[goodsound]

After weeks of trying to determine which direction to go for building a headphone amp, I have finally narrowed it down to these two chips -
One is a complete solution (TPA6120) whereas the other is a output buffer(LT1210).

What I am looking for is a simple, compact/portable headphone amp.

If its the LT1210 then it would have to be a "buffer-only" amp, no other driving stages please.

So which one should I go for ? Has anyone listened to both ? Any suggestions you could provide for helping me make the decision would be highly appreciated!

Thanks!

<edit:> I will be using it with a Grado SR-80 but that is not a constant. It could change to another headphone down the line. And I will be using it with a variety of sources (Sound Card, PDP, PCDP, DAC, CDP)

 

[rickcr42]

Why the LT1210 instead of the much better sounding LT1010 for the buffer ?

 

[goodsound]

well my understanding is that typically in a "buffer-only" configuration i.e. buffer outside the feedback loop you want to start with something that has higher output current. and LT1010=150ma, LT1210=1000ma.

is either better sounding than the TPA ?

 

[rickcr42]

Couple of points/quesitions

1-what on eath do you need a full 1A output for ?
2-large amounts of current delivery is not the only thing you need to look at but IF that current is delivered in a way pertinant to audio use
3-look at the battery drain specs then run away fast
4-thos is a current feedback amp and NOT a true buffer even though can be used as any other opamp type in unity gain and as a buffer
5-Tried it and thought it sucks but that is me
6-I know of no portable duty rig that is not fine with 100mA current output no matter what others say
7-The LT101 kicks it a*s in every aspect and that no contest.especially run at class-A operation though not real good for battery use


Quote:

is either better sounding than the TPA ?

The TPA is an map,not a buffer.I have four,have had them for over a year and still have not felt the urge to use them but others report good sound

 

[goodsound]

its not about "output" current.

would you mind sharing the schematic of the LT1010 that kicks @$$ and also the LT1210 that sucks ?

 

[rickcr42]

Straight from the data sheets with the exception of I use a jfet front end to the LT1010 (not needed except in extreme cases,me

)

The LT1012 is not meant for audio and while that alone does not preclude its use the dta sheet really does tell the tale.If you want to drive a servo motor for an automation or robot however,great chip,but only if it is heat sinked properly.The 1012 is also too high a drain on the batteries for portable use and finally...why do you feel you need a full 1 Amp current delivery ?

That is 10WPC AB power amp territory and with it a huge heat sink is needed or you will not even get close.

The LT1010 can get 150 mA linear current delivery and can be biased with an external resistor for seriously deep Class-a operation.Other than the support circuitry internal to the chip it is really a BUF-02 clone (obsolete ADI part) without the jfets (and why I added my own

),will operate at low voltages and has a civilised idling current until you "tweak it" for higher bias-all good for portable duty as a straight buffer and to me personally sounds better than other buffers in actual use.

 

[goodsound]

just to make sure we are not talking about different chips - I am referring to the LT1210 and not the LT1012.

 

[rickcr42]

I realise that or would never have asked why you think you need a 1A output current.the other just a typo

 

[Tomo]

Hey,

No getting hostile people. LT1210 seems usable. Noise level is decently now. Pretty darn fast and you can use unity gain. Not to mention high bandwidth. Please do becareful, however. It is a current feedback amplifier (or CFB amp). Thus, you should be real careful about the feedback you install. (See specs for details.)

LT1010 is an entirely different beast since it is a true buffer. And, also it seems it is quite stable outside of the feedback loop! Confirm this plz.

Tomo

P.S. Hey I might try LT1010 out. Thanks.

 

[rickcr42]

Quote:

No getting hostile people. LT1210 seems usable. Noise level is decently now. Pretty darn fast and you can use unity gain. Not to mention high bandwidth. Please do becareful, however. It is a current feedback amplifier (or CFB amp). Thus, you should be real careful about the feedback you install. (See specs for details.)

one of those "looks good sounds not so very" chips.If a person is hell bent on using something there is no opamp.current feedback amp or buffer that can't be used to make sound but does not mean a good choice.

