Summary: Portable amp for low impendence phones Design options & component selection
May 24, 2002 at 9:28 PM Thread Starter Post #1 of 10

Nicwix

Survey says....
Joined
Aug 5, 2001
Posts
96
Likes
1
Exciting Times: Discussion on Portable amp for low impendence phones

The story so far ...

I started the Headwize thread thread on Headwize (id 2703): Best opamp-buffer combination for a portable multi-loop amp for low impedence phones assuming the multi-loop configuration was in fact the answer to my needs. I also placed a cross-reference thread in this Head-Fi forum - thread 12157 ).

The only real variables that I was concerned about were which opamp and buffer and buffer would be best to use.
confused.gif


This developed into a discussion on well various chips performed at lower voltages (as per datasheets or actual testing) and the associated current requirements, and hence battery capacity needs and options.

At that point PRR replied with a few helpful comments, such as:
Quote:


"I don't know what "multi-loop" is doing in there; you want good sound, not a certain number of loops."

"Your requirements are not synergistic."

and

"You can get ample speed and current with 2V total supply, IF you design for it. Putting on more voltage to overcome losses in a circuit not optimized for low voltage speed/current is just wasteful."


Aghast to discover my requirements were unsynergistic
frown.gif
, I restated the design goals as:
1. high-end sound quality
2. small, portable package (board and battery)
3. good battery life (say 20 to 40 hours)
3. high current capability
4. as minimalist as possible (I like to know what each component is there for)
5. reasonable cost and easy to source components
6. relatively easy to build and package

PRR also seemed to be questioning the need for loops.
redface.gif
In his response, PRR went then on record saying he was "not convinced it [the multi-loop topology] is best". No wilting violet is our PRR.
tongue.gif


PRR then suggested some simplified output configurations (not full designs) and later suggest that the LMH6642/6643/6644 3V, Low Power, 130MHz, 75mA Rail-to-Rail Output Amplifiers might be worth trying.

I expressed my concern about the output impedence of buffers (specifically the EL2001) without feedback - which led to PPR suggesting that the EL2001 was an old design that had limitations - this pricked some holy cows indeed.
eek.gif
At that point Morsel though her hat into the ring (
wink.gif
- Morsel sans hat), then ppl weighed in big time.
very_evil_smiley.gif


The discussion continues hotly in the Headwize thread.
evil_smiley.gif
evil_smiley.gif
evil_smiley.gif
We look forward to the ongoing debate - anyone else willing to venture an opinion?


In the meantime, tangent has kindly pointed out that the META42 does meet the stated requirements pretty well (at least in the opinion of some)!!!

Nix
biggrin.gif


ps (Added after tangent's post below) The various smilies in this post (actually I maxed out the smilies) are intended to clarify that this whilst poking a bit of fun around, only goodwill is felt towards all participants - and though we may have heated debates, we will still remain friends.

All the best - Nix
 
May 25, 2002 at 4:19 AM Post #3 of 10
Nicwix, you must realize that my writings in these various threads have been part fact and part opinion. There's also some limited experience thrown in there.

Listen to PRR -- he knows where his towel is. I'm not saying that the META42 isn't worthwhile -- just that if you want the best answer, you should listen to all the opinions, all the facts, and then make a decision.

Good luck with your quest!
 
May 25, 2002 at 6:46 AM Post #4 of 10
Actually, there are no actual facts - only one's opinions of the facts
biggrin.gif


And actually I think it is great that people have contrary views - nothing better than a healthy debate. And well done to PRR for being prepared to stand up and justify his views. I respect the views of both ppl and PRR - that makes it interesting when they differ.

I think that you and Morsel have done a fine job with the META42 - the design appeals to me a lot and building in flexibility the way you have obviously added an extra level of PCB design difficulty.

Nix
 
May 25, 2002 at 8:55 AM Post #5 of 10
Budgie:
I think your Right I think Nicwix is just trying to Create Contraversy by his post. I might do well to avoid Replying to His posts in the future> I do have better things to do with my online time than Debate with People.

Nicwix:
Maby you should just Design this Amp your self then share with the rest of us what you did.

Tangent:
I have great Respect to the technical knowlage of PRR, he is one of the few on this fourm that has an Extensive know-how of Electronic Design. However I stand by my reply to that thred. My reply is technicaly accurate and the EL-201 dose indeed sound wounderfull as well as Measures well. In comparrison the BUF-634 sounds slow. The BUF-634 is a great part and i have used it alot in Headphone Amps. If the EL-2001 becomes unavailable i would use the BUF-634 without reservation. The EL-2008 and 2009 is IMHO the Best output stage available in Monolithic form to drive Headphones but since thay are Gone I am working on my own Discreet component design to replace them i will share that with this fourm when the it is complete.
 
