[qusp]

A qiuck question... are there any inherent flaws in using a chip like AD797 as a buffer?  If it sounds good, y not?

yes, as ive covered here a number of times, its really not what I would call unity gain stable. they are stretching the truth in the datasheet, it needs a very well laid out and well executed PCB to make this work properly, its very prone to oscillation otherwise and not exactly happy driving capacitive loads (headphones), or low impedance loads*

basically it requires a heap of boxes ticked that are unlikely to be present on a PCB not designed for it

*its not exactly high in current, though higher than many, 30mA isnt exactly what I would expect from a buffer. For example the BUF634 or LME49600/10 put out 250mA and are truly unity gain stable.

 

[kiteki]

Ah I see thanks.  I think BUF634 sounds like rubbish, so how about OPA627 what if I use that as a buffer what happens then?

 

[SpudHarris]

Have you tried HA5002? it's my favourite buffer of all...

 

[qusp]

Ah I see thanks.  I think BUF634 sounds like rubbish, so how about OPA627 what if I use that as a buffer what happens then?

haha, I think its more likely the way you (or whatever ebay dac it was) used it sounded rubbish, to say the Jung diamond buffer itself (the discrete circuit both BUF634 and LME49600/10 are based on) sounds rubbish, is quite frankly.... rubbish.

opa627 is unity gain stable, but again current limited for true buffer applications. it will work with a good amount of headphones though, as many dont need much current. very low impedance headphones/IEMs will perhaps present some difficulties if driving direct with opa627

Spud: nah I never bothered with the HA500X, I like the LME49610 because of higher voltage and good current/power capability if used on a suitable PCB and if I want more, or something really different I build something discrete, since I now have a pretty large parts bin, with many and varied mosfets, jfets, transistors, passives etc I can build most things I need on the spot.

BTW soz mate I just saw the message you sent me last week, did you sort that out or should I take another look? its been a while since I used my amp so i'll have to dig it out to check for you.

 

[kiteki]

I tried ordering it, ceramic and all, then they cancelled my order and said they couldn't find it.  Right now there's only one on ebay, selling for $25 a piece, seems like way too much hassle.
 
Likewise, I placed an order for LT1010, then noticed the seller had too much negative feedback... cancelled it.
 
Can I convert an LME49600 into DIP form by twisting it around?

 

[kiteki]

Originally Posted  qusp /img/forum/go_quote.gif

btw soz m8

 
No worries m8, I'll yuse da oppah627 as a buf den, since it shud hav enuf current 4 da IEM applekayshuns no?

 

[qusp]

I tried ordering it, ceramic and all, then they cancelled my order and said they couldn't find it.  Right now there's only one on ebay, selling for $25 a piece, seems like way too much hassle.

Likewise, I placed an order for LT1010, then noticed the seller had too much negative feedback... cancelled it.

Can I convert an LME49600 into DIP form by twisting it around?

no, you cannot, it uses PCB for heatsinking and the tab is live negative voltage. people have done it with the use of isolated heatsinks, but personally I think its a silly idea and you should just use BUF634

No worries m8, I'll yuse da oppah627 as a buf den, since it shud hav enuf current 4 da IEM applekayshuns no?

I wasnt talking to you, since it still appears you havent read anything about ohms law, I guess you dont know that IEMs are the least likely to cope with it. you'de have better luck driving HD600 or even DT880/600 than many IEMs. high impedance = low current, low impedance = high current

how in hell people can mod stuff for years and still not read up on Ohms law I have no idea

 

[SpudHarris]

[COLOR=FF00AA][/COLOR]

BTW soz mate I just saw the message you sent me last week, did you sort that out or should I take another look? its been a while since I used my amp so i'll have to dig it out to check for you.

No worries pal, strangely enough I got a mail the following day off Ryuzoh asking me to review a new DAC and he says the Fi-q works best via mains with switch in position 3. I still love mine but of late most of my listening is with HE-6 via a Topping TP60 speaker amp and In my limited experience it's the best I've ever heard..... Thanks anyway bud.

 

[SpudHarris]

Double post - Sorry

 

[kiteki]

high impedance = low current, low impedance = high current

 
So a DT770 / 600 ohm is low current, and a sensitive 16 ohm IEM is high current?  I don't get it, but thanks anyway.

 

[fmlfml]

I used to have a Audio GD OPA Earth,
it sounds quite good and you might want to try it out!

 

[Mad Max]

Well, at the very least those Audio-gd opamps are slow enough to be dumped into most circuits without them overheating, and they already have a wire for grounded psu bypassing.
The stock MKT bypass caps don't quite do them justice, in my opinion.  You have to replace them with Mundorf film cap(s) or Russian PIO K42-Y2 cap(s).  I've gotten major improvement in sound with ceramic and tantalum caps, never tried the Mundorfs or PIOs.  I still have Sun, my favorite of the three for its energetic and forward presentation.
I dislike how incredibly bulky they are compared to chip opamps, and they can't sound as good either.
 
Now I remember that I never got around to trying out 827, 4627, and 797 with those Russian capacitors.

 

[kiteki]

to say the Jung diamond buffer itself (the discrete circuit both BUF634 and LME49600/10 are based on) sounds rubbish, is quite frankly.... rubbish.

 
I'm not familiar with diamond buffers but the BUF634 isn't transparent sounding.  From what I've seen around h-f and diyaudio I'm not alone in that opinion.  I'm using it in an iBasso amp, not an "ebay DAC", not that there is anything wrong with ebay DAC's!, is there?
 
 

you'de have better luck driving HD600 or even DT880/600 than many IEMs.

 
Can you expand on this?  So let's say a full-size headphone and IEM of equal impeance, let's say 50 ohm, doesn't the full-size ideally need quite a lot more mA?

 

[kiteki]

I still have Sun, my favorite of the three for its energetic and forward presentation.

 
Have you tried any other discretes than from Audio-gd?  I know qusp mentioned that the AD797 inventor is releasing a DIY discrete.

 

[Mad Max]

Quote:

 
Have you tried any other discretes than from Audio-gd?  I know qusp mentioned that the AD797 inventor is releasing a DIY discrete.

 
Yes, when I said that Sun is my favorite, I did mean to imply that I tried Earth and Moon as well.  These three plus OPA2134 and LME49710 were the first opamps that I ever rolled.  Earth was my favorite... but then I tried ADA4627-1B and never looked back... and then I began to rediscover my opamp collection with extra psu bypassing, so that is when I revisited the Audio-gd opamps.  I ultimately sold Earth and Moon and kept Sun.

No matter what I did to Moon, I never liked its weird sound. Versus OPA2134, LME49710, and a few LT and other TI chips, I liked Earth a little better as it sounded more "right".  4627, 797, OPA827, and several other TI and Intersil chips all make Earth sound muddy, overly colored, and unnatural by comparison.  But I compared the latter bunch versus Earth with all of them having extra psu bypassing.  It seems to me that without the extra bypass caps, Earth may smoke those others in sound depending on the circuit.  Well, many of those chips aren't great candidates for blind rolling anyway.  =p
 
Sun can seem to have a strangely roller-coaster-like FR without upgraded psu bypass caps, really bizarre.
Oh, then there's the matter of the A-gd opamps having fatiguing treble when they are new and not burnt in.  =/