Quote:

LT1010 is an entirely different beast since it is a true buffer. And, also it seems it is quite stable outside of the feedback loop! Confirm this plz.

I use them 100% open loop stand alone zero gain stage and to be honest could not live without my simple "buffers".
A person with a good preamp or a CD player with a volume control really has zero need for a headphome amp but does for an Impedance Converter/High current Driver if they want to use headphones or drive long lines and even in a portable type set up many times it is just "muscle" that is needed and not gain so an amp there is again overkill.

I like the 1010 over the HA's and the BUF634 but that is preference ,kinda like how some like one opamp over another.

you should check a couple out Tomo,i think you may like the resulting sound but do get the 5 Pin "CT" package so you can get a proper heat sink on it then use the bias resistor at around 20 ohms

 

[jcx]

if you mean portable as in battery power, the tpa and 1210 don't come out on top of most lists

at headwize you can find a project thread (or two) using the AD8397 since it swings R-R on the output, which neither of you choices do

on pure specs the cfa amps also seem to loose when you carefully look over the distortion (tpa6120 imd curve looks unfortunate for a "0.00014 %" distortion amp)

I've prototyped headphone amps with cfa op amps inside good audio vfa op amp feedback loops, using gain in the cfa output stage as Walt Jung shows in his multiloop circuit examples - even a tl072 can give great distortion # (including imd) with a cfa doing the heavy lifting for it

 

[rickcr42]

the main problem as I see it is-using a feedback amp as a buffer where an open loop actual buffer will do.The jung style multiloop VFB opamp/CFB opamp addresses one set of problems and opens the door for another set which is by having the buffer in the loop and the loop attached to the load any reflections or RFI can not only get in but be fed back to the inverting input.
Since this signal is not also present at the non-inverting input there will be no common mode cancellations as there would be with RFI on the input and will be treated as a signal to be amplified.
Yes there are ways to isolate the output and yes they are effective but so is simply leaving the buffer outside the loop as a true voltage follower.

CFB same thing.Isolate the output or risk contamination because it uses feedback and why I personally choose to use a true buffer run open loop to a "feedback amplifier" used as a buffer.

Quote:

at headwize you can find a project thread (or two) using the AD8397 since it swings R-R on the output, which neither of you choices do

This op-amp has also been discussed extensively here at head-fi.Do a search.No need to go off site

 

[peranders]

TPA6120 alone is pretty nice as headphone amp. The current capability is more than enough. According to me it's more than OK.

My TPA6120 look like this:
http://home5.swipnet.se/~w-50674/hif...0_overview.jpg

 

[goodsound]

Quote:

the main problem as I see it is-using a feedback amp as a buffer where an open loop actual buffer will do.The jung style multiloop VFB opamp/CFB opamp addresses one set of problems and opens the door for another set which is by having the buffer in the loop and the loop attached to the load any reflections or RFI can not only get in but be fed back to the inverting input. Since this signal is not also present at the non-inverting input there will be no common mode cancellations as there would be with RFI on the input and will be treated as a signal to be amplified. Yes there are ways to isolate the output and yes they are effective but so is simply leaving the buffer outside the loop as a true voltage follower.

Ok. Now I see what you’re saying.

Quote:

I use them 100% open loop stand alone zero gain stage and to be honest could not live without my simple "buffers". A person with a good preamp or a CD player with a volume control really has zero need for a headphome amp but does for an Impedance Converter/High current Driver if they want to use headphones or drive long lines and even in a portable type set up many times it is just "muscle" that is needed and not gain so an amp there is again overkill.

And that is exactly what I believe and want to implement, hence the whole “buffer only” scheme. But then I bumped into the TPA6102. Its sheer simplicity and favorable opinions drew me towards it. So that became an option as well.

Thanks for bringing up the AD8397, yet another good option!

Ok so we are past the first stumbling block – the lt1210. But I still don’t have enough information to decide between these great choices!

 

[morsel]

Have you seen the PINT and Mini³ projects? They use the AD8397.