A version of an O2 Desktop Amp (ODA) - Page 11 - diyAudio
Go Back   Home > Forums > Amplifiers > Headphone Systems

Headphone Systems Everything to do with Headphones

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 17th May 2013, 12:43 PM   #101
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
John-

Yeah his dissapoearing is really the hangnail. This ODA project is a direct result.

From the various posts here and on his blog a lot of people were waiting for the ODA design he mentioned many times but never produced. I was too! Then he dropped off the radar. I waited for 9 months or so after he dissappeared to work on this version, once it seamed clear the official ODA would never happen.

I do agree that his big point was that performance did not degrade. One thing he certainly made crystal clear in all his writings is he was not in it for profit himself. Nothing would make me happier than if RS would suddenly re-appear and be willing to go through the ODA design here and modify it to his liking and specs. We all finally would get our (official) ODA.

From my standpoint, until I hear differently from RS / NwAvGuy, I'm going on the read of the license that home, personal, DIY is fine for derivitive works but commercial sale is not.

I do see one change I'm going to make. I just re-read what license wording have on the board. I had "GNU" on the brain and put GNU public license wording on the PCB, forgetting that he had used Creative Commons. I'll change that to specifically state it is NwAvGuy's original O2 Creative Commons license that applies, which was my intent, to prohibit commercial sale without his OK.

Hey thank you for bringing this whole issue up! It did need to be addressed. Anyone who does measurements on this ODA version, if that ever happens, needs to be fully aware of NwAvGuy's license terms.

Last edited by agdr; 17th May 2013 at 01:08 PM.
  Reply With Quote
Old 17th May 2013, 01:09 PM   #102
diyAudio Member
 
Join Date: Mar 2012
Yeah, I have been waiting on the ODA since it was first mentioned on Nwavguy's
blog. I am a big fan of what he has done, but I think we have done our due diligence on waiting for him to return. Agdr has stated that he wants to see his version of the ODA to measure as well or better than the O2. I don't see a problem here.
  Reply With Quote
Old 17th May 2013, 08:02 PM   #103
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Default V1.2 updated Gerbers, layout, schematic

As per the license discussion, here is a revised layout with the license wording on the PC board changed to reflect NwAvGuy's Creative Commons license.

I've made a few more upgrades. I realized that 5 of the 4.7uF coupling capacitors per channel are adequate rather than 6, chopping $10 off the cost. The corner frequency of the O2's 2.2uF + 40.2k pair is around 1.8Hz. With the 6 4.7uF caps and 4.99k the ODA previously was around 1.12Hz. Now with 5 caps the corner frequency is around 1.3Hz, still better than the O2.

And... getting rid of those caps freed up some PC board space to make another attempt at adding back in a linearization loop around each channel's paralleled buffers. Occured to me a couple of days ago that although the LME49990 is apparently way to fast to be using in a loop with the slow NJM4556AL buffers, the NJM2068 (the gain chip in the O2) is a lot more in the ballpark and should work. The unity gain bandwidth of the NJM2068 is 27Mhz vs. 110Mhz for the LME49990, vs just 8Mhz for the buffers. Still faster than buffers, but at least a fair shot at keeping the NJM2068 stable.

Plus, the best part, looking at the gain/phase curve the NJM2068 starts out at 0 degrees, going down to -90 (90 degrees lag) at unity gain. Recall from past posting that the NJM4556A has a 30 degree lag which starts dropping rapidly after the audio band. Together the loop should have a fairly large phase margin. Enough that I wouldn't be worried about the various headphone capacitive loads that could be attached. TI still hasn't come up with the gain/phase plot for the LME49990. The OPA627 that I tried for the loop at the start of this thread and eventually oscillated at 70K started out with the fairly large 90 degree phase lag, which together with the NJM4556A's lag resulted in no phase margin.

RocketScientist / NwAvGuy wrote that he measured the NJM2068 at having lower distortion than the NJM4556A and that the distortion numbers of the NJM4556A are theerfore the limiting factor in the O2. So if the loop here works properly, there is at least the possibility of reduced output stage distortion vs. the O2. I've laid it out so that the NJM2068 loops on the output don't have to be used. The parts can be left unpopulated and 1206 0R jumper resistors soldered in to bridge input to output, in which case the design reverts to the previous V1.1 without the output stage feedback loop.

