Do you use a Buffer by itself? Do you use an opamp by itself?

[Tomo]

Hey all,

Some of us have experienced nice results bypassing the voltage gain stages of our amplifiers. (Some of us never build the voltage gain stages. ME!)

Some of us have experienced nice results from single opamp and not using a buffer. (I did this too.)

You all throw in some comments.

Ciao,

T

 

[rickcr42]

Both the above Tomo.

straight med-high current Op-amps for portable amps (use zero opamps at home) and have both discrete and monolithic current amplifiers at home for line driving/headphone driving.

Available "in house/in use daily" buffer/drivers are :

1-Straight LH0033 in the ceramic DIP package @ 70mA Class-A
2-Straight LH0063 in the TO-5 package @ 150mA Class-A
3-JFET White follower/LT1010 in the TO-220 package @ 70mA Class-A
4-JFET White follower/SE MOSFET in TO-220 package @ 120mA Class-A

I have a gain stage (6SN7 GC/CF0/WOT+DC Outs) but never use it until I need the extra gain for driving my power amps which are all two stage driver tube-output tube designs so I guess you could say my buffers are my headphone amps even though ALL the controls except a mute switch are in my passive control console.

 

[twofishy]

rickc, are these all on seperate PCBs for easy swapping? Would be nice if you could post some details on how you connect the 6SN7 to these

 

[rickcr42]

Quote:

are these all on seperate PCBs for easy swapping?

No,each is a stand alone project that ended up being a keeper over the years so have found a place anywhere I needed to either drop the output impedance or raise the output drive : long line drivers with 600 ohm output transformers,headphone drivers following passive or high impedance tube stages,driving LCR stages,the record output send to a remote dubbing panel from my control master console then more at the dub panel output as a distribution amp to the actual decks,cue output of my passive summing mixer,anywhere I need to either drive a long line,split a single line into multiples or drop tube stage impedance without having the voltage gain reduction of a transformer so I have a bunch

Quote:

Would be nice if you could post some details on how you connect the 6SN7 to these

Piece of cake man.Any tube pre can be run into a straight buffer stage used as an impedance reducer/high current driver as long as the buffer has a jfet input otherwise you need to add your own.Monolithic Bipolar input buffers don't like being driven by impedances over the 100-600 ohm range without input buffer assitance (monolithic fet input devices no problem) and if you REALLY read the data sheets it is pretty much spelled out so since a single-ended tube stage is a high impedance device the match is not a good one without compensating for the impedance mismatch.
If the tube stage is transformer coupled you satisfy the buffer input drive requirements but since a transformer is a gain reduction device with even the least of them being 4:1 reductions you may not have enough voltage drive for the application so why I have added "choice" to my own system.

The first stage of the 6SN7,the grounded cathode gain stage section,has a cap coupled output that taps the signal before the cathode follower stage.Do a google on VTV 6SN7 line stage or 6SN7 line stage and you will get hundreds of hits with many showing what is known as a "special" output as described.Makes good sense to use only as much in the signal path as needed so why I went that way.

The cathode follower stage has both a cap coupled output for the "high gain" setting and a transformer coupled out put for the "low gain" setting so again I have the option to use what I need and no more depending on what the present need is.

Satisfy the input requirements of the buffer and it will take care of the output drive-simple..........

 

[NelsonVandal]

Good thing you started this thread Tomo.

Please show the schematics of your successful projects just using buffers as headphone amp, or other good minimalist projects. It's time to throw all those unnecessary, crappy and expensive amplifying stages away!

Bad opamps are killing good music!

 

[rickcr42]

Quote:

Please show the schematics of your successful projects just using buffers as headphone amp

pretty much variations of a theme so for background

http://www.headwize.com/projects/sho...awliw1_prj.htm

http://www.stereophile.com/solidpreamps/54/

hint-download and install this program : http://sector7g.wurzel6.de/pdfcreator/index_en.htm the go to the Stereophile article,scroll down until you see "print this article",click that and a window will pop up that has the entire article intact as a single file,go to "file" on your browser then click "print" and when the window opens choose "PDF Creator" as the printer and print the whole damn ball of wax directly to your desktop for either later printing out a hard copy or storing on a CDROM with other articles

The National "LH" buffers are obsolete,have been for a long time but there are a few kicking around the internet at a premium price plus I beleive are being "cloned" by another semiconductor manufacturer that purchased the rights though not sure on this.

