diyAudio (
-   Solid State (
-   -   Class A hybrid power dac project (

hbc 9th February 2011 01:09 AM

Class A hybrid power dac project

I am making a power dac, as part of my on going personal research in to sound quality. Purpose of, to achieve minimum of everything, signal path, active components, no feedback etc to try and hear the raw sound of "digital" so to speak.

2x tda1543
ecc88 as I/V converter
shunt volume control
class a(b) bipolar darlington output buffer with dc servo
about 10W

I have a prototype up and running and now sounding fairly together, am posting interesting observations for anyone interested:

first interesting shock was how much effect a normal electrolytic decoupling across my vbe multiplier had, changed to oscon much cleaner!

not sure if this has been done elsewhere, but inverted WS to one of the TDA1543 and the left right connections, this gives a sort of 2x oversampling with slight cost of 11uS delay, sounds happier like that.

all for now, hope to post more if any interest as i need to write up this project, but busy so not much time.


ontoaba 9th February 2011 02:20 AM

Is that powered DAC, or power DAC?

hbc 9th February 2011 06:36 AM


it is a power dac, it has SPIDIF input and drives speakers directly.

hbc 9th February 2011 08:12 AM

1 Attachment(s)

hbc 10th February 2011 01:02 PM

1 Attachment(s)
one Class A (b) biased output stage, about 300mA bias

ontoaba 10th February 2011 03:20 PM

1 Attachment(s)
But, I see that you amplify already analog signal. Isn't a power DAC like this one?(see my pict), My picture is Digital to (Analog)Power Converter. Sorry, I don't really know the differences.

hbc 10th February 2011 09:33 PM

Yes you are right, it is not that much of a power dac... It is about getting the (normal) dac as near to the output stage as possible with hopefully the least number of the most transparent audio stages i.e. a triode and a class A buffer. I know a bit about switch mode psu, but not sure I would personaly try that approach ! I would always be tempted to have all the current sources running in to the bottom of a cascode, which is effectivly what I have got in my circuit.


ontoaba 12th February 2011 02:37 AM

I ever make a power DAC once and yes only once using that time based switching, and only 8bit resolution 33KHz. It need at least 4 parallel switch to discharge inductor current to zero each cycle. It is become too complex, rejecting supply ripple with precise A*B circuitry, reducing Mosfet timing error, and many more.
Any way, what is your goal? more linearity or more power, from your circuit?

hbc 12th February 2011 11:29 AM

The most transparent digital playback...?

I know I am a noob to posting here, but I have been doing this stuff for 25+ years. and have a pretty good home made set up, that I can listen to almost anything (music) and it sounds good, anyway.

The Mrs has gone out and its daytime and I have been spending a few hours listening to this thing..... ****... yes it has a sound of its own, or maybe its my speakers, I need to test it on some other settings.

I need to make these and sell them to every record company, because they would never make a bad sounding CD again if they had to listen to it on

got my Isely brothers super-hits, its your thing, great intro, loved it in the car for a couple of minutes, great for a pulsating exit of petrol station moments, but then it went down hill fro there on, and normally got spat out after. I totally know why now, what a mess, where as pull out an AAD loyd cole big snake, sounds delicious, (my CD collection is a bit inaccessible due to moving house and having free roaming small destructive hands at work).

Attempted description of sound, if possible: all constituent sounds are free of each other, they start and stop independently in there own space. some familiar CDs Duende, a man called Adam, man what were they thinking, gratuitous compression I had never noticed before, totally obvious now, it is like the recording is de-constructed.

so it goes. I have "integrated it now, it no longer needs 3 separate power plugs, so time to move on to some evaluation.

Attempted some measurements last night, not really conclusive. My Sound Technology 1700 has broken (haven't used it for 10 years), sniff, I am hoping it is just the oscillator, as I don't really need that bit.

I couldn't get Paraxis to work, But ARTA works ok (W7 64bit), Have access to a prism sound set up at college, but I like to work at home. At any recommendations for free software I could use to check this thing out on.



I love the recorded sound time machine/ travel, shut your eyes and you are some where else, some other time effect. I like to know how things work, I don't like bulshit myths that transistor amps sound bad because of crossover distortion, why do SETs sound nice etc etc, I want to know the truth!!!

ontoaba 12th February 2011 03:43 PM

But, have you ever using transistor for your I/V converter instead of ecc88? and is that converter using current that flow into resistor to get voltage across it?

May be the truth is only one of a million right now.:)

All times are GMT. The time now is 06:33 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio