So even tho its early without the dac-end 2 being broken in. Would you say the new dac is better / more potential than the gigawork dac with transformer outputs?
As that is the same dac I've been using, looking forward to building the dac-end 2 as more and more parts are arriving daily
well... its been playing music all morning.
the dac-end clearly has potential. its very natural sounding, superb top end (maybe slightly rolled off? not sure but its definately sweet with no hint of sibilance. This is using jan phillips - actually looking at audiodesigns tests it does begin to roll off before 20khz). Generally i would say that it has a warmer presentation, slightly more bodied sound.
The old transformer dac was very good though, it sounded much much better than my cambridge audio dacmagic, not doubt about that. bear in mind i used sowter transformers (expensive and very good)
which is better? not sure yet. Which am i keeping? well i sold the tranformer dac yesterday so that i have confidence in the dacend.
The dacend will satisfy my tweaking needs more. I already can see the potential for a 6n6p ouput stage, gyrator or CSS loaded, led biased. i do love the 6n6p.
the dac-end clearly has potential. its very natural sounding, superb top end (maybe slightly rolled off? not sure but its definately sweet with no hint of sibilance. This is using jan phillips - actually looking at audiodesigns tests it does begin to roll off before 20khz). Generally i would say that it has a warmer presentation, slightly more bodied sound.
The old transformer dac was very good though, it sounded much much better than my cambridge audio dacmagic, not doubt about that. bear in mind i used sowter transformers (expensive and very good)
which is better? not sure yet. Which am i keeping? well i sold the tranformer dac yesterday so that i have confidence in the dacend.
The dacend will satisfy my tweaking needs more. I already can see the potential for a 6n6p ouput stage, gyrator or CSS loaded, led biased. i do love the 6n6p.
Last edited:
Thanks for giving your feedback.
Have you planned to test different output capacitors such as PIO or Teflon? (yes crazy)
I can see a lot of potential in this project. The one with output transformer does not give much to tweak. The transformer is more responsible in that case but if good one it gives a very special sound. So if you say this DAC End is up there then is really good.
I love 6n6p too. Have you tried the 6n6T Beijin? The top is unbelievable.
This weekend I will run my DAC. I will publish results soon.
Last edited:
i dont plan to change the output caps. i messed around with caps with another project and found the obbligatos to be the best value.
one area i will definately play with is the cathodes of the tube stage. From experience the sound is influenced by the bypass caps. Currently have panasonic FC in there. I'll probably try oscons (as recommended) and another thing to try is get rid of the resistors and caps all together, replacing with leds (two yellows in series should be about 3.8v typically i have found led biasing gives a slightly clearer sound.
Question for the designer.
Why run the 5687s at the selected operating point. Looking at the datasheet looks like a more linear region could be obtained?
one area i will definately play with is the cathodes of the tube stage. From experience the sound is influenced by the bypass caps. Currently have panasonic FC in there. I'll probably try oscons (as recommended) and another thing to try is get rid of the resistors and caps all together, replacing with leds (two yellows in series should be about 3.8v typically i have found led biasing gives a slightly clearer sound.
Question for the designer.
Why run the 5687s at the selected operating point. Looking at the datasheet looks like a more linear region could be obtained?
Last edited:
Regarding bypass caps, I will try between Oscon and Elna Silmic but I think at the end I suspect I will end up putting a pair of RIFA silver bullets.
I'm afraid Oscon will have problem with bass. The top will be good but bottom... mmm i don't know. Rifa by experience makes good balance.
Let us know.
Audiodesign knows E182CC and 5687 really well.
I guess the operating point is at a balance between linearity, amount of swing needed, noise figure and sound. Low current= less noise also sometimes the tube sound less hard.
Last edited:
Avoid long term dry soldering joints
Because the 2.2R resistors are running very hot this could be an idea how to avoid dry solder joints problems in the future by dissipating the heat from legs.
The aim is to keep the soldering joint under 80 degrees.
The resistors can work as hot as over 200 degrees no prob.
Because the 2.2R resistors are running very hot this could be an idea how to avoid dry solder joints problems in the future by dissipating the heat from legs.
The aim is to keep the soldering joint under 80 degrees.
The resistors can work as hot as over 200 degrees no prob.
Attachments
Last edited:
less hard because of more distortion probably.
thats where are i think. Its not too bad, but you can see why people run it hotter. A bit more B+ and a higher current would be slightly better, but then you are into the realms of a different powersupply.
I am not criticizing - it sounds great. I just wonder what audiodesigns reasons are for that point. He probably has 1000X more experience than me so would be good to get some insight.
A simple mod with this could be to ditch the 47k resistors and build a simple cascode CSS with led ref (keeping the bias scheme). That would flatten the load line and probably reduce distortion a little (although its only 0.2% anyway! we arent swinging much voltage anyway, so i really am being picky i think. It would increase the gain too, which probably isnt wanted.
Also - i know others have said it, but Thanks to quaghao and audiodesign for their efforts. This is a very fun project with countless opportunities for tweaks.
I love that i have just put an image up from 1959.... 25 years older than me. Hasnt technology moved quickly!!!!!
An externally hosted image should be here but it was not working when we last tested it.
thats where are i think. Its not too bad, but you can see why people run it hotter. A bit more B+ and a higher current would be slightly better, but then you are into the realms of a different powersupply.
I am not criticizing - it sounds great. I just wonder what audiodesigns reasons are for that point. He probably has 1000X more experience than me so would be good to get some insight.
A simple mod with this could be to ditch the 47k resistors and build a simple cascode CSS with led ref (keeping the bias scheme). That would flatten the load line and probably reduce distortion a little (although its only 0.2% anyway! we arent swinging much voltage anyway, so i really am being picky i think. It would increase the gain too, which probably isnt wanted.
Also - i know others have said it, but Thanks to quaghao and audiodesign for their efforts. This is a very fun project with countless opportunities for tweaks.
I love that i have just put an image up from 1959.... 25 years older than me. Hasnt technology moved quickly!!!!!
Last edited:
I use the 6n6p in my Welborne/Cato linestage with a battery powered filament - really nice
For those experimenting with cathode bypass please keep us posted. I have decided to start with Elna Simlic IIs with a Russian PIO .1microF. I am about 3 eBay parts away from completion.
Also any thoughts about using a 200microF Solen at C5 on the HV PS.
Thanks
Bob
Bob
quick question. I was just doing final checks before putting the top on and putting in the rack.
The salas shunt for the digital 5v actually drops to 5.2v with the load. quick inspection and i have put a 33ohm resistor instead of the 27ohm. is this likely to be my problem - too little current from shunt, or should i look else where. would you even bother about it?
The salas shunt for the digital 5v actually drops to 5.2v with the load. quick inspection and i have put a 33ohm resistor instead of the 27ohm. is this likely to be my problem - too little current from shunt, or should i look else where. would you even bother about it?
I don't know how and why is set here, but I can tell you that 33R will give about 60mA when 27R will give about 75mA. If you want good Zout there must be enough spare. Also how do you set at 5V when there are TL341 next? Don't they need a margin for drop on their 10R? TL431's are substantially noisier above 1mA self current BTW. 0.2-0.5R Zo, 50nVsqrtHz...I would skip em and remote sense the main reg directly. Your project will benefit much if you watch margins and settings IMHO.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.