Who tried Class D only to return to S/S or Tube



And what were the reason you did a backflip back to S/S or tube.
As there are a few pro Class D threads being hammered at the moment, I thought I'd put this up, to get some perspective.

Cheers George
128x128georgehifi

Showing 7 responses by ait

I owned Gilmore Raven stereo amp for a few years, then upgraded to the Raptor monoblocks. Was happy with them for several years as well.

Then I built a pair of two stage, direct-coupled SET monoblocks, with 6E5P driver and 833C power tubes, producing 200wpc. These are so much better from bass through treble that I was stunned. More real-sounding, more musical, and more authoritative. They make me want to listen to music as often as possible. I'll not go back.
Oh, and prior to the Gilmore Class D amps I owned Flying Mole Class D monos. The Gilmores were better than those, but the FMs were still nice sounding. I also auditioned NuForce Ref9 SE monos and thought they sounded awful - harsh and screeching are the descriptors that come to mind.

I don't see myself with anything but my 833 SET amps from now until I will them to my son.
Kijanki, in your zeal for Class D I'm afraid you mischaracterize linear power supplies. Not all are cap input.

Choke input linear supplies do NOT draw power in "huge narrow spikes". In fact they draw power continuously throughout the conduction cycle, resulting in smooth, fast and clean power delivery. No "huge amount of capacitors" are needed since choke input supplies have very good inherent regulation.
I use choke input in everything I build, including amps, preamps, DACs, even for DC IDHT tube heaters. Best feature though is no RF nasties flying around the room, which are inevitable with high frequency switchers.
I showed the current waveforms through the first PS Capacitor in each case. It's clear that in one case there are sharp spikes through the cap, in the other there are not. 
The rectifier is not directly charging the cap, it's charging the inductor, which stores the energy in its magnetic field, then the inductor output charges the cap. Filter topology is LCLC.

Kijanki,
I believe you misunderstand what a critical inductance choke input supply is. The input choke serves to smooth the current draw - I have attached two simulations of the power supply I use in my amps showing the current draw in the first power supply capacitor. The left half of the picture is the supply as-is (with choke input); the right half is the same supply set up as cap input.

http://i1243.photobucket.com/albums/gg559/Quagmire22/Choke%20vs%20Cap%20Input%20Supply_zpsnywzj0ww.p...

Notice that with choke input the draw is a smooth sinusoid. With cap input the draw is as you describe - sudden sharp pulses. The energy storage of the input choke smooths the current draw and also contributes to the very good regulation of the choke input supply.


Your link shows a cap input supply, by the way. Put a choke between the rectifier and cap and you get a choke input supply.
Kijanki,

The choke needs to be of "critical inductance" at minimum, for this type of supply to work properly.

Lcrit = V(dc) / I(mA)

So higher voltage and/or lower current require a bigger choke. This type of supply was often used decades ago when capacitance was expensive compared to chokes; once capacitance became cheaper it fell out of favor in the mass market, but it retains its unique advantages to this day.

Happy reading!