Marsh headphone amp from Linear Audio

[sonidos]

Originally Posted by janneman

Not that it did help any

Umm..no, sure didn't. Couldn't pass up those bargains just across the border!

To RNMarsh, after we all get our hands on building the head amp, I'll definitely put up some feedback on this thread. I have a pair of Senn HD580s that I hope kick some *** with this circuit.

[RNMarsh]

Originally Posted by sonidos

Umm..no, sure didn't. Couldn't pass up those bargains just across the border!

To RNMarsh, after we all get our hands on building the head amp, I'll definitely put up some feedback on this thread. I have a pair of Senn HD580s that I hope kick some *** with this circuit.

Let the fun begin !!

[dirkwright]

Originally Posted by coluke

Guys, we're talking about an open-loop small-power buffer capable of 0.001% THD into 30 ohm - it seems there's quite a lot of good engineering behind these figures...

L.

Why would anyone leave a buffer hanging outside the feedback loop? That makes no sense to me.

[simon7000]

Originally Posted by janneman

Thank you!...


What I find interesting is that people are willing to shell out $ 100 for a boutique cap on a simple "I did, and it's great, sound stage has never been as good"- type post, but have issues with getting a printed book for $ 28 that REALLY can advance them in understanding what it is they are doing. ...

jan didden
editor/publisher
Linear Audio

Jan,

You have made an assumption that is clearly not the case. There are several cases you might want to consider:

1. There really are folks here who spend almost no money as their budgets are limited.

2. It just might be possible there are Ego's hanging out here that are larger than mine and so don't recognize the advantage of reading others contributions.

3. There are those who as long as the milk is free, won't buy a cow. So what if the milk is a bit sour (Or the content is not at as high a standard.)

4. Then there are folks like me. I won't buy any new printed content when I don't have time or room for it. I'll buy material when things slow down! I still need to build 6 to 10 more bookshelves!

5. Then there those who are here to primarily show off. You for some strange reason assume that everyone who writes they have a Royce really does. Painting RR on the hood of a Yugo just isn't the same.

6. Finally there are some who just want to build stuff and there are lots of free projects. They don't get that some of the designs are really not very good.

Oh I just re-read #2, My mistake there can't be Egos larger than mine!

ES

Now as proof just look at some of the comments on the circuit discussed here by folks who have not seen the article, don't know the author's history and generally don't have a ticket to the clue train. (Currently found in ant issue of L.A.)

[jan.didden]

Sure Ed, all good points.

I didn't per se literally mean chosing a $ 100 cap for an LA volume, rather I wanted to contrast the easy spending on parts without iron clad reasons, versus spending something on increasing your understanding.

I'm sure you'd agree on that in principle.
Now if you'll excuse me, I must go paint 'Lotus' on my Chevy Cruze...

jan

[Joachim Gerhard]

As far as i can see it the buffer has 100% feedback in a local loop.

[simon7000]

Originally Posted by janneman

Sure Ed, all good points.

I didn't per se literally mean chosing a $ 100 cap for an LA volume, rather I wanted to contrast the easy spending on parts without iron clad reasons, versus spending something on increasing your understanding.

I'm sure you'd agree on that in principle.
Now if you'll excuse me, I must go paint 'Lotus' on my Chevy Cruze...

jan

Just leave the hood up and parts scattered around. It will look just like the real thing!

[dirkwright]

Originally Posted by coluke

Guys, we're talking about an open-loop small-power buffer capable of 0.001% THD into 30 ohm - it seems there's quite a lot of good engineering behind these figures...

L.

So, is 0.0007% good enough? Four transistors and two diodes (which may count as transistors since they are CRD's). Since there is no loop feedback to reduce the output impedance, then you really have to juice it up to get a low enough impedance to drive 30 ohms at low distortion. Output level is 731mV rms. At 1V rms output, distortion goes up to 0.001%. This is across 30 ohms of course, as you wanted. These are Mr. Cordell's transistor spice models, btw.

[simon7000]

Originally Posted by dirkwright

So, is 0.0007% good enough? Four transistors and two diodes (which may count as transistors since they are CRD's). Since there is no loop feedback to reduce the output impedance, then you really have to juice it up to get a low enough impedance to drive 30 ohms at low distortion. Output level is 731mV rms. At 1V rms output, distortion goes up to 0.001%. This is across 30 ohms of course, as you wanted. These are Mr. Cordell's transistor spice models, btw.

Now run a variation of parameters.

[dirkwright]

Originally Posted by simon7000

Now run a variation of parameters.

Nah, you can do that. Go ahead. You're not my boss.