New Doug Self pre-amp design...

[owdeo]

Originally Posted by sandyK

owdeo
You could always look a little closer to home, for designs by David Tilbrook from Au. who designed many well received projects such as the AEM 6000,and ETI 5000 PAs, as well as some other designs such as the Ultra Fidelity preamp. One of those from close to 30 years ago also used NE5534s !
DIYAudio member, and C.S.I.R.O engineer Suzy J even got to update one of those amplifiers using modern surface mount components, and was even ripped off by an ebhay clone of her hard work. Perhaps you even got to meet David Tilbrook in person at Jaycar's Gore Hill store many years ago ?
Of course, that was back in the days when you had an interest in high quality audio reproduction, and studied to become an E.E.

SandyK

I thought you knew I had an Ultra Fidelity preamp! It's one of the preamps I'm comparing the D.S. design to subjectively. I had to change the compensation to make it stable though...

I have an old ETI5000 courtesy of Jaycar ex-demo. After I rewired it properly and replaced all the dodgy components it works well and sounds pretty good to me - certainly much better than the "Pro Series 3" that EA published that was supposed to replace it (I built one of those at the time while I was at uni and sold it immmediately afterwards).

I've heard the AEM 6000 years ago in a friend's very expensive system with Duntech speakers and an air-bearing turntable. It sounded wonderful. I therefore have collected the parts to build one over the years and they are still sitting waiting for me to finish my PCB design (I've recently discovered that Protel 99SE doesn't work properly on Vista or Win 7 as I can't import my libraries...aaarrrrgh!). I've seen Suzy's excellent work, but wanted to keep mine through-hole to use the parts I already have and also at the higher power rating.

I'm quite happy with the Leach amp anyway - it's a good preamp that I really need now.

Those were the days at Jaycar Gore Hill - we had lots of fun back then. I never met David Tillbrook but sure wish I had. From his articles and the sound of his designs he seemed to be just as willing to listen as to measure...

[sandyK]

I thought you knew I had an Ultra Fidelity preamp! It's one of the preamps I'm comparing the D.S. design to subjectively. I had to change the compensation to make it stable though...

Hi Owdeo
I am well aware of that.
My reply was a little tongue in cheek, in order to make the point that you aren't exactly inexperienced in these areas, and not afraid to correct even Silicon Chip magazine as you did recently , with a a correction published recently in their ERRATA section .
Kind Regards
SandyK

[Kindhornman]

Owdeo,
I do understand what you are saying. Like you I started in a different part of this sound field. I started very young working with live sound and that was a learning experience. I also have had many friends over the years who are from the Jazz field and I have spent many a time sitting and listening to rehearsals in peoples homes and studios. I played a little guitar years ago but didn't keep that up.
I know of others who work in design and they also suspect that the feedback loop has more going on than can be described with traditional testing methods. When you think about what is happening it makes some sense. Though we think that the speed of an electron traveling is at an almost infinite speed there is still the fact that it has to travel in a loop and be superimposed over the original signal. There just seems to be something in physics that says you can't be in two places simultaneously..... So if this is some kind of phase shifting or let's just call it a Doppler effect something is beyond what we are testing for. If we had a perfect world there wouldn't need to be a feedback loop. The back emf wave from any loudspeaker must be interacting with this is some way and driving this servo loop in an uncontrollable fashion.

[AndrewT]

Have you measured the Pre-Amp to check that it meets the target specification?

[SY]

Originally Posted by Kindhornman

I know of others who work in design and they also suspect that the feedback loop has more going on than can be described with traditional testing methods. When you think about what is happening it makes some sense. Though we think that the speed of an electron traveling is at an almost infinite speed there is still the fact that it has to travel in a loop and be superimposed over the original signal. There just seems to be something in physics that says you can't be in two places simultaneously..... So if this is some kind of phase shifting or let's just call it a Doppler effect something is beyond what we are testing for. If we had a perfect world there wouldn't need to be a feedback loop. The back emf wave from any loudspeaker must be interacting with this is some way and driving this servo loop in an uncontrollable fashion.

Some basic misconceptions never die, do they? I think you have some basic reading to do on control theory and feedback, not to mention basic physics. In another thread, I had just recommended Bruno Putzy's excellent article on feedback that appeared in Linear Audio, and I'll recommend it again- it will clear out much of the nonsense you've been fed.

[Bonsai]

I am afraid feedback does not work like that. Even if it did, the c. 1ns transit time around an amp circuit would hardly matter at audio would it?

I am hardly in a position to challenge DS but I would suspect that in this design, the fact that the opamp stage driving the active volume control is operating in class B ( i.e very small class AB region) at low signal levels may have something to do with it. However, the contribution is very small as evidenced by the low measured distortion, and this is only a wild guess. And, this assumes the implementation is free of any oscillation or other anomalies.

[Bonsai]

Owedo, one thing you could try is to load the stage driving he active gain control to one of the supply rails with about 5-10mA. This would shift the cross over point away from 0. Might address your sound issue - worth a try.

[Kindhornman]

Sy,
I take no offense to you pointing me in the right direction. I'll go read that information and hope I can find it. A link to the article itself would be nice. I'm here to learn and no expert on any of this side. Just a newbie with much to learn and looking everywhere to find the information. It isn't always easy to separate the truth from fiction, opinions abound. I'll try and keep my foot out of my mouth.

[Kindhornman]

Sy,
A follow on to my last post. I went to the Linear Audio site and registered for it. So this is a publication to purchase, a technical self published journal in effect. Is there somewhere that I can go to read this article without having to buy an entire volume of papers? I'll look around on this but I assume some of this is in D. Self's design books and also Cordell. Is this paper more theoretical in nature rather than application specific. Thanks in advance.

[nereis]

Originally Posted by Bonsai

Owedo, one thing you could try is to load the stage driving he active gain control to one of the supply rails with about 5-10mA. This would shift the cross over point away from 0. Might address your sound issue - worth a try.

Do this with a resistor to the positive rail and not the negative rail, or use an active current source. I recall reading that the ne553x output stage is better at sinking current but not sourcing it.