Code:
I used to worry about eliminating caps but less so now I've built various fantastic sounding projects that include them, Valve DAC being a case in point.Where can I find the "CD-optimiser article"?
Ths evening I finally found time to sit and have a proper listen to DSD512 vs DSD256, helped by finally paying to upgrade from HQP3 to HQP4 so that the music doesn't stop when you're into listening to something.
Although I am able to play DSD512 without drop-outs I noticed very quickly that when there are silences in the music there is a low-level static like noise present - when the music starts the noise isn't discernable, as if it gets turned off.
On the same music passages, at DSD256 the noise floor is all but inky silent to my old ears and the music is so beguiling I'll happily stick at 256.
I found this music to be particularly revealing of the static noise because of the breathing pauses and the passages being short. It's a album too, excellent playing and recording;
Although I am able to play DSD512 without drop-outs I noticed very quickly that when there are silences in the music there is a low-level static like noise present - when the music starts the noise isn't discernable, as if it gets turned off.
On the same music passages, at DSD256 the noise floor is all but inky silent to my old ears and the music is so beguiling I'll happily stick at 256.
I found this music to be particularly revealing of the static noise because of the breathing pauses and the passages being short. It's a album too, excellent playing and recording;
Attachments
Funnily enough I was just reading that Marcel - no sooner had I made the last post than a little voice told me to look at the documentation. I also tried DSD128 and having said DSD25 was inky, DSD128 is inkier! Maybe there's life in these old ears yet.
I'm getting very tired now so I won't fiddle as I know where that leads but I will try the trimming process and report back. Having been delighted thatIi actually made it well enough for it to work I've been happy to just listen and enjoy the undoubted qualities of this DAC but I think now is the time to put in a shift or two and squeeze it for everything its got.
I'm getting very tired now so I won't fiddle as I know where that leads but I will try the trimming process and report back. Having been delighted thatIi actually made it well enough for it to work I've been happy to just listen and enjoy the undoubted qualities of this DAC but I think now is the time to put in a shift or two and squeeze it for everything its got.
We’ve had different experiences with circuits, I suppose. My order of preferences are DC coupled>transformers>capacitors, so I gravitate towards stuff that I can do that with.
You could get rid of the DC blocking capacitors and anode resistors and connect the anodes of the upper valves straight to a centre-tapped transformer with not too high turns ratio, like a Sowter 3292, 3603 or 1465, with the centre tap grounded. You would still need filter capacitors from the anodes to ground.
You then can't trim the anode DC voltages for the lowest possible noise floor anymore. The muting circuit with the JFETs would probably still work. Trying to play back music while the DSD inputs are not connected will saturate the transformers and increase their second-order distortion until you demagnetize them.
If the centre tap should get disconnected, you get the full 300 V across the insulation of the transformer. Even if the insulation should break down, touching the outputs would not be lethal because the tail resistor of the bottom valves will limit the current to about 10 mA. In fact with two 402 ohm termination resistors from the secondary to ground, you would simply get a -2.01 V DC voltage at the output.
You then can't trim the anode DC voltages for the lowest possible noise floor anymore. The muting circuit with the JFETs would probably still work. Trying to play back music while the DSD inputs are not connected will saturate the transformers and increase their second-order distortion until you demagnetize them.
If the centre tap should get disconnected, you get the full 300 V across the insulation of the transformer. Even if the insulation should break down, touching the outputs would not be lethal because the tail resistor of the bottom valves will limit the current to about 10 mA. In fact with two 402 ohm termination resistors from the secondary to ground, you would simply get a -2.01 V DC voltage at the output.
Not really - I haven't really sat down and done any sort of rigorous or semi-rigorous comparison and I have no plans to do so. I mentioned the two DACs that I've most recently been using in the link in my previous post. All three are using HQPlayer and streaming to Beaglebone Black based renderers (HQP NAA) and have isolation/reclocking at their inputs. I also have a Soekris DAM1121 DAC but its used in a different context. What more are you asking for?
My assessment against the DACs I mentioned is purely based on my audio-memory of that recent use but my preference for Marcel's DAC was immediate and unequivocal.
My assessment against the DACs I mentioned is purely based on my audio-memory of that recent use but my preference for Marcel's DAC was immediate and unequivocal.
Some really concentrated listening and adjstment has bourn fruit this morning - I've been tweaking as per the documentation and I'm currently listening to DSD512 with a noise floor around what I was getting with DSD256 yesterday and DSD256 now has a really black background. It's quite tiring listening so hard for a couple of hours so I think I'll leave it for now, just enjoy some music and maybe see if i can wring a little more out of it another day.
A recording like the one I referenced last evening is really good for this tweaking because being a solo flautist playing short pieces with lots of breathing 'pauses' you can really pick out the noise floor changing.
The process reminded me a little of the old days of tweaking cartridges for geometry and tracking, though turning potentiometers is a lot easier to execute.
Here's where we are with the GB;
diyAchim - 1 PCB
Raj!1 - 1 PCB (UK)
dubulup - 2 PCB (USA)
ericoto -1 PCB
HenSch - 2 PCB (Germany)
mterbekke - 2 PCB (Netherlands, confirmed)
rfbrw - 1 PCB
Pistollero - 1 PCB (Portugal)
luchoh - 1 PCB (USA)
pgour - 1 PCB
agent.5 - 1 PCB (USA)
spm - 1 PCB (UK)
Total = 15
I'll give my friend the weekend to consider the smd soldering proposition and give him a nudge if I don't hear from him by Monday.
diyAchim - 1 PCB
Raj!1 - 1 PCB (UK)
dubulup - 2 PCB (USA)
ericoto -1 PCB
HenSch - 2 PCB (Germany)
mterbekke - 2 PCB (Netherlands, confirmed)
rfbrw - 1 PCB
Pistollero - 1 PCB (Portugal)
luchoh - 1 PCB (USA)
pgour - 1 PCB
agent.5 - 1 PCB (USA)
spm - 1 PCB (UK)
Total = 15
I'll give my friend the weekend to consider the smd soldering proposition and give him a nudge if I don't hear from him by Monday.
I think this is fair. From what I know so far, the toughest test for this is going to come for a well-structured DAC that has direct DSD processing (AKM family is good for it). I have one such animal running off batteries in critical locations. I’ll make the compare as best as one subjectively can. Not going to help those that are hounding for info right now, though.