I'm going to pull a RocketScientist / NwAvGuy at this point and disappear myself. John has some very good points about the licensing questions with the ODA. I'm not going to sink any more time into it. But these Gerbers should be fully functional if anyone should want to give it a try, although for home use and not commercial sale.
Attached Files
File Type: pdf ODA circuit.pdf (681.4 KB, 77 views)
File Type: pdf ODA layout both layers.pdf (885.4 KB, 39 views)
File Type: zip Gerbars for ODA v1.2 1.6mm green HASL double layer 80x160mm.zip (224.7 KB, 32 views)

Last edited by agdr; 17th May 2013 at 08:07 PM.
  Reply With Quote
Old 17th May 2013, 08:21 PM   #104
diyAudio Member
 
Join Date: May 2005
Location: Lisbon Portugal
I am no expert in circuit board layout, but this layout could be improved quite a lot! You should search the web for proper circuit board layout techniques and apply them here.
  Reply With Quote
Old 25th May 2013, 04:48 AM   #105
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
I've had some email. I'm not leaving the forum and the audio hobby world, just this project until such time as the license issues John mentioned become clearer. I've run a couple of tech businesses over the years and I take contracts and intellectual property seriously, as I'm sure John does with JDS.

For what it is worth, after some discussions this week, I don't have any answers but I do have a bunch of interesting pointers for anybody else looking into the O2 licensing problem.

As for the Creative Commons licenses, they claim to be written to not override the normal exclusions of Fair Use and Public Domain / Prior Art. In the CC-BY-ND license that RocketScientist / NwAvGuy used for the O2 amp:

Creative Commons — Attribution-NoDerivs 3.0 Unported — CC BY-ND 3.0

scroll down to “with the understanding that” and the bullet points about “public domain” and “Fair Use”

Fair Use is the doctrine that says a copyright holder cannot dictate how a person uses the copyright “for their enjoyment”, along with commentary and criticism of the work. There apparently has been a pretty good body of case law built up over the years:

Fair use - Wikipedia, the free encyclopedia

As it applies here, my best understanding of it all is still that non-commercial DIY work would constitute Fair Use. Internet postings of non-commercial DIY modifications may be a first amendment / free speech issue, same as publishing a how-to book on a subject.

Public domain and/or prior art relates to works already in the public domain. In the case of the O2, it is full of what are likely public domain works-
  • Half wave power supply
  • Voltage doubler
  • Regulators using off the shelf chips
  • Non-inverting op-amp and unity gain buffer op amp stages
  • Use of NJM4556A in a headphone amp (the Grado RA1)
  • Input RF filter (out of one of Doug Self's books)
  • Parallel op amps (the Meier parallel amp)
  • Trickle NiMH battery charging
and some things that may be unique-
  • The power management circuit
  • Pot and coupling cap in the middle
Although public domain works can't be copyrighted, collections of them can:

Welcome to the Public Domain :: Copyright Overview by Rich Stim :: Stanford Copyright and Fair Use Center

which may apply to the O2. But for that type of copyright to hold up in court, it may need to be extremely specific and narrow in scope, such as a portable headphone amp with 2 batteries (maybe even as specific as 2 NiMH 9V batteries), a power management circuit, with a certain circuit arrangement, etc. Otherwise, if applied broadly, such a copyright would apply to any portable headphone amp with batteries and two amp stages.

Which then goes to the issue of derivative works. Derivative works are their own chunk of copyright law:

Derivative work - Wikipedia, the free encyclopedia

An issue that comes up is "when is a work different enough so as not to be a derivative, but rather a unique work?" In the case of the amp here a few points of difference from the O2 amp are:

* No batteries, no charging circuit, not portable
* No power management circuit
* Different chips in the gain stage, two singles vs. one dual
* Different regulator circuit with different chips and 2 voltages, both different
* Different case and PCB form factor
* Unique layout, not a copy of the O2 layout
* Clipping indicator circuit
* Relay protection circuit
* SMD along with through hole parts, both sides of the PCB rather than one
* SIPs instead of DIPs for the output stage

but similar things

* pot in the middle with coupling caps
* parallel buffers, although more of them, using the NJM4556A chips
* 3.5mm input and output jacks, although other too

I don't have an answer here. I wish I did! But it certainly is a good question. Is it a derivative work or a new work?