The Military/Aerospace ceramic DIP package I doubt you will find nor the LH0063 and why I guard mine as if a treasure coveted by others

Whats makes these parts nice is they are old school inefficient designs being more a "four transistor discrete on a common chip" than high density monolithcs and also because of this low density design can run class-A for the majority of operation there being no need then to cram a bazillion devices into a battery operated hand held device.In fact the 63 is the size of a small power amp when mounted on the heat sinks yet I figure the output is no more than 2 Watts (if that

) and the current delivery an embarrassment when compared to modern buffers.

BUT !

Quality over quantity.Having gobs of current is not always a good thing if that current delivered is better suited to driving a servo-motor than it is headphones and the LH0033/63 are musical even if current shy by modern standards.

The closest I get is the LT1010 which is again current shy when compared to the commonly used buffer chips and AGAIN to my ears has a better SQ than the competition but it is a bipolar input device so if you can't drive it with an impedance in the 50-100 ohm range you are feeding the device with a source that can actually showcase the ability and why most toss an op-amp in front,which as you noted is an SQ killer because IT will dominate the final sound of the amp.
Then they take a buffer that is fine "out of the loop" and like everyone else doing buffers (no damn imagination

) screw it up even more by creating a path from the output terminal to the opamp input terminal which worse case lets in cantamination AND all this while not adding any loop protection from this potential inlet of RFI so a worst case nightmare scenario...........

The answer was provided by a once readily available "Mods" file at the Welborne site that has been absent for many years but using ANOTHER TOOL (Rick likes his tools

),the "Wayback Machine" was able to dig out a copy of the perfect solution,a solution I beleive designed by Erno Borbely of Hafler and Borbely Audio fame.....(10MB download !!!!!!!! But a virtual design manual for audio so worth it.)

link

Scroll down to the Headphone Amp page and there it is.Simple,low cost and will work with ANY buffer not having an FET buffered input for driving the internal output stage (BUF634

).

Not rocket science but working backwards away from all the complications most seem intent on adding and that ALL impact the SQ of the end project.Simple is always best when there is a choice as are options so it is my way to add as many usable features as I can when it comes to the Switching/Input/Output matrix but to make the active circuitry as minimal as I can and still have it do the job asked.Far better to concentrate on feeding the active device than to throw more active devices in the mix,shaking it it then HOPING what comes out the other end is not a total sonic butcher

As usual just my opinion.Not the only path just the path that serves me best in MY system according to my own taste in what makes for good sounding audio

oh yeah-The Welborne article shows 45 ohms 1/2 watt for "R3" .This sets the bias of the output stage,how far into Class-A it will go,so for more you would decrease this resistor,less increase it but there IS a limit to how far you can go before burning up the chip !

I have used as low as 20 ohms and fried a couple while others survived but they got noisy over time from the abuse so the lowest possible value would be 22 or so ohms with the best "safe" value all else considered 25 or 27 ohms.I also suggest you go to a higher wattage part here and to steer clear of resistors with a high inductance.My personal fave' in that position is a big a*ss 5W Mills available from Sonicraft or Welborne Labs

 

[Tomo]

Wow wee,

Rickster said what I wanted to say about chip buffers.

I use the simplest of all discrete buffers. Szekeres amp. Yep. If you build it right and use it right, it sounds a thousand bucks. (especially with a lot of manufacturers started to make low-impedance phones)

Tomo

 

[twofishy]

Tomo, are you the same "Tomohiko Ishigami" that is listed in the project addendum for the Szekeres amp?

 

[Tomo]

Hey,

Yes. I am that person last time I checked. I am a bit more geeky now, though.

Tomo

 

[twofishy]

Wow you built that amp back in 2001...does it still work?

 

[twofishy]

Quote:

Originally Posted by rickcr42 /img/forum/go_quote.gif 1-Straight LH0033 in the ceramic DIP package @ 70mA Class-A 2-Straight LH0063 in the TO-5 package @ 150mA Class-A 3-JFET White follower/LT1010 in the TO-220 package @ 70mA Class-A 4-JFET White follower/SE MOSFET in TO-220 package @ 120mA Class-A

Hrmn, can't find any of these Linear Technology chips from Digi-key/Mouser...where did you get these from?