May 25, 2002 at 9:29 AM Post #6 of 10
Quote:

And actually I think it is great that people have contrary views - nothing better than a healthy debate. And well done to PPR for being prepared to stand up and justify his views. I respect the views of both ppl and PPR - that makes it interesting when they differ.


You might want to refer to him as PRR
smily_headphones1.gif


PRR definitely appears to have a very sound grasp of the issues at heart.
 
May 25, 2002 at 10:57 PM Post #7 of 10
Quote:

Originally posted by ppl
Budgie:
I think your Right I think Nicwix is just trying to Create Contraversy by his post. I might do well to avoid Replying to His posts in the future> I do have better things to do with my online time than Debate with People.


ppl

My apologies if I have upset you.

As I see it, the whole point of this forum to debate the pros and cons of different designs, is it not?

As I have said from the outset, my intentions are to build the multiloop and hence requested recommendations on components. As we all know, the multiloop is endorsed by Jung, yourself and is widely supported in this forum.
PRR (whom we also respect) questioned my design goals and chosen solution (which I appreciate). He also was willing to challenge the group thinking on multiloop and EL2001 and back his views with details. No one has ever done that before.

When two people, with the depth of knowledge that you and PRR have, disagree on a design issue, there is great value in debating the difference - that way leads to greater understanding for everyone and hopefully improvement. And if anyone else is able to contribute, that would be wonderful.

Tangent has assessed the multiloop against my design goals and also provided useful feedback on his measurement and testing of potential component choices. There has also been useful discussion on battery options - which obviously interacts with your component choices.

I am genuinely interested to understand if there alternate designs that better achieve my goals than the multiloop. To that end, yes I am trying to stir debate.

But, if my approach did not work for you, you have my sincere apologies.

All the best, and please do continue with your great contributions.

Nix
 
May 25, 2002 at 11:04 PM Post #8 of 10
Quote:

Originally posted by ppl
... I am working on my own Discreet component design to replace them i will share that with this fourm when the it is complete.


ppl

Dare I ask, will it work on a low voltage supply?

Nix
 
May 25, 2002 at 11:13 PM Post #9 of 10
Quote:

Originally posted by Joobu

You might want to refer to him as PRR
smily_headphones1.gif



I edited all the above posts - it's a bit like using an AD823 in place of AD843
smily_headphones1.gif


Quote:

Originally posted by Joobu

PRR definitely appears to have a very sound grasp of the issues at heart.


Indeed

Nix
 
May 26, 2002 at 9:35 AM Post #10 of 10
As I see it PRR was claiming that Class AB Output stages will not have an Impedance rise at the Crossover Point? That selecting a value of Emmitter resistor in an Emmitter follower stage will result in a flat impedance vs load Current ?

If the above is what was implied then I take issue with that since First any Non class A output stage will indeed have an impedance rise at the crossover point unless you eather add series resistence between the load and the Output stage or as I think Prr is trying to say that using a higher value of Emmitter Resistor, whitch is in Essence doing the same thing.

Regarding the EL-2001 I have sed Before and I will say it again this Is the Best sounding Headphone driver I have Heard to Drive Headphones With at least 100 ohms of Impedance. The Rise in output impedance is not an Issue at these impedances. I have also stated alot of times that i dont like this Buffer for use with loads lower than 10 ohms Like My Sony's. For them I use the EL-2008 (Best Bang Per Ma) or as an alternative use at least two (2) EL-2001's in Parralell and Perfer (4) if batteries Premitt. or you can use 10 to get the Output curent and sound of the 2008. ALSO NOTE That the Output Impedance is Reduced by the number of Buffers used in Parralell.

I would also like to Add that since My first recomendation of this Part to this Fourm Not one Person that has used it has sed anthing Negitive about its sound. This Buffer is used by some Quite well respected Audio Designers as well. (Ill Leave them nameless Don't ask who thay are, Non disclosure)

I also stated that I like Walt's Multiloop design and I think PRR implied that is is not the number of loops used. This also is Contrary to what i and alot of others have found using Opamps. I do not like to use any output stage Open-Loop, Yet I also like to Keep the Open Loop bandwidth well above the Audio band so as to maintain a constant amount of negitive feedback and thus Constant Distortion across the Audio range. The most Elagent way i have found to do this with a generic opamp/Buffer combo is Walt's method. I use it I like it It sounds good to me.

SO IF ANYONE TAKES ISSUE WITH THE ABOVE so be it my intent in my contrabutions to this fourms is to inform of what I works for me, Will it work for you, Heck i dont Know for 100% Certen Because No one can antisipate what someone else will do with an Idea, Be that mine or someone Elses. Let you Ears decide.>>> PPL


FYI> http://www.head-fi.org/forums/showth...ght=ppls+beast and also see http://www.head-fi.org/forums/showth...ghlight=review and also see http://www.head-fi.org/forums/showth...=&threadid=473
 

Users who are viewing this thread

Back
Top