That means you directly connect Pos and Neg pins of the dac output to your amp.
But first you need to measure that there is no DC residue on these outputs. Just measure the DC voltage between Pos and Neg, if this is zero or very low, like 0,01V that it would be OK.
You know some capacitors also show some DC leakage on the output (I measured 0,007V dc on a new AN paper in oil cap)
Just make sure your amp and speakers do not get damaged by DC!
Regards,
you are welcome...
... I am not a technician, but my stomoch "tells" me that we will need something like this not only for the 12 Volt but the supply of the 5 Volt as well...
The isolator recommendation is not to make this better than Russ' solution, it is meant to improve the rpi i2s sound if you are using the rpi. Obviously, an ARM solution that can use a low jitter RCK(s) must be better than the rpi's clock. But if you are using the rpi *right now*, the isolator is a nice and cheap improvement on the sound, since Russ' product is still not available. Besides, I'm sure he will be isolating the bbb form the cape and would not be surprised if he uses the NVE isolator.I think I will wait for the BBB with the Cape from Russ. I expect he can make it nicer than I can
Well, for others, besides Supersurfer, that are using the rpi, try the isolator. I think this will be my last rpi/i2s post here since Russ' solution negates all other approaches. Enjoy.
Last edited:
Check this: P5 header
And this: building-open-embedded-audio-applicance
I am a Linux noob so do not ask difficult questions on low level software issues
First I used a distribution with mpd build by a friend of mine, later on RaspiFi and now Volumio.
Volumio is very easy to install and use, and has I2S output on the Pi and within the coming months probably also for beagle bone black!
thanks a lot for the links ...
best regards
Yes, back in England now after a nice little adventure at the weekend!
Thanks so much to Reinhard and Klaus for the invitation, for taking care of us all with food, drink and wonderful hospitality
So nice to travel so far and meet up with a group of total strangers, but to feel straight away like a bunch of friends all excited together
Klaus's horns and the whole, incredible, purpose-built listening room are truly a thing to behold
And so great to get a good feel for the different variations of Doede's DDDAC design and hear the difference side by side that the modifications and power supplies make to the detail and character of the sound. I'll write a bit more when I have a spare moment, I just wanted to say a quick thanks and how it was great to meet everyone there
cheers,
James
Thanks so much to Reinhard and Klaus for the invitation, for taking care of us all with food, drink and wonderful hospitality
So nice to travel so far and meet up with a group of total strangers, but to feel straight away like a bunch of friends all excited together
Klaus's horns and the whole, incredible, purpose-built listening room are truly a thing to behold
And so great to get a good feel for the different variations of Doede's DDDAC design and hear the difference side by side that the modifications and power supplies make to the detail and character of the sound. I'll write a bit more when I have a spare moment, I just wanted to say a quick thanks and how it was great to meet everyone there
cheers,
James
I think a power supply with low impedance is most beneficial in the analog domain, so for the 8v stages of the dac. The digital domain is less critical in this regard. Digital benefits more by good power supply rejection of noise and isolation of digital noise:so the noise produced in the digital domain cannot enter the analog domain of the dac or amplifiers.
Low impedance supplies can be made by simply overdimensioning; like Bernts supply, of by smart design and use of shunts
The supply of James shows this nicely by bypassing some resistors and the fuse. In this way you will have created a bit more ripple on the dc output but the benefit of lower impedance is much greater than a completely flat dc line. The dc will also be regulated again by the regulators on the board so we can easily get away with this.
Overregulating is not good
I think I will try an unregulated raw power supply on the dac, just to see what it does.
A nice approach will be: big transformer, heavy rectifiers, nice elna silmic caps and chokes ( no resistors in the path); nothing more.
If only I had more time to experiment
I think a power supply with low impedance is most beneficial in the analog domain, so for the 8v stages of the dac. The digital domain is less critical in this regard. Digital benefits more by good power supply rejection of noise and isolation of digital noise:so the noise produced in the digital domain cannot enter the analog domain of the dac or amplifiers.
Low impedance supplies can be made by simply overdimensioning; like Bernts supply, of by smart design and use of shunts
Interesting thoughts.
The supply of James shows this nicely by bypassing some resistors and the fuse. In this way you will have created a bit more ripple on the dc output but the benefit of lower impedance is much greater than a completely flat dc line. The dc will also be regulated again by the regulators on the board so we can easily get away with this.