 

[Tomo]

Hey,

I think LH00XX aren't made anymore ... But you can make LH0002 youself with cheap transistors. (I think)

Yep twofishy. That thing is still alive in a way. I smited the amp with my heavenly fist and recycled the parts to make a better one. Here is the picture.

The chasis looks far more sexier, but the transistors are the same IRF510s I used on that amp you see in the addendum. ... Hard to believe, but these things last far longer than a lifespan.

Cheers to the olden times, and the coming glory. Heehee.

Tomo

 

[rickcr42]

Quote:

Hrmn, can't find any of these Linear Technology chips from Digi-key/Mouser...where did you get these from?

digikey

http://www.digikey.com/scripts/DkSea...S&Cat=32965194

Quote:

I think LH00XX aren't made anymore ... But you can make LH0002 youself with cheap transistors. (I think)

LH002,LH0033,LH0063-all simple toplogies but I think would take a bit to get a "clone" to work exactly the same.The LT1010 is pretty straight forward too if you disregard all the current limiting and complicated biasing to the left side of the schematic.A a simple input stage,simple output stage

Quote:

Hard to believe, but these things last far longer than a lifespan

mine were soldered to boards,removed and bolted directly to panel mounted heat sinks with the passive parts soldered directly to the pins,removed and put on different boards,back to the panel......

You're right man,tough b*stards.

Quote:

Cheers to the olden times, and the coming glory. Heehee.

you got that right.The early pioneers of audio had it mostly right and were mainly limited by the materials available then NOT the circuit designs but things being what they are "progress" usually means different not better so the next 50+ years was spent screwing things up in many cases (something still popular

) while the Ultra-High end is going backwards,rediscovering the obscure or applying old designs to modern deices.Even the best solid state usually comes from studying old tube technology and in fact it took me treating my Szekeres Class-A Mosfet Amp like a Class-A Pentode Amp rather than a solid state amp to finally get to the point where I was comfortable bolting the lid on and leaving it there

 

[rickcr42]

More........

Look at the first schematic,the jfet follower

http://www.nalanda.nitc.ac.in/indust...semi/AN-48.pdf

damn near identical to the "suggested" buffer in the Welborne Mods Catalog.


now the LH002 that preceded the 0033

http://cn.100y.com.tw/pdf_file/LH0002.PDF

if you look the text right above and to the left of Fig.5 it reads how the bandwidth is tied in with the source resistance and how it should be kept to within 50 ohms to 300 ohms.More evidence supporting the theory that bipolar transistor input buffers need to be driven from a low impedance source and why you usually see the data sheets recommend an op-amp at the input but also be aware data sheets are written to the lowest common denominator so will always suggest the CHEAPEST method to get great test results and not the BEST method for top performance when the "test" is the human ear and music the tet signal.No steady state test tone,even a warble,can approximate the moment to moment changes and dynamic contrast of music so while you can design for a very pretty graphic it may have zip to do with how the final design will sound and why a smart designer uses bench tresting to "see" if there are any obvious problems needing to be addressed but will "listen" to get the final design nailed down (good place to insert the heafdi mantra of "looks good I bet it sounds good" wise crack

)

if you look at the LT1010 schematic at the bottom of this file

http://www.linear.com/pc/downloadDoc...28,P1219,D1683

it looks damn complicated until you REALLY look at it then realise the entire buffer stage has 7 actual active devices and surprise ! Is actually a single ended design.The problem is the input-Q16,a bipolar transistor where a jfet follower like in the LH0033 would have been nice because here too it needs to "see" a low Z source at the input or bandwidth suffers so what is the answer to using the chip (and many others with like inputs) ?

Add the damn source follower to the input and call it a day !


BTW-works real well with bipolar input op-amps too

 

[twofishy]

Quote:

Originally Posted by Tomo /img/forum/go_quote.gif Yep twofishy. That thing is still alive in a way. I smited the amp with my heavenly fist and recycled the parts to make a better one. Here is the picture.

Ah, that blue led behind the volume control is a good idea...looks nice too