Marsh headphone amp from Linear Audio

His Msc Thesis? Or one of his web pages?

I haven't looked in a while, I have the complete Ms thesis and I don't know if he posts it. Bruce Hoffer is acknowledged at the start and Samuel claims 20dB improvement over all previous equipment (though I don't think he considered the CLT1). The second part is like a service manual and I assume no secret sauce is left out and every detail is there relay boards, power supplies, etc., not a weekend project though.
 
Devil's in the details. RM's circuit runs in class A, the figure you reference doesn't. IMO, class A is no big deal, but I can't argue with RM's results, the measurements Jack showed were outstanding.

I've got no dog in this hunt, I hardly ever use headphones, but the article was educational, and it's interesting to understand the thought process that an experienced and accomplished designer uses. No doubt others are far beyond my abilities and this is all old hat to them, but I'm just a dumb chemist.

OK Sy, no problem. The addition of a couple of resistors to Fig 3a is all it would take I guess to make it the same. It's an interesting idea. I suppose I can explore it more later. Do you know what this particular output stage is called?
 
Disabled Account
Joined 2012
This is almost interesting to read the fall out of the article on a headphone amp. For those who wont read the article, the derivation from historical times has been included and my own role. There is no claim to be a new circuit topology, either. Instead of rewording what I have said in the article I have to leave it at that.

There is no money involved. I took no writing fee, either. There are some good info there or I wouldnt have bothered. No one has yet talked about the auto EQ of the ear/headphone method and results. Thats a totally new app.

My comments on the amplifier circuit (not what was shown here) are: Make one that performs better in all this amps measured areas, use fewer transistors and be cheaper to make. Not one spec or area but All. I spent over $2000 on discrete buffer circuits, some published by others, and line level amp circuits and pcb's etc. I bought and tried all sorts of transistors but stopped and moved on when the results were so good. It is the transistor type/make and the currents chosen to run them and the topology makes it so you can drop in a part without testing them and get >-100 db thd. And cost what a low level IC costs.

Anyway, I will be happy to discuss anything about it. But, I am especially interested in you guys building automated EQ for headphones as I showed in the same article.
 
Last edited:
Member
Joined 2005
Paid Member
Richard,

I built your double chamber bass reflex cabinet with dual Dynaudio 24w100 in each and they served me well for a decade as subwoofers for Quad 63s. Built just the line amp of your 1980 preamp article as the AD input devices in the phono stage were hard to come by. I replaced most of the ceramic and tantalum coupling cap in my electronics after the Audio article on capacitors. I added the input stage cascode and DC servos to Hafler and early Krell power amps from the AA Pooge 3.
Thanks for the articles then and now.
 
Last edited:
Disabled Account
Joined 2012
Richard,

I built your double chamber bass reflex cabinet with dual Dynaudio 24w100 in each and they served me well for a decade as subwoofers for Quad 63s. Built just the line amp of your 1980 preamp article as the AD input devices in the phono stage were hard to come by. I replaced most of the ceramic and tantalum coupling cap in my electronics after the Audio article on capacitors. I added the input stage cascode and DC servos to Hafler and early Krell power amps from the AA Pooge 3.
Thanks for the articles then and now.

Wow! You've been a busy guy and for a long, long time. Nice to get a long range perspective. Thx. -RM
 
Hi Richard, what about its output impedance? I'm working on an open-loop hp buffer/amplifier - actually a standard CF open-loop topology working very well indeed. My tests show that in order to avoid a dramatic THD raise at lf the output impedance must be kept extremely low. In my first prototype it was about 2.5 ohm, but now I'm managing to reduce it further to 0.1 ohm or so. Did you manage to measure THD on low Z (~ 30 ohm) * real * load?

L.
 
.....For those who wont read the article....There is no money involved.......

Those who won't pay to read the article would be more accurate.

The fact that you took no fee does not mean that there is no money involved in reading it!

I'm sorry to be pointing this out to you. I know that you do a great deal for this forum; if you like, you're one of the good guys.

But despite its non-profit (which I can well believe) nature, Linear Audio is a commercial concern, albeit a small one. Non-profit does not mean non-commercial - I don't know how things are in the States, but not-for-profit limited companies are very common here in Britain.

The rules of the forum state that:

The regular forums are non-commercial and should have no advertisements or overtly commercial threads or posts

The OP's initial post was, whether intentionally or otherwise, a teaser for a Linear Audio article, the schematic for which will not be posted here. It was an advertisement.

Rules that only apply to some, no matter how good or noble the some concerned, are never a good thing.

I strongly suggest that Linear Audio, by all accounts a deliberately better than average publication in this field be given its own section in the forum. I think this would benefit everyone.

Finally, I'd like to apologise to those whom I have undeservedly insulted or offended.

 
I have built R.Cordell's distortion magnifier plus I use AP.

Me too, results? In the end I find the source nulling techniques to far exceeed the capabilities of brute force THD measurement. Somehow these techniques were denigrated by folks that used Hafler's version of them and did not find corellation to listening tests. Anecdotal, arbitrary, and unscientific.
 
Disabled Account
Joined 2012
Hi Richard, what about its output impedance? I'm working on an open-loop hp buffer/amplifier - actually a standard CF open-loop topology working very well indeed. My tests show that in order to avoid a dramatic THD raise at lf the output impedance must be kept extremely low. In my first prototype it was about 2.5 ohm, but now I'm managing to reduce it further to 0.1 ohm or so. Did you manage to measure THD on low Z (~ 30 ohm) * real * load?

L.

The thd in the article was quoted as into 30 Ohm load and it was well better than .001%. Higher load Z would give lower thd. A search on the Internet of headphone Z showed many were below what IC's could handle well... that is why I went for 30 Ohms for design goal.
 
Disabled Account
Joined 2012

Those who won't pay to read the article would be more accurate.

The fact that you took no fee does not mean that there is no money involved in reading it!

I'm sorry to be pointing this out to you. I know that you do a great deal for this forum; if you like, you're one of the good guys.

But despite its non-profit (which I can well believe) nature, Linear Audio is a commercial concern, albeit a small one. Non-profit does not mean non-commercial - I don't know how things are in the States, but not-for-profit limited companies are very common here in Britain.

The rules of the forum state that:


The OP's initial post was, whether intentionally or otherwise, a teaser for a Linear Audio article, the schematic for which will not be posted here. It was an advertisement.

Rules that only apply to some, no matter how good or noble the some concerned, are never a good thing.

I strongly suggest that Linear Audio, by all accounts a deliberately better than average publication in this field be given its own section in the forum. I think this would benefit everyone.

Finally, I'd like to apologise to those whom I have undeservedly insulted or offended.

yes, it will cost you money and yes, the magazine might make a profit. I happen to enjoy the higher level of technology discussion there and hope others would also. I try to teach a little as I go and give away what I have learned from time to time. Just because WE love music and making/building it.