Overregulating is not good
Then what do you think of swapping regulators ?
Maybe, like you said, it's better to make the PSU more direct, at a cost of some ripple, and add top notch regulators after ??
Yes! If we have improved the regulation further downstream with the shunt regs, then a very simple, very low impedance power supply should work very nicethink I will try an unregulated raw power supply on the dac, just to see what it does.
A nice approach will be: big transformer, heavy rectifiers, nice elna silmic caps and chokes ( no resistors in the path); nothing more.
If only I had more time to experiment
I'd love to try that to compare, but it gives me a few questions:
If there's no voltage drop from a regulator or resistors, then we need a lower voltage Transformer. 9v ac would come out somewhere around 11.5v dc by my calculations. Guido's specs seem to suggest around 3v headroom so that should be OK for the 8v shunts? Or would you go larger?
Any recommendations for suitable chokes?
What size of Silmic caps do you reckon?
Thanks,
James
Thanks Supersurfer, it does indeed work very nicely with 4 boards into a 20k load. I previously had two boards and less than 5k load which gave me a bit of distortion.
David
Hi guys,
I found some nice chokes:lundahl ll1694
I am thinking of this:
Simple but large EI transformer-rectifiers( have to find some nice ones, quality here is very important)-cap- choke-cap-choke-cap
I have some black gates 3300uf16v, and also elna silmic 4700uf 16v.
The transformer can be a 9v type.
Regards,
I found some nice chokes:lundahl ll1694
I am thinking of this:
Simple but large EI transformer-rectifiers( have to find some nice ones, quality here is very important)-cap- choke-cap-choke-cap
I have some black gates 3300uf16v, and also elna silmic 4700uf 16v.
The transformer can be a 9v type.
Regards,
Yes but you have to deal with higher radiation with EI cores... and in a low current circuit...
Seems to be pretty hard to overcome in a single chassis, unless you find the very few manufacturers capable of winding EI cores with some "secrets" that make the difference but costs a lot.
Generally where would you buy EI cores ? I can't find online stores for them.
Seems to be pretty hard to overcome in a single chassis, unless you find the very few manufacturers capable of winding EI cores with some "secrets" that make the difference but costs a lot.
Generally where would you buy EI cores ? I can't find online stores for them.
See a sheet metal shop and fold up a zic coated box that fully encloses the tranny. We use an EI in the Killerdac with no ill effects, they only leak from where the laminations join. Make sure the shield is ferrous material. Also an r core tranny is worth looking into as well...
Chuz,
Drew.
Chuz,
Drew.
Hello
Has anyone tried to change the power supply of the DDDAC for some power supplies or voltage regulators from Paul Hynes ?
Thanks
thank you, for all those who asked for a closer look onto Bernd's monster controlled power supply, now "finished" in a cabinet, weighing more than 20 kilograms, here it is:
and for more have a look here:
http://killerdac.com/forum/index.php/topic,842.200.html
No. 205
but please, don't ask me for details, regarding your questions go ahead and ask Bernd himself:
berndwieland11@web.de
thanks
An externally hosted image should be here but it was not working when we last tested it.
and for more have a look here:
http://killerdac.com/forum/index.php/topic,842.200.html
No. 205
but please, don't ask me for details, regarding your questions go ahead and ask Bernd himself:
berndwieland11@web.de
thanks
Last edited:
Ok, here’s one for you. I’ll probably be shot for suggesting this. But it’s been bothering me that with only using single-ended output I’m in affect only using half the power/gain output of the DDDAC, by connecting either ENG or POS and common and I could do with a bit more gain anyway. I was also curious about the amount of DC de DDDAC produced. Well in my case it was 0.008Vdc&0.000Vdc R/L respectively. So I thought what the heck, it will be a nice experiment to do away with the output buffer caps and since I knew what Rload resistance was in my power amp (I put it there), I got rid of Rload in the DDDAC too. So now I’ve connected POS/NEG single-ended RCA out to my amp. I’m sitting here listening to it now and I can not put any fault to it, if anything it sound better with +/-15Db more gain (someone will be able to do the math on this). I also made some changes to my PSU so it’s hard to tell what’s what, but less is more in my book & no output caps..
Any comments?
Any comments?
Last edited: