Dual Mono DAC AK4490/AK4497

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I received my Yanasoft dual mono AK4497 board back from my good friend Ales today; he's done a nice job of the smd soldering...

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


I have my Lundahl 1676 transformers for the output so I need to order the rest of the parts and find the time to see how it sounds.

Have you done any there?
I had a pair of Lundahl LL1545A in my lab, so I tried this.
A dual mono AK4490 demo board from my friend DimDim and each differential output direct to each transformer winding. The trasformer output was a SE L2U LL1545A setup.
The output has a zero almost dc offset and the output is 2.71Vrms.
If you like to have a LPF at 50KHz you must set an appropriate resistor parallel at the SE output with input resistance correlation of next stage.

The result was nice but not outstanding for my taste.

I am wondering if the transformer direct connection of differential AK output is the most fine solution.
I have seen some other implementation like this http://www.audiodesignguide.com/DAC32/Suppler-V-I-LL1684-for-DAC.jpg
If you see the differential output of the dac connected direct to the transformer primary but there are some parallel resistors and capacitors to the earth.
I have seen the same scheme again with some different combinations of 10uF+6.8-20K to the earth.
Is there any suggestion about this?
 
Last edited:
Have you done any there?

No, I've not made much progress; I needed to order an LCD screen to control the DAC so used that purchase to try out the 'From Japan' proxy shopping service but the downside is that delivery takes a long time from Japan!

I will give my transformers a try when I get the build together but I'm also having some second thoughts and have ordered a Broskie BCF board to potentially use as a balance to SE buffer for the outputs from the AK4497s. I had good results with a BCF on the output of an AK4490.

Ray
 
Lemon, i've build a single version with the super v/i. Schematic exactly like audiodesignguide shows in your link.

So generally transformer output works nice with ak4497, at least without summing.

Once i get the time to write some arduino code for syncing, i will try the super v/i on dual version using simple resistors for summing the two ak4497 like shown in the yanasoft manual.

Chris

IMG_1878.JPG
 
More like : this DAC cover now DSD512 native (DSD256 DoP) and PCM 768khz at 32bit or 24 bit those bit depth selectable trough the PC driver, bitperfect host everytime in full output volume and DAC manage the up and down of volume by host mixer or on front panel. Studio quality headphone amplifier with it''s own volume knob , menu to set DSD and PCM filter, on both oled display separately both right and left output level meter, mode played like DSD or PCM bit depth on left screen and on right one display the current filter set. This first prototype has only behind (I will provide photo later) a line out unbalanced and a SPDIF (coax & optical) output, plus one USB A port (USB 2.0) connected to my customized XMOS
 
Yes is on the 216 one, and I should check with XMOS but I suppose we are not able to share modified source from the USB audio 2.0 provided by XMOS. Anyway the hard work done to modify the source who XMOS itself has not reach will be difficult, I know you Dimdim we discuss about the XMOS on your Site sometimes ago, my adventure and personal cost and time invest should have a result and I will be happy to have someone peraps like you to build a 4 layer pcb board for this prototype and other planned models, and we have a possible exchange of coding part ;)
 
@chris303
@lemon

Guys,
One of dacs I have is ak4490 and xmos216 with bitperfect volume control. I use Super V/I LL1684 stage. I tried different caps and got different sound signatures. However thanks to @analog_sa in this forum I found that best cap is no cap at all in this stage! Seriously ...
Just remove those 10uF, 0,1uf, 0,01uf and 10k resistor and short pin 9 and 10 from LL1684 directly to ground. The sound is amazing!. I bet it will also work for ak4497. It is best output stage for akm for me, maybe expect Finamet Teramoto transformer. I did not tried them however.
See Super V/i converted (LL1684 ) thread.
 
Last edited:
@chris303
@lemon

Guys,
One of dacs I have is ak4490 and xmos216 with bitperfect volume control. I use Super V/I LL1684 stage. I tried different caps and got different sound signatures. However thanks to @analog_sa in this forum I found that best cap is no cap at all in this stage! Seriously ...
Just remove those 10uF, 0,1uf, 0,01uf and 10k resistor and short pin 9 and 10 from LL1684 directly to ground. The sound is amazing!. I bet it will also work for ak4497. It is best output stage for akm for me, maybe expect Finamet Teramoto transformer. I did not tried them however.
See Super V/i converted (LL1684 ) thread.

WRT shorting pin 9 and 10 to gnd:

You don't need to short any of the transformer secondary to ground.
In fact that will effectively take away one of it's virtues - galvanic isolation
and separation of grounds.

Just 'float' the secondary and let it get it's ground ref from whatever it is
driving.

Terry.
 
WRT shorting pin 9 and 10 to gnd:

You don't need to short any of the transformer secondary to ground.
In fact that will effectively take away one of it's virtues - galvanic isolation
and separation of grounds.

Just 'float' the secondary and let it get it's ground ref from whatever it is
driving.

Terry.

Hi,

Pin 9 and 10 are on the primary side.

I agree on secondary. No no need for connecting it to ground. It is floating in my setup.
 
I haven't had time to touch my Yanasoft dual AK4497 board since the New Year but have now picked the project up again and am awaiting the parts to finish populating the board. The Super V/I approach, using Lundahl LL1676 transformers) is essentially what I will try as my output, though I also plan to experiment with using transformers to sum the outputs too. See the diagram attached.

BTW, for expediency, here's a link to the Super V/I thread;

http://www.diyaudio.com/forums/quanghao-audio-design/246579-super-v-i-converted-ll1684.html
 

Attachments

  • V DAC output LL1676 2_1.pdf
    18.4 KB · Views: 292
Sorry, maybe it is OT, but would it also be possible/make sense to add an super V/I stage to the AK4495 SEQ?
I would connect AoutLN and AoutLP to the Ends of tthe prim. coil and tie the middle contact of the coil to ground and same for the right side...anything to be considered?

I tried super v/I ll1684 for both ak4495 and ak4490. It works for both. However ak4495 does not have dc offset, so connecting it to super v/I only make sense if you want to sum the signal.
 
You mean I could just take ground and AoutLP (AoutRP) to use it as unsymmetrical out?
Isnt the coil of the transformer a passive LP filter aswell which could substitute the Opamp LP filters of an usual output stage?

Yes you could. I had ak4495 successfully connected like that with tube buffer. Depends to what you are connecting it to. You need try. Impedance match played important role here.
Yes transformer works as LP filter and yes it substitutes opamp. Downside of using transformer is RCA or XLR cables should be relatively short.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.