Upgrade AKM4493 DAC

I have a few year old Papu TA12 DAC with bluetooth. My first impression of the unit was bad. With the stock NE5532 op amps it was both flat and the top end was still harsh.
I replace the op amps with some OPA1612Q1 and it didn't get any better. Tried some OPA1656 and it was about the same. I put it on the self and ran something else.
A few weeks ago I got a Douk K4 DAC with the ES9038Q2M and it works nice with quite good sound out of the box. I change out the op amps with the ones from the TA12 and it was really good after that. I took the stock NE5532 from the K4 and put them into the TA12 just to test them.

The TA12 really sounded better with the "other" newer NE5532 from the Douk. I'm guessing the 5532 that came with the TA12 were fake or just low grade.

I think the AKM boad just might be a nice DAC if I can upgrade a few parts. I could use some help identifying what parts might make the most improvement.

I dont know if the caps are real or fake either. I'm trying the find the coupler caps from the DA which is differential and they don't use a summing amp. Looks like the 390pf is it. Ideas welcome
.
Thanks
Dave


DSCN0584.JPG
 
I have seen far worse China crap than this. It is extra part now.

I bought the S.M.S.L D300 with the Rohm's DAC chip BD34301EKV for my main DAC. It really sounds great with rich timber and the board is well done. That leaves me 2 DAC that I can play with and see if I can get them closer to the sound of the MUS-IC chip-D300 setup. I really do think that MUS-IC chip will do well if they can get the price down over time.

For this board all I want to do is clean up the PS and upgrade the coupling caps and that's it. I have a little space for bypass caps so I will. Then roll a few op amps and see what I get. Its not too bad right now with the better 5532.

I dont quire understand on this one how they feed the op amps yet with out tracing it a bit more. They are 2 channel OP amps with dif inputs but I think they only use 1.

If it fails then no loss, I will toss a little time at it.

 
That must be the worst layout I've seen... ever.
Glad you like it.😵

Looking at how nice many of the $150 to $200 DAC are, the time and money on this one will be limited. I thought there was a lot of love for the AKM chips.....

I think I will throw some film bypass/de couplers on what I can on the back of the board. Found some surface mount Panasonic 1uf 12V that are cheep and will fit. I have a few polystyrene caps coming in that I will try to de couple the input voltage of the AKM chip and perhaps the bluetooth card from the top.

If I dont need to do coupler caps between the AKM and the op amps I will call it a day.

Anyone have a pinout of the AKM4493? They want me to join the club to get it.

Thanks so far guys.

 
  • Like
Reactions: StevenCrook
I received my caps the other day and did the install job. I did remove two of the Nichicon 470uf for two Rubycon 820uf. I bought 15 but it looks like the stock Nichicon might be real and not fakes but I can be 100% sure. I'm only going to change the 2 and leave it for now. I put them on their side to allow for more room DIP8 for larger op amps boards.
DSCN0595.JPG
 
Last edited:
Perhaps not the answer you were looking for but do yourself a favor and put it back on the shelf. Layout is garbage and low-grade cheap parts throughout.

Exactly. Opamps are overrated. Each one works optimally only in a specific circuit and rolling them is a ludicrous nonsense. Unless you want to colour the sound, i.e. making it even less faithful to the source. In a good circuit with a proper layout the 5532 will sound essentially transparent.
 
And the back side.

I did go back and flip all the Wima's to be the same orientations on both sides. If anyone know what the outside layer might be for the future it might be helpful. I don't have my scopes at this houses so could not to a quick look at the noise pickup. I did have 2 of the Muses 8820 to try. I did have some OPA1612 in that I got from Ebay but I'm thinking they might be fakes or just not a good match for this DAC.

My first impression after adding the caps was not too great. I let it run all day and checked it later that night.

The sound had changed quite a bit. With the Muses it really was better. It now sounded better then my other cheep DAC. I am impressed with the Muse op-amps. Really nice tone and timber.

DSCN0599.JPG
DSCN0598.JPG
 
The finished DAC next to my A10 amp. I would disagree about doing mods or op-amp rolling. My other DAC is the SMSL D300 with the new MUS-IC DA chip and it really is worth trying out but before the mods I could hardly stand the AKM and with the new op amps and the cap upgrade the difference is fairly small between the $400 D300 and the $69 AKM. After mods Im still have under $100 in it.

I did stick the 8820 in the Douk with the ES9038Q2M. It has a better PS with chokes, better bypass and the like. With the 5532 op amps it was just too bright and brash. With the 8820 it is much better. I'm going to try some others but I do have to say good things about the 8820

I consider this upgrade a success. The op amps were cheep and the caps we under $10. It is now a usefull nice souning board.

 

Attachments

  • DSCN0603.JPG
    DSCN0603.JPG
    294.2 KB · Views: 185
  • DSCN0592.JPG
    DSCN0592.JPG
    501.6 KB · Views: 145
Wima and other film cap sound can vary depending on the specific dielectric, how long the caps have been polarized to a given DC bias, and the typical type of AC signal going through them. Some high quality film caps have taken a week or even two weeks to fully settle in. Sometimes turning around a cap can change the sound more because of suddenly changing the DC bias on each side rather than because of changing what the outside foil is attached to. IME to really know the effects of component changes like that can take some patience. Just sayin' 🙂

Regarding opamps, there are some circuits that can work perfectly fine with a variety of opamps. In fact, PMA did some blind listening tests of some well-designed generic non-inverting unity gain buffer audio opamp circuits to see if anyone could hear a difference. At least one person showed they could hear differences of low-level distortion in double-blind testing.
 
Last edited:
Wima and other film cap sound can vary depending on the specific dielectric, how long the caps have been polarized to a given DC bias, and the typical type of AC signal going through them. Some high quality film caps have taken a week or even two weeks to fully settle in. Sometimes turning around a cap can change the sound more because of suddenly changing the DC bias on each side rather than because of changing what the outside foil is attached to. IME to really know the effects of component changes like that can take some patience. Just sayin' 🙂

Regarding op-amps, there are some circuits that can work perfectly fine with a variety of opamps. In fact, PMA did some blind listening tests of some well-designed generic non-inverting unity gain buffer audio opamp circuits to see if anyone could hear a difference. At least one person showed they could hear differences of low-level distortion in double-blind testing.
I almost made it through the long op amp thread, got to page 120 of 125...... Like everything it depends on where your system is at and how you listen. Using the Ipad and ACC bluetooth I could hardly tell any difference between op-amps. With a laptop and APTX-HD it was not hard to hear a difference between op-amps. For some reason the USB connection sounded harsher. I need look at the source more with that. Perhaps HD streams will perform better.
 
For the AKM board I did pick up a few polystyrene caps. I could use them for bypass or as couplers. If you guys have any ideas as to the best place to use them that would be helpful. I have some 750pF 33 V that are small enough to to get on the surface mount parts and other larger ones.
I was thinking of removing the 390pF around the op amps and adding the 750.
The DAC is sounding nice now and I'm doing a op amp little rolling with it.
Thanks guys
Dave
 

Attachments

  • DSCN0602.JPG
    DSCN0602.JPG
    492.8 KB · Views: 113
  • DSCN0582.JPG
    DSCN0582.JPG
    385.3 KB · Views: 106
Those small value capacitors around the op amps are likely part of a filter and not the power supply. You’re going to be better off sticking with the same values unless you can use the data sheet and reference the suggested circuit.
 
Sorry if i am rude, but i don"t understand yout mods, because you do not understand what are you doing... I see very bad ...for example 820uf ZL on signal output ???
The 820uF is just a PS filter cap not a coupling cap. All the larger caps on the board are filter caps. I changed the op amp caps to a slightly better cap for two reasons. First, if the stock caps are fake then the new cap will have a lower ESR so have better performance. The Rubicon replacement 820 does have the lowest ESR I could find for a standard electrolytic but really not much better than the stock if they are real.

I also needed to make room for discreet op amps so the caps needed to be on their side. If you have ideas on how to make the board sound better that would be nice.....