Page 5 of 7
#61
Posted: Fri Mar 09, 2012 12:35 pm
by pre65
On my e-bay AK4393 DAC I have a couple of 51R 2W resistors to drop the 17V AC from the transformer down to 15V AC.
Do these resistors need to be decoupled ?
If so what is an approximate capacitor value ?
#62
Posted: Sun Mar 18, 2012 12:45 am
by colin.hepburn
Can anyone tell me what the EBay AK4393 DACs chip output is i.e. if it's a voltage or current output as I want to try a valve buffer using a Russian valve I have ether the 6n2p /6n23p or does it not make any difference which valves I use on the dac output
#63
Posted: Sun Mar 18, 2012 1:23 am
by Nick
It can be either, but there is a 2.5v DC offset on the output.
#64
Posted: Sun Mar 18, 2012 1:41 am
by colin.hepburn
Thanks nick I did see it had the 2.5v offset when I wired in the FET Buffer does the 2.5v offset give a problem to the buffer
if ether valve or FET I did have a 0.22uf cap on the dac outputs but then removed them I plan to use just a cathode follower to start with
#65
Posted: Sun Mar 18, 2012 9:40 am
by Nick
Thanks nick I did see it had the 2.5v offset when I wired in the FET Buffer does the 2.5v offset give a problem to the buffer
It will depend on the buffer, can't be any more specific.
#66
Posted: Sun Mar 18, 2012 12:05 pm
by Mike H
Surely you can raise the cathode Voltage to compensate?
#67
Posted: Sun Mar 18, 2012 1:34 pm
by colin.hepburn
colin.hepburn wrote:Ok after some confusion with double sided pcbs boards I have had the Ebay FET Buffer up and running the last copual of weeks
As far as the sound improvements go well I would say it's about the same as the step up from the stock NE5532P was to the 4562NA op-amp in my view the 4562NA was a bit more open airy more space between instruments and more detailed a round vocals
Adding the FET buffer kept what the 4562NA op amp was doing but sounded smoother with more width and depth to the sound stage so is a worthy upgrade to the Ebay 2496 Dac fother upgrades to this dac with be the input cap and a couple of oscon caps in the digital psu plus maybe some better Low Impedance 2200uf 63volt
Buffer PSU
Is a Rod Elliott. Project 05 Rev-A LM317/337 regulators
Which has an uneven voltage outputs
17.7v out from the 337
14.9v out from 317
This maybe my oversight in not checking if the two 2,2k resistors on the LM337 side were matched
The FET
Seems to work ok with the voltage difference on the LM 337
The Dac appears to have 2.5volts offset on its output from the Dac chip i first added 1uf poly caps to the dac output this killed the bass and sounded thin change to the Russian K42Y-2 0.22uF PIO capacitor excellent sounding caps and this sorted it out and got it back to the description above I have a match pair of the Russian FT-2 0.1uF teflon capacitor to try i have a second 2496 Dac which i will add a valve buffer too
The Regulator schematic
Just let me say I know nothing about dac chips and buffers as I have never used them before
so I have two of the AK4393 DACs one with fet Buffer and the second will have valve buffer so let's take one at a time to try and avoid confusion lets deal with the FET buffer first which is up and running and as I said in my last post I am aware of the 2.5v offset after checking the dac outs on the DMM
so my thought was to add DC blocking I added 0.22uf caps to the dac outputs
but this seemed to add a tad of distortion so i removed them so the sound is cleaner but the gain could be better i feel
see buffer sch
#68
Posted: Sun Mar 18, 2012 2:44 pm
by richardcooper2k
Hi Colin. My understanding (which is limited) is that the output from the DAC chip does not have a flat frequency response. This is corrected by the resistor network around the opamp. So if you take your output before these and don't correct it you won't have a flat frequency response.
Also the 93 and 96 chips require slightly different filters to get a flat response.
Forgive me if you've already dealt with this and i missed it
#69
Posted: Sun Mar 18, 2012 4:02 pm
by colin.hepburn
Hi Richard
Thanks for your reply
No I haven't done any filtering from the dac as I thought the buffer took care of that unless I missed it on the thread at diyaudio
I was hoping for a response from the members on here who are messing around with this vary dac could add something to this
As it stands the sound with the buffer added is very good not having a flat frequency response never even crossed my mind
#70
Posted: Sun Mar 18, 2012 4:39 pm
by richardcooper2k
If it sounds good; that's the main thing
#71
Posted: Sun Mar 18, 2012 5:49 pm
by Paul Barker
richardcooper2k wrote:Hi Colin. My understanding (which is limited) is that the output from the DAC chip does not have a flat frequency response. This is corrected by the resistor network around the opamp. So if you take your output before these and don't correct it you won't have a flat frequency response.
Also the 93 and 96 chips require slightly different filters to get a flat response.
Forgive me if you've already dealt with this and i missed it
I'm no egg spert but I thought the network was to block a supersonic oscilation I don't recall the frequency, which is less a problem in the valve world.
#72
Posted: Sun Mar 18, 2012 6:22 pm
by colin.hepburn
I really don't know anything about dacs even looking at the spec sheet for this one it may as well be printed in chine's as I said it seems to sound very good better after removing the caps I had on the dac outputs
although the caps did zero the 2.5volt offset it the thing I am trying to get to is if the offset is a problem and what i can do about it and how to do it
#73
Posted: Sun Mar 18, 2012 6:57 pm
by Mike H
Paul Barker wrote:I'm no egg spert but I thought the network was to block a supersonic oscilation I don't recall the frequency, which is less a problem in the valve world.
44.1 kHz? Hence the op-amp based low-pass filter, which may involve a small choke as well. I'm going by the old Philips type schematics.
#74
Posted: Sun Mar 18, 2012 7:12 pm
by Nick
Mike H wrote:Paul Barker wrote:I'm no egg spert but I thought the network was to block a supersonic oscilation I don't recall the frequency, which is less a problem in the valve world.
44.1 kHz? Hence the op-amp based low-pass filter, which may involve a small choke as well. I'm going by the old Philips type schematics.
Its a up-sampling DAC, so the clock rate on the output is way higher than 44k, thats one of the good things about this DAC, it only needs simple (or no) output filtering.
#75
Posted: Sun Mar 18, 2012 7:42 pm
by colin.hepburn
Yep ok thanks nick that's how I now have it output direct from dac pins 23 AOUTL+ and pin 21 AOUTR+ to buffer input sound is good
but the gain of the buffer seems to have dropped a tad how can I change the buffer to get the gain up a tad although it don't need much