Discrete Buffer vs. IC
Mar 29, 2006 at 4:00 AM Thread Starter

#### bingalls

Argghh...I have a feeling these may be dumb questions, but I've searched high and low for answers without luck..honest, I have.

The datasheet for a BUF634 lists quiescent current draw and maximum current output. How does one determine these 2 numbers for a simple discrete buffer such as this?

Is there a formula to come up with this or am I way off track?

Mar 29, 2006 at 4:09 AM

#### star882

Output current is determined by the SOA of the transistors.
Quiescent current will vary depending on exact component values and matching. Best way to find out is to build it and measure.

I really recommend an IC since it has been carefully engineered to work good. Not to mention a lot smaller and possibly cheaper!

Mar 29, 2006 at 4:12 AM

#### Pars

##### Can Jam '10 Organizer
Yes, but the discretes sound better than a BUF634 (IMHO). Do a search for a BUF634 thead here with posts by user mono. His version of sijosae's discrete sounds better in my son's Pimeta than the BUF634s (dual) in wide BW mode do.

Be aware that the discretes do not have output protection, so a short on the output (such as unplugging phones with the amp on) can fry them.

Mar 29, 2006 at 4:34 AM

#### rickcr42

##### Are YOU talkin' to me?
not all buffers are created equal.I would take a LH0033 over a discrete buffer with it being a tossup between the LT1010 or discrete otherwise.but that is for a stand alone voltage follower/buffer.

If we are talking about "booster stage" following an opamp then it would be a SEPP Class-A follower stage

Mar 31, 2006 at 3:11 AM

#### Tomo

##### DIY tube amps can be SHOCKING
Hi,

Quiescent current draw is the current that the chip consumes when there is no signal.

Maximum output current is a little controversial factor. It may mean the maximum current the chip can deliver to a short circuit. Or it may mean the maximum output to a specific load. You will need to refer to the specification sheet for details. (I know spec sheets are hard to understand, but you got no better source of information. ... Other than begging for it in forums
)

In terms of sound quality, I would go for BUF634 anyday. It is far easier to build with and to maintain. I never need to match transistors. Also, the footprints are easier to deal with. Thus, it is much easier to build excellent humm-less circuitry with BUF634.

For me, who is a quite imperfect DIYer, building better circuitry means more than perfecting audio circuit topologies. In fact, for me, "upgrades" did less improvement compared to "careful rebuilding." DOUBLE THUMBS UP for BUF634.

Tomo

P.S. Or LT1010

Mar 31, 2006 at 4:44 AM

#### nikongod

##### DIY-ku
i will oppose the seeming masses.

in the case of just building for fun TRY IT. build a descrete buffer stage. nest it with your lm-6171 and bass boost... play around, try new stuff.

as long as your "experiments" dont destroy anything too expensive (and even "cookie cutter bulders destroy parts) you have only some small expendatures to make, and may find a sound you absolutely love. at the very least if you like doing this stuff it will be fun.

Mar 31, 2006 at 5:37 AM

#### rickcr42

##### Are YOU talkin' to me?
Quote:

 In terms of sound quality, I would go for BUF634 anyday.

a shame you will never get to try the obsolete LH0033 buffers tomo.They flat out just sound great even though on paper they suck in comparison to modern buffers.Mine are the ceramic DIP MIL versions and a better buffer i have never heard including any discretes i have heard.

OK.Maybe my pair of LH0063's but those puppies are like Mongo-Buffers and not a nomal parts with their 4x5 inch heat sinks for only 100ma output current due to the Class-A operation

the BUF634 is about a simple to use as it gets and will not shame most systems even used as a straight open loop buffer just hanging off the output of a CD player ot line stage but this is in a serious shortage mode because of the transistion to lead free parts and apparently all the old leaded parts are sold out everywhere and the "green" replacements are slow in coming as everyone trying to build the Meier HA-1 is finding out (see thread on Corda HA-1,Corda Cross-1 boards herein DIY).

Mar 31, 2006 at 8:21 AM

#### Tomo

##### DIY tube amps can be SHOCKING
Well ...

I was not entirely being honest. I like the idea of buffer amp (discrete or integrated). But I think the gentleman above is referring to the buffer application in Buffer-Opamp pair.

Tomo

P.S. You know I like discretes. But you probably know my primary amp is ... an opamp amp! (I am building Szekeres amp now and it's gonna take over till I build LT1010 amp!)

Mar 31, 2006 at 1:41 PM

#### rickcr42

##### Are YOU talkin' to me?
Nice Tyre,very nice indded.So what do you think of the sonics ?

Quote:

 But I think the gentleman above is referring to the buffer application in Buffer-Opamp pair.

Likely yes but that does not mean the buffer section is an afterthought.a good stand alone buffer works in the loop even though I do not like to use them that way (leasves the loop open to problems at the output).If it is good alone it should be good in combination

Quote:

 P.S. You know I like discretes. But you probably know my primary amp is ... an opamp amp! (I am building Szekeres amp now and it's gonna take over till I build LT1010 amp!)

where is the old one ? You are the first human I know that actually built the Szekeres and way back in the dark ages when I was getting back into audio electronics were the one I went to for answers.Since then we took it to different directions with you going to active loading and me passive but the proof that yours existed is in the Headwize library so what gives man ?

Finally blow it up from all the experiments ?

Mar 31, 2006 at 7:21 PM

#### tyre

Quote:

 Originally Posted by rickcr42 Nice Tyre,very nice indded.So what do you think of the sonics ?

I haven't had a chance to try them out yet. I'm currently using an AD815 dsl driver as a buffer at the moment and it seems to work good.

Speaking of LH0033's, I just recently won a set of 12 on ebay for \$15! I'm thinking of trying a protoboard PPA with 4 LH0033's per channel.
How well do these things work in parallel? Would each buffer need a resistor on it's output?

Apr 1, 2006 at 3:55 PM

#### rickcr42

##### Are YOU talkin' to me?
Quote:

 Speaking of LH0033's, I just recently won a set of 12 on ebay for \$15! I'm thinking of trying a protoboard PPA with 4 LH0033's per channel. How well do these things work in parallel? Would each buffer need a resistor on it's output?

My personal preference would be to not waste a buffer having superior sonics in a design that would comprimise the sound.That is BUF634 or HA territory and would be a waste of time to use a 70mA buffer to get what "on paper" a single BUF634 can get.

This buffer is best suited as a pure follower/driver stage and that means no loops,no opamps,nothing but the buffer or you degrade its sonics by forcing the sonic signature of the "looped" circuit on it and that would be a shame since this puppy runs very deep into class-A (why it is only 70mA
) and is really in essence four internal devices.

I run mine "stand alone" in its own box with a shunt regulated power supply.On the back panel is the intput "trim" control to match my Grados to the 300 ohm transformer headphone output on my line stage so no matter what headphones i use the volume is the same when the master volume is rotated.

100K in Z,use the DC blance control,add some resistance to the output to take the cable out of the equation and hang one off a CD player otr preamp and you have a "pure" headphone driver with good sonices.Toss a transformer on the output and it makes a good long line driver