Schematics for this DAC board
What do we have for schematics?
I have them for "LM-DAC3 V5" and "DAC USB V2.0"; My "Lars Audio V1.0" is similar but with a few mistakes. I found +/-12V runs to the DAC socket pins 17,18!
I started drawing out a schematic to prevent any smoke
These PCB's are only 1/2oz copper and very fragile.
What do we have for schematics?
I have them for "LM-DAC3 V5" and "DAC USB V2.0"; My "Lars Audio V1.0" is similar but with a few mistakes. I found +/-12V runs to the DAC socket pins 17,18!
I started drawing out a schematic to prevent any smoke
These PCB's are only 1/2oz copper and very fragile.
The CS8421 ASRC up-sampler core is run off 3.3V when the design is 2.5V 
datasheet max. is 3.5V. I'm leaving this alone- you could add a LM1117-2.5 on top the AM1117-3.3
Heat is usually why you keep the core voltage low. At 192kHz Icc is high at 80-100mA and the chip's lifetime will prob. suffer.
I found 47uF tantalums best at lowering noise on VC. 100uF electrolytics didn't cut it.
datasheet max. is 3.5V. I'm leaving this alone- you could add a LM1117-2.5 on top the AM1117-3.3Heat is usually why you keep the core voltage low. At 192kHz Icc is high at 80-100mA and the chip's lifetime will prob. suffer.
I found 47uF tantalums best at lowering noise on VC. 100uF electrolytics didn't cut it.
I finally got this DAC board sounding good
I finally got this DAC board sounding good. Before, I couldn't listen to it for more than 10 minutes before reaching for the power switch, which had me wondering what was up.
Spent a day with the scope on 1mV/div using optical and OPA627's.
Adding SMT caps as close as possible to the IC's makes audible improvements:
0.1uF's on (DIP) pins 6,7 VD and 21,22 VA power for the DAC IC.
The CS4398 smurf board has a skinny trace between (DIP) pin 15/17/22. Found a lot of noise (50mVpp) on pin CS4398 pin 17 trace impedance is the problem. Had to go up to 100uF X5R before the noise was reasonable. HUGE improvement in sound, gone is the midrange harshness and detail is back with some velvet.
Added 2.2uF X5R on CS4398 pin 15 (FILT+) like above, across ground fill to the IC pin.
On the motherboard, I shorted the 4 electrolytics on the CS4398 outputs, 2 on the output RCA's and have only +/-3.3mV of DC offset at the RCA's with OPA627's.
The Vregs for +/-12V are the noisiest and 10uF tantalum caps from ADJ divider to GND reduce this.
Added 0.47uF X5R and 10uF tantalum decoupling for the op-amp's rails.
It's musical now, certainly worth the hassle of soldering SMT parts on a cheapo PCB.
I finally got this DAC board sounding good. Before, I couldn't listen to it for more than 10 minutes before reaching for the power switch, which had me wondering what was up.
Spent a day with the scope on 1mV/div using optical and OPA627's.
Adding SMT caps as close as possible to the IC's makes audible improvements:
0.1uF's on (DIP) pins 6,7 VD and 21,22 VA power for the DAC IC.
The CS4398 smurf board has a skinny trace between (DIP) pin 15/17/22. Found a lot of noise (50mVpp) on pin CS4398 pin 17 trace impedance is the problem