And finally there is the issue about NwAvGuy disappearing, which turns out to be more of a can of worms than I realized. The concept of an "Orphaned Work" would seem to apply:

Orphan works - Wikipedia, the free encyclopedia

Orphan works in the United States - Wikipedia, the free encyclopedia

Apparently the law here is a mess after the Copyright Act of 1976 did away with the requirement to actually register a copyright with the PTO. So now there is no way for anyone to know if the original copyright holder or their representative is still around, much less how to contact them. Apparently various legislative solutions have been proposed since, but nothing passed or solved.

So yes, from my read of it, in the US right now it is apparently possible for an original copyright holder to suddenly show up after years of dissapearance and still own the copyright.

Interesting stuff!

What I may do is eventually break up the layout I already have here into "public domain pieces" as individual PCBs, like the power supply and gain stage, which would seem to get around the issue of copyrighting a specific collection of public domain pieces as an complete work.

Last edited by agdr; 25th May 2013 at 04:56 AM.
  Reply With Quote
Old 27th May 2013, 06:10 PM   #106
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
copyright only protects an exact copy of an expression - not the ideas/technology - only active, valid patents can in principle protect published circuits from being reused/reimplemented by others

even for a schematic copyright - only the exact, particular drawing can be subject to copyright - anyone can fire up their own favorite schematic entry tool and re-enter the exact same circuit - and freely publish it - protect their own drawing of it under copyright - but the topology, circuit ideas, methods, implementation, parts choices are not copyright-able

likewise copyright can protect the exact grahics image of a particular PCB layout of a circuit, or the Gerbers, even the PCB itself is considered an copyrigh-able graphic work - but anyone can redo the layout without facing restriction from copyright

Last edited by jcx; 27th May 2013 at 06:23 PM.
  Reply With Quote
Old 28th May 2013, 03:08 AM   #107
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Hey thanks for the feedback, jcx! Very helpful.

Yeah I've had some additonal PMs, emails, and discussions and the net result is that there apparently isn't a problem here. A "derivative" work apparently has to be vastly more similar than this amp. Someone has even pointed out that RS himself referred to the ODA as a separate project that he would produce at some point.

In fact, it may be the case that if I believe this amp isn't an O2 "derivative" then I should be leaving the mention of NwAvGuy off the board entirely. Still looking ito that one. That is unfortunate since I really respect his work and wanted to give credit were due! RS clearly was vastly better at this stuff than me with a lot more audio industry experience.

So apparently no problem after all. But thanks again to John for bringing it up! Needed to be discussed and would have come up eventually.

Back to the technology... if anyone etches the last set of Gerbers I posted, two things need to be done. First I've realized that the 0.51R CLC resistors really should be a higher wattage for longevity given the half wave charging spikes. I've changed those to 2512 2W SMD parts, but an easy fix on the board is leave the two surface mount 1206 SMD resistors off and just solder a through-hole axial 0.51R 2W between the two pins on the cap and inductor that the SMD traces went to for each resistor, right under the board. Plenty of space in the case for that. Or the two 0.51R resistor pads can simply be shorted. Those resistors are just a small tweak to lower the Q of the filter circuit to get rid of a tiny amount of overshoot.

Second the NJM2068 loop around the output buffers is likely not going to work, so leave that chip off and use the two 0R bypass resistors that I included on the layout to feed the output buffer chips directly from the coupling caps. I was able to reproduce the triangle wave oscillation that I saw with the LME49990 loop chips in an LT Spice simulation using the NJM2068 chips instead and the output buffers. I believe I know what is going on now. It is likely more related to slew rate, in having a faster slewing chip (the NJM2068) powering a slower slewing chip (NJM4556AL at only 3V/uS). The slow chip trying to keep up with the faster slewing chip produces the small-amplitude triangle waveform.

This one was a little tricky to figure out. The audio has a swing of up to 10V rms with the 17Vdc rails. So with a 3V/uS slew that would be about 3.2uS for 1/4 of the wave or a frequency of 78kHz for the full cycle. Still adequate for audio. But the problem is with much smaller amplitude swings (noise) which takes considerably less time even at 3V/uS, and can result in small amplitude, but fast frequency, triangle wave oscillation.

There really isn't a workable audio-capable chip with a slew less than 3V/uS to wrap around the NJM4556As that I can find, so the best solution is just not to go there with this design and leave the NJM2068 chips around the buffers out entirely. Just feed the buffers directly from the coupling caps, as in the O2 by using those 0R 1206 resistors.

Last edited by agdr; 28th May 2013 at 03:15 AM.
  Reply With Quote
Old 28th May 2013, 06:12 AM   #108
diyAudio Member
 
availlyrics's Avatar
 
Join Date: May 2011
Quote:
Originally Posted by agdr View Post
...That is unfortunate since I really respect his work and wanted to give credit were due! RS clearly was vastly better at this stuff than me with a lot more audio industry experience.....
Exactly my thoughts!
__________________
Thanks,
availlyrics
  Reply With Quote
Old 29th May 2013, 03:56 AM   #109
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Default Output balancing resistor measurements

I've attached an interesting set of measurements below. These are the DC output offset voltages at idle, with no input signal, for each half of each NJM4556AL output chip. One bank of 3 paralleled chips (6 outputs) has 1 ohm resistors, same as the O2, while the other bank has 3.32 ohm resistors.

With just two chip halves paralleled the O2 amp has an effective output resistance of 0.5R. With 6 chip halves paralleled and using 1R resistors the effective output impedance is around 0.17R. With the 3.32R resistors it goes back up to 0.55R, similar to the O2.

Smaller balancing resistors cause higher balancing currents to flow and increase chip power dissipation, but pull the various outputs closer together. The larger 3.32R resistors reduce the balancing current and allow the chip outputs to move closer to their natural idle state.

One of RS's theories was that the two halves of any given NJM4556A chip would be fairly well matched since they are on the same die, vs comparing halves of different chips. That seemed to be the case measuring the O2 with the 1R balancing resistors. But with the 3.32R resistors unbinding the outputs a bit it actually doesn't look like the case. For example, from the last entry, -83uV difference from the average on one chip half and +305uV on the other. Which means that going to multi-chip balancing in this amp, with 3 chips per channel, shouldn't produce much of a different result in terms of matching than with just one chip per channel in the O2. Just more current capability.

Going into these measurements I was not expecting the DC output offsets of the various chips to be so well matched. I was expecting things more on the order of around +4mV output on one chip and -2mV on the other to produce the 2mV average, resulting in large(r) balancing currents. Instead they are all positive voltages just a few tens or hundreds of uV away from each other.

The balancing currents involved are tiny - hundreds of uA - so neither 1R or 3.32R will heat up the chips much and subtract from the total chip 800mW dissipation capability, even at +/-17Vdc power rails. The chip quiescent current dissipates more than that. Either value would work. I've read a suggestion from a power op amp manufacturer in an app note to use balancing currents per chip (chip half in this case) of around 1% of the chip's output current capacity when paralleling chips. For the NJM4556AL that would be 70mA / 100 = 700uA. It looks like a balancing resistor value of around 0.5R would be closer to that, from the measured numbers here.
Attached Files
File Type: pdf ODA output resistors 1.pdf (37.6 KB, 17 views)

Last edited by agdr; 29th May 2013 at 04:21 AM.
  Reply With Quote
Old 29th May 2013, 04:24 AM   #110
diyAudio Member
 
abraxalito's Avatar
 
Join Date: Sep 2007
Location: Hangzhou - Marco Polo's 'most beautiful city'. 700yrs is a long time though...
Blog Entries: 109
Send a message via MSN to abraxalito Send a message via Yahoo to abraxalito Send a message via Skype™ to abraxalito
Quote:
Originally Posted by agdr View Post
One of RS's theories was that the two halves of any given NJM4556A chip would be fairly well matched since they are on the same die, vs comparing halves of different chips.
I was so surprised by his theory that I did some measurements on about 25 of these chips. Whereas most were fairly well matched that I reckon was because the offsets were fairly low anyway. I did find one chip with a big difference in offset between the two halves. Only one device is needed to invalidate a hypothesis.

Since then I found an article from TI which contradicts RS's theory. I put the link up on my blog - http://www.diyaudio.com/forums/blogs...datasheet.html
__________________
Seek not the favour of the multitude...rather the testimony of few. And number not voices, but weigh them. - Kant
The capacity for impartial observation is commonly called 'cynicism' by those who lack it.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Building the O2 Headphone Amp 6L6 Headphone Systems 37 20th October 2014 05:46 PM
Help with BantamDAC and O2 Amp jschristian44 Headphone Systems 2 27th January 2012 06:22 PM
O2 Headphone amp for sale! agdr Vendor's Bazaar 10 9th December 2011 04:05 AM
o2 amp building jtktam Vendor's Bazaar 0 3rd November 2011 10:52 AM
Questions before getting parts for the O2 Amp dewasiuk Headphone Systems 3 12th September 2011 05:33 PM


New To Site? Need Help?

All times are GMT. The time now is 08:29 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2