True Digital Audio Amplifier

[langtuhanoi]

Re: PowerDAC

Hi John Hope,
Your PowerDac3 is intergrated digital amplifier with digital input such as I2S and SPDIF or DAC?
Thank

Hi wytc0:

Sorry I didn't see your previous posts because I unsubscribed from the thread, which seemed to be losing focus.

I have already posted some of the schematics of the existing PowerDAC 1 and 2, which I don't have any problem with sharing to the forum.

I am now developing PowerDAC3 - inbetween trying to work and live up to the high expectations of my cats - and there has been commercial interest shown in it by an established UK company. It wouldn't be appropriate for me to post details of the new developments at this time.

Kind regards

John Hope

 

[John Hope]

No, the present version requires external signal processing from AES to 8-bit PWM values, ie interpolation, PNPWM and noise shaping. The previous versions, PowerDAC1 and PowerDAC2, had these functions in a separate processing module.

I intend to incorporate these functions into PowerDAC3 during this year - and improve them by using Floating Point processing - so that by the end of the year (I hope) the PowerDAC 3 will be AES (or I2S) in and speaker out.

 

[ebdrup]

Where can one buy a John Westlake True Digital Amplifier?

Where can one buy a John Westlake True Digital Amplifier?
I'm looking for a Resonably priced True Digital Amplifer (Class D)for my squeezebox:
http://www.slimdevices.com/pi_duet.html
I need somewhere between 60 and 100W
I want it to be class D because of the low power consumption and because it will be on all the time, and the perfect amp shuts down automatically when there is no digital input for a while, and starts up automatically when there i s input again.

Kind Regards,
Allan Ebdrup

 

[Wingfeather]

Hi John,

Are all your PowerDACs based on PNPWM? I've been simulating this in MATLAB recently (along with the other PWM linearisation schemes that are around, for comparison), and found that it suffers from the effect predicted by Peter Craven in his 1993 paper:

However, an effect which has so far been ignored is the nonlinear treatment of the quantizer noise [in the PWM modulator]. As is well known, a heavily noise-shaped quantizer is always "busy," this "busyness" producing a peak displacement of the pulse edge of typically ±2bit clocks. Referring to the conventional arrangement of Fig. 4, this wideband noise at the output of the noise shaper will intermodulate with itself in the nonlinear PWM, and some of the intermodulation products will fall back into the audio band.

I've found that this manifests as extra noise power at LF, which makes the shape of the noise-floor of the output signal deviate significantly from the shape of the noise-floor of the (noise-shaped) input signal. As far as I can tell, only Peter Craven's nonlinear noise-shaper will actually compensate this effect. No feedforward system (e.g. PNPWM, dynamic FIR, etc) will ever be able to.

Can you comment on this?

 

[jkeny]

hi

the Panasonic receiver i own ( sa-xr45 ) has tas5036 and 5182 on board , h-bridge , no feedback loop and voltage control-based volume
the Tact 2150 too , i think
The sound is amazing but not natural

alain

I'm modding a Panasonic SA-XR57 based on TAS5076 & 5182 TI chips and brought it to a recent audio meet where it was easily trounced by a tube amp - admittedly it was a €6000 Air Tight integrated amp but the naturalness of this amp stunned me!

I was running my output stage on a 12V 7aH battery which gave ample volume in my room (which isn't small) but in the open plan area where we were listening to all the equipment, I had to turn the volume up to -26 to satisfy the punters present

Anyway will be returning shortly, I hope, with another two batterries hooked up to it, although I don't expect it to sound as good as the Air Tight.

(it did get "best bang for buck" award but this isn't the criteria I'm interested in)