp, that has most of the desgin points of the desktop version nwAVguy has on his blog site. But unfortunately its not a very friendly DIY project at all...it challenges even folks with good soldering skills. But I like challenges and at an age of 67 years with my tired eyes its getting more difficult. I have to use my walmart 3X readers to help see wht the heck I am doing and an external light source to illuminate etc....seems as you age you loose visual acuity and need more light to see stuff especially small SMD parts.

I guess you could build and O2 on a bigger board and probably it would work and sound pretty good. Nwavguy spent lots of time on the layout and his idea of portability to meet his objective design specs as well with the ODAC. So laying it out on a larger board may or may not meet the specs? To some specs area  baseline and to others they are not as important.

I have played with many, many amps doing blind listening testing here and very seldom have I heard really bad results of amps with a decent design. I am very skeptical about any new amp or electronic thingy that claims its more open or airy or dreamy or whatever. I do like getting them and listening and making my own "objective" tests.

Ok I need to go print out the booster board layout and schematic so I can "weld" one up!

Alex

[dB Cooper]

I guess you could build and O2 on a bigger board and probably it would work and sound pretty good. Nwavguy spent lots of time on the layout and his idea of portability to meet his objective design specs as well with the ODAC. So laying it out on a larger board may or may not meet the specs? To some specs area  baseline and to others they are not as important.

Alex

He designed the board to fit a particular enclosure and "roped off" a large part of the board for two bulky, heavy 9V batteries (and the associated power management/charging circuitry). Once those constraints were set, the rest of the layout became a 'given'. Not saying he did a bad job, quite the opposite, just that portability as a goal created other limitations such as the power supply design. If anything, I would expect that spreading the circuit out, either by reclaiming the battery/odac space or simply by giving up the iffy "portability" and going to a larger board/enclosure should result in fewer problems with stray capacitances and whatnot. But what's done is done. I'm not dissatisfied at all, especially now with the booster board, but I wish there was an easy-to-assemble, good performing, cheap & cheerful headphone amp kit with O2-level performance that someone with my close vision could build. It would be simple for someone with the right skills to eliminate all that dead space and simply spread out the circuit differently on the same size board, but that would surely violate the license terms. And the market for kits isn't what it used to be.

[adydula]

OK, I got my parts and everything I need to build the Booster Board!

I started this morning around 9am and completed it around 3pm'ish, with a lunch break and a few eye "strain" breaks! LOL.       

Here is  picture of the board as I was about 20% through the installation:



Heres the completed O2 Booster Board installed in the O2, two op amps have to be removed and the board plugs into
the original O2 board and three wires need to be added to make it functional. There are no adjustments it just works!




I had some minor issues with it not working correctly on first power up, and it was due to my poor soldering of one of the
827 op amps, a quick reflow of three pins corrected that!!

I have not objectively listened to it and compared to an stock O2 but I have one on hand and hope to get to this tomorrow.

The first impression is it works well, plays my T90s with authority and is absolutely clean, and silent...

Stay Tuned!
Alex

[adydula]

One of the issues with the O2 has been its "power supply" and people wanting a more robust power supply. I have in the past a few years back run the O2 with a lab spec Lambda DC power supply and compared the sonic output with blind AB testing and I could NOT reliably tell the difference between the two....

Also when you run off of batteries, its pure DC...so all those nasty AC power supply worries well just vanish! Again I cant tell any difference between all of them.

Alex

[Armaegis]

but I wish there was an easy-to-assemble, good performing, cheap & cheerful headphone amp kit with O2-level performance that someone with my close vision could build.

The cmoy kit from JDSlabs is $42.
Old school biosciencegeek is floating around ebay
Australian Fred has his own site these days: http://fredsamplifiers.com.au/index.php?main_page=index&cPath=3
AMB mini3 kits go for around $100 http://www.glassjaraudio.com/product.sc?productId=25&categoryId=2
Millett Starving Student kits used to float around $200-300
Meier Audio has boards and a few parts for his original Porta Corda.

[adydula]

Thanks for the list!!

One thing is db Cooper also mentioned "eye sight" yes there are kits and DIY stuff available, as we old farts age our eyesight gets poorer and gets harder to solder and see small discrete components.

The O2 Booster was not extremely difficult but was challenging to get the parts down and soldered well. I have been doing this for years and for me and my eyes I needed a magnifing glass(walmart 3x readers) a bright light to be able to read the bands on the diodes and pin 1 IC locations....The smd parts are SMALL.

Also if your not careful and drop one you will be hard pressed to find it again!!! Ha!

That said I would do all this again!! the designer of this booster board had a long relationship with nwavguy and lots of back and forth about the O2. Instead of just commenting AGDR went ahead from an enginnering perspective and offerred "improvements" on the O2 base that to some mean a lot and to others is just hype.

Having all these here has allowed me to listen and decide for myself!

Its a great past time for me in my retirement!

Alex

[Armaegis]

Oh I forgot, Sjostrom has a bunch of projects and boards too: http://sjostromaudio.com/pages/index.php/hifi-projects
A few use through-hole components, but most of them start getting into smd.

[adydula]

Ok we have spent the morning listening to a normal O2 vs the O2 with a booster board.

I setup the two amps with my SPL meter with a 1Khz test tone to 85db. If you fiddle with the volume knob its easy to get confused and make an amp that doesnt sound that good a whole lot better....

The source was an HP Spectre laptop to a ODAC to the amps. The music player I used was JRiver running in Bitperfect mode setup. The OS on the laptop is Win10 with all the updates, drivers etc...

I used my reference Beyerdynamics T90's (250 ohm) and a set of Audio Technica MSR7's (35 ohm, 100 dB/mW sensitivity). The sensitivity for the T90's: 102 dB SPL at 500 Hz at 1 mW. (1 mW is 500 mV at 250 Ω, thus 1V is 4 mW or +6 dBm into 250 Ω, so 1V gives 108 dB SPL. From Ken Rockwell.com review).

I am a pretty objective analytical person and dont believe in a lot of the snake oil that out there all over the audio world. Many of the amps I have built in the past few years all sound pretty good with subtle differences on the audio side and vast differences on the versatility side....mostly mechanical connections etc.

That said, as I listened to both my objective side was telling me that there isnt a lot of differences between the two...but subjectively I was drawn to the Booster board O2 again and again. It seems to be "clearer, more open, beautiful bass response and all the details in the background were there in spades. I played many very well recorded songs that I am very familiar with, and even though the O2 was excellent the O2 with booster board just drew me to its sound....could very well be subjective bias in my brain, but I do know that the booster board does not hurt or impair the O2 in anyway.

I am wondering if the slew rates have anything to do with this?

The major thing I noticed is the volume knob setting on both varied alot with the two headphones. The 35 Ohm MSR7's did not need much volume pot rotation at all around 8-9 o'clock, the T90s 250 ohm cans needed a lot more rotation to get the 85db, around 10 o'clock position.

What was immediately noticed it the turn on thump is gone...the booster board is totally silent and has a "black background". With the volume all the way up there is no hiss, hum or noise of anykind.

The amp comes alive around 3 seconds and even with the music source hot the turn on is smooth and not abrupt.

So all in all the booster board is a great upgrade for the Original O2 and sounds well "transparent", "open", "natural, and .......all those other subjective adjectives!

Great job AGDR!!

Alex

[adydula]

My AGDR stack:
1st gen Cmoy, the two altoids tins have the Super Cmoys...one O2 is stock the other has a booster board in it,
the blue box is the inverting version of the O2 and the red bottom box is the ODA or what AGDR thinks the O2 desktop aka ODA would have been like.






The small red box is an ODAC.

Alex