Rerurn to Romy the Cat's Site

Welll, you still did not express yourself what are the sonic verdure of DHT from your point of view. You are not obligated I juts proposed you to do it if you feel that have any observations on the subject.

guy sergeant wrote:

I started by supplying DC via a simple bridge circuit using fast recovery diodes and capacitance which significantly reduced the noise but which also flattened the pallete of colours the amplifier seemed to have at its disposal. Where it had been eager and colourful it became much less engaging. I experimented with various regulators but still felt that this 'washed out' character persisted. It may or may not be because of the fact that the passive components used in heating the cathode are in the signal path.

I personally very much NOT in a position to express an expert or reference opinion about electronics but I have my own undeniable experience that in my mind (and for me only) overrides anything that others say. I experimented with the very same issues you described during the times when I built my 7788-7721 phonocorrector and where I used DC on filaments. I made-many power suppliers for filaments and spend quite a time to listening them.

Tube rectification was the best form a perspective of colors but I never was able to get necessary bass and necessary “rhythm” form tube rectification. I was trying 3-4 different rectifiers and they were all the same, thought there was some difference between them.

The SS diodes driving caps very good bass but the devises that those power supplies drive practically unlistenable tonally. At low current they do sound fine but at higher currents for whatever reason this setting is absolutely disastrous. I found a compromise by lowering the first cap after the bridge with increase of current (of course there is a resistor between the firs and second cap).  The ultra fast but soft switching diodes are better diodes but in this application they do not sound good as well. In order to get any more or less civilized sound at 1A-2A I was forced to drop the first resistor to nF level.

The salvation come when I put a ultra fast but soft switching diodes with input choke. It was like an instant bliss- it had a full bass, full “rhythm” and dynamic but no deliration effects. Eventually I learned that the firs cap after indictor should be small and the first LC chain should form some kind of mimic of a reversed Bessel filter.  Sure the choke should have the necessary minimum inductance to the given current but it very easy to get. Interestingly that this first LC does not necessary kills a lot of ripples, in fact it will be a lot of ripples after it. However, it is ironic but ripples actually do not affect tone. I was driving a lot of equipment with juts a simple LC filter (with very small cap) and I loved how it sounded. Sure it has nose but even there thought the noise it was clearly auditable that it was very problem-free supply from a perspective of musicality. So, I personally do not recognize the input LC chain as a “filter” – it hardly filters anything but rather my “default entry point” form power supply to the rest of the circuit. (in case on constant current of course) From here everything is very simple: one more RC chain with the cap as big as your finance, space, ego or stupidity allows you.

This is what I learned worked for me and since I ma very restricted in my electronic diversification I use the same pattern in PS everywhere. If I make a PS for DHT it would unquestionably go for fast dies with LCRC, probably for a fully balanced LCRC with the last cap of enormous values. What until you guys see the PS for the upcoming Zaratustra II amps – it will be cartoonish…. But still all around the very same idea…

Rerurn to Romy the Cat's Site