A New Take on the Classic Pass Labs D1 with an ESS Dac

[Russ White]

Originally Posted by opc

Hi Guys,
Russ:

1) Could you post some measurements for the legato? I've checked your website and I don't see anything there. If you've got access to an AP I'd prefer that.

2) think the interest in this particular circuit is that it's a simple, single stage/single device design with no feedback and THD performance that is so good it's unmeasurable with state of the art measurement equipment. Can you think of a reason it wouldn't be interesting? Power dissipation is the only drawback, but at 42 watts, it's a small fraction of what even a small class A amp would draw.

3) The legato is a two stage design with 4 active devices per branch and feedback. I'm sure it works quite well, but it's a fundamentally different circuit. I'd also be surprised if the Legato's measured THD+N was better than -112dB (measurement floor of the AP).

4) I like simplicity, and I like the idea of minimizing component count and circuit complexity. There's nothing special here with this circuit, I just took it a little farther than others have, and took the time to tweak the individual parts both with listening and measurements. I've posted all my results here for everyone to see, so there's no black magic or mystical claims. It's just a really good/cheap/simple circuit that performs incredibly well at the cost of a little bit of heat.

Cheers,
Owen

1) No I don't currently have access to an AP (I have in the past, and will in the future), but I will certainly measure it more thoroughly at some point. Still that said I have measured it with what I have and it's distortion is below the measurement floor my personal gear. I have posted the simulation FFT which is much better than a similar simulation with just FET in two ways. First input impedance is much lower. If you would like to measure a Legato with your AP I can send one to you.

2) I don't disagree that the D1 is both an excellent and simple circuit. It is both.

3) This is only partly true. It is only a two stage design if you use the buffer. There is no global feedback, the only "feedback" if you will is the CFP. I can't see any reason to find that objectionable. It is also most definitely *NOT* a fundamentally different circuit. That sounds like a mystical claim to me... It is a common base amplifier. You should not be surprised if the Legato measures as well or better then 112db. It certainly simulates much better.

4) You have eliminated much of the practical simplicity by creating huge heat sinking and mounting requirements. But I agree, it is an electronically simple circuit. Nobody is interested mystical claims I am sure.

So now that I have answered you let me be clear. I think your doing a good job. I think your work is great. I am encouraging you.

I am really only commenting that if people really want it I can resurrect my working "D1" PCB which I had already tested. It is designed to optionally mount to a heat sink(Not required).

Cheers!
Russ

[qusp]

Russ, i'm sure you meant well mate, but coming in here and basically advertising legato and an old D1 like circuit wasnt really that well thought through IMO and I think opc handled that admirably. I mean this isnt a product for profit, but all the same, the more people that buy the PCB in the GB, the cheaper it is for us; so coming in and name dropping a PCB in a development thread when you have a paid for thread of your own and already a high profile raises some questions WRT motives

[Russ White]

Originally Posted by qusp

Russ, i'm sure you meant well mate, but coming in here and basically advertising legato and an old D1 like circuit wasnt really that well thought through IMO and I think opc handled that admirably. I mean this isnt a product for profit, but all the same, the more people that buy the PCB in the GB, the cheaper it is for us; so coming in and name dropping a PCB in a development thread when you have a paid for thread of your own and already a high profile raises some questions WRT motives

What motives do you think I have? I am not trying to sell the PCB I was offering it up (the gerbers) for FREE. I was only offering to share the layout that worked well for me.

I only mentioned the Legato in passing, because it was my work playing with the "D1" type circuit that led to it. See really I was just contrasting two ways to attack a problem. The problem being low transconductance of FETs.

I solved the problem with a CFP. This allowed me to keep the power requirments modest. OCP solved the problem by pumping tons of current through the FETs. Neither is wrong. They both are good solutions but with unique challenges.

Peace.

Cheers!
Russ

[qusp]

haha well OK like I said, raises some questions. sall good i'm not making any accusation, just letting you know how it could and kinda did look from this side

[Russ White]

I am not sure why you think I should not communicate my thoughts on the subject. I think I am clear when I say I want this project to succeed. I am only contributing to the conversation.

[qusp]

Russ, mate you did indeed offset this with some good information and well meaning praise which is why I framed it how I did. without that it would have looked a blatant ad. dropping your own products names 'legato' and 'D1' and then boasting higher performance after moving on from a circuit like this; saying how surprised you are at how popular such a circuit is; all in a development thread that is getting a bit of attention and you really cant see how this could look? I really dont think you meant it to, which is why I made sure to start with that statement.

[Russ White]

I find that often the way you look entirely depends on perspective of the looker. That is something I can do little about. My personality is such that even knowing such perspectives exist that might view it badly wouldn't be likely to constrain me in any case.

In the end, I only wish the best.

Cheers!
Russ

[opc]

Hi Guys,

Back on topic please...

I'm almost ready to get the routed version posted, so if everyone can give it a look over once it's up, I'd appreciate it.

Looks like I'll be ordering 30 boards to start, and I'm going to guess it'll be $20 per PCB but I'll have a better idea when I know how much shipping will be and all the little add on charges that happen when you ship something from the US to Canada.

Anyone with suggestions on how to manage the board sales? Something like a sign-up list to help me keep track?

Russ:

I'd be glad to measure a Legato if you'd like to send one. I already have a Buffalo II I could try it with, and if you have any special power supply requirements then you'll have to provide those as well. We have excellent lab supplies here though, if you're alright with that.

I have access to both a brand new AP 585 as well as an older but fully equipped System 2.

How much current were you running through the single FET circuit you were using? When I first started out I had small SMD FETs and little 1W SMD resistors, but the circuit outgrew them pretty quickly. You can still get pretty respectable performance with minimal power dissipation, but it really is a case of more is better with this circuit.

Back to routing!

Owen

[ronu9400]

Hi, Owen

I apologize for hijack, but I'll make it quick. I've been trying to reach you via PM & follow-up post(s) after our Ottawa-area DIY meet, but doesn't seem to have worked...maybe I'm on probation as a new member?

Anyway, I won the PCB's from you at that meet & research suggests that this is for your Moskido layout/build. If possible, I would like to get the PDF (schematic & parts list) you mentioned and offered to make available - I expect it is beyond my current abilities, but maybe I can start gathering parts & experience.

Thanks & again, my apologies
Ron (fyi, you can e-mail to: suzaneurquhart at rogers.com)

[opc]

Hi Ron,

I'll get those sent off to you this evening. I might have a few parts for it as well that you could have. Not much but enough to get you started.

Cheers,
Owen