I'm going to build a second unit but use the Salas shunt regs for the +-12 VDC on the analog section.http://www.diyaudio.com/forums/group-buys/188974-gb-thread-salas-ssr1-1-bib-shunt-reg.html
I will leave the SMD reg's in place for the low voltage but feeding them from the shunts should help them as well.
I will leave the SMD reg's in place for the low voltage but feeding them from the shunts should help them as well.
Last edited:
Feeding the regs from the shunts might help, but not that much.
The question is that the final regulator will imprint its main characteristics to the audio. If you preserve the SMD regs, which are not shunt types, they will imprint their quality.
But you do have to pre-regulate what goes to the SMD regs, because they can't dissipate the full voltage coming from the raw supply.
The question is that the final regulator will imprint its main characteristics to the audio. If you preserve the SMD regs, which are not shunt types, they will imprint their quality.
But you do have to pre-regulate what goes to the SMD regs, because they can't dissipate the full voltage coming from the raw supply.
Sure,noise will be cumilative but since the shunts will be superior in this regard to the LM78/79's I suspect there will be some improvement.I am however of the opinion that providing the purest DC possible to the output buffers(vs the digital IC's) will result in the most significant gains sonically.
One might also try to omit the L2 inductor before DVDD (pin 2 on 439X)..
AKM 2004 applications guide somehow do not recommend placing inductors.. On page 4 it says 'do not place inductors in the power supply lines, as these restrict the flow of current to the chip and can reduce performance') Might be easy to try with/without L2 if there is some difference (assuming your supply has some care upstream already (?) ) (best to measure)
L2
Bypass as close to the consumer pin: AKM designers are of course talking about smd ceramics btw, which can be placed much closer to VD consumer pins than the throughhole components on our 2496 mini (C18 for example which are, few mm away and might render them more ineffective (?) )
And, adding bypass to Oscon should not be done (according Dario, Eva, Elvee etc.. look this for more information http://www.diyaudio.com/forums/powe...ilm-caps-electrolytic-caps-4.html#post2257381)
In this case its possible to put your Oscon into place originally designed for the small bypass (example; if you have some 100uF Oscon in C30 >> put it in C18 position you get the idea ^^)
AKM 2004 applications guide somehow do not recommend placing inductors.. On page 4 it says 'do not place inductors in the power supply lines, as these restrict the flow of current to the chip and can reduce performance') Might be easy to try with/without L2 if there is some difference (assuming your supply has some care upstream already (?) ) (best to measure)
L2
An externally hosted image should be here but it was not working when we last tested it.
Bypass as close to the consumer pin: AKM designers are of course talking about smd ceramics btw, which can be placed much closer to VD consumer pins than the throughhole components on our 2496 mini (C18 for example which are, few mm away and might render them more ineffective (?) )
And, adding bypass to Oscon should not be done (according Dario, Eva, Elvee etc.. look this for more information http://www.diyaudio.com/forums/powe...ilm-caps-electrolytic-caps-4.html#post2257381)
In this case its possible to put your Oscon into place originally designed for the small bypass (example; if you have some 100uF Oscon in C30 >> put it in C18 position you get the idea ^^)
Last edited:
I have 2 Salas shunt board I´m thinking of doing this with.Or would it be better to use them with th A+5 and A+3,3v and leaveing the D+3,3v as is?
Edit: Does the VL VD get 9v from the regukator?
Edit2: See now that VD VL is 3,3v what is the 9v used for then,is only as prereg for the D3,3v?
Last edited:
- Yes assume the A+5 and A+3.3V will benefit more from your 2 salas boards (in the future you could well add some lownoise design to D+3.3V and use ears
) keep feed as short as possible- yes LM7809 is prereg to VD+3.. eventually because of how much heat the tiny SMD regulator can dissipate (voltage drop) (?)
Better make one change at a time and listen what you like. /me think it can sound incredible good with low jitter transport
Since that i'm not using opamps in my ak4396 based soundcard hence direct-out mod, single clean 5v source is enough to supply all dacs and 3.3v digital parts via National's LM1117-3.3 . I didn't compare 1117-3.3 variants of other companies but National's noise figure and other specs are promising.
LM723 voltage regulator is very stable and clean source for this kind of applications in my humble opinion. LM723@5v + 22.000uF + power transistor(s). This is my current setup. You can find lots of schematics based on LM723 around.
LM723 voltage regulator is very stable and clean source for this kind of applications in my humble opinion. LM723@5v + 22.000uF + power transistor(s). This is my current setup. You can find lots of schematics based on LM723 around.
IMO the priority would first be to provide the output buffer with the best supply possible.If you have enough boards to also do the 3.3 and 5 volt analog reg's then by all means do it.
If your transformer for the digital supply has two secondaries you could use one for the third salas shunt board and use it to feed the A+5 voltage to the DAC chip.Some trace cutting may be required and the complexity increases substantially.
OK, I just stumbled on this thread and have to give my 2 cents.
TDA 1541 or AK4396 is not a debate as long as there is any OpAmp in the
path. Sendler, a few others and me keep on suggesting to just get rid of anything after the dac exept for a coupling cap. Man guys it is not so difficult to just try it out. If any output stage ist needed use tubes like the Lampizator of Lukasz or a D1 like discrete approach, but get that crap out of the signal before you talk about sound quality.
Greets,
Klaus
Additionaly, if your amplifier has a cap at input (yes, mostly there is), you can bypass it (amplifier's input cap) by doing what Radian talk about above. This will make you closer to your source even closer. Commercial and DIY amplifiers have this cap at input besides some exceptions. I'm assuming your amplifier has NOT any possible dc generating circuit, opamp etc. at preamp side. This is important. Otherwise there will be a risk for your speakers while avoiding input cap.
I'm using axial type metal body military grade NOS 150-220uF tantalums as coupling cap. This may became strange but, i see no weakness on bass, mid and treble with these tantalums.
This is a similar nos Sprague tantal that i have.
Last edited:
It is correct. AK4396 datasheet page 34 already states that analog output voltage is 2.8V p-p.
I just put together this kit DAC 2496 (AK4393) DAC KIT With CS8416+AK4393+5532 and also purchased an info HA usb/spdif converter, with the idea of (eventually) copying my CD collection onto a portable hard drive, if the sonic performance justifies it.
No mods yet to the DAC, but I have ordered an OPA627 opamp.
After briefly evaluating the set up through headphones(I find it is easier to critique component changes with headphones) using:.
1. laptop playing uncompressed .wav files through the usb/spdif converter and B4 buffer preamp
2. digital out from my Onkyo DX4555 into the DAC/B4.
3. headphone direct out from the CD player,
I found that whereas 1. provided a lot of detail, a little bright, maybe(I have 70 year young ears), the bass too was very light. Number 2. revealed similar detail etc, but with fuller bass. Number 3. was the least musical.
So I assume the info HA usb converter is, to my ears, the weak link in the computer setup.
Can anyone share their experiences of USB/spdif converters using this DAC?
Peter
No mods yet to the DAC, but I have ordered an OPA627 opamp.
After briefly evaluating the set up through headphones(I find it is easier to critique component changes with headphones) using:.
1. laptop playing uncompressed .wav files through the usb/spdif converter and B4 buffer preamp
2. digital out from my Onkyo DX4555 into the DAC/B4.
3. headphone direct out from the CD player,
I found that whereas 1. provided a lot of detail, a little bright, maybe(I have 70 year young ears), the bass too was very light. Number 2. revealed similar detail etc, but with fuller bass. Number 3. was the least musical.
So I assume the info HA usb converter is, to my ears, the weak link in the computer setup.
Can anyone share their experiences of USB/spdif converters using this DAC?
Peter
Thanks for all the help so far.
I am preparing to populate the pcb and have a few questions.
I plan on not populating op amp side of the board and going the directly out from the dac with a capacitor. I found a 2x 12v power supply in an old Reckhorn active crossover and would like to use it to power the DAC. The power supply has only one ground wire and 2x 12V wires. Is this a problem with the DAC board? Could I just bridge the ground terminals and connect the one ground wire from the power supply?
Thanks again and can't wait to finally hear the dac.
I am preparing to populate the pcb and have a few questions.
I plan on not populating op amp side of the board and going the directly out from the dac with a capacitor. I found a 2x 12v power supply in an old Reckhorn active crossover and would like to use it to power the DAC. The power supply has only one ground wire and 2x 12V wires. Is this a problem with the DAC board? Could I just bridge the ground terminals and connect the one ground wire from the power supply?
Thanks again and can't wait to finally hear the dac.
You're not specifying if that 2 x 12v is rectified or regulated either. Can you measure what voltage output are you getting with some load? It might be interesting to regulate it, if it already isn't, in order to lower the voltage.
Remember that you shouldn't send more than 4 volts DC difference to the smd regulators, as their dissipation is rather small.
Of course, if you are not using the output amp, you will not need the negative supply.
Remember that you shouldn't send more than 4 volts DC difference to the smd regulators, as their dissipation is rather small.
Of course, if you are not using the output amp, you will not need the negative supply.
I think I hear an echoooo.
As for the 4 volt differential that would be more than plenty to keep the 1117's from dropping out,but the datasheet indicates up to a maximum input of 15v is allowable.For that matter the on board 7812 is functioning as a pre reg.feeding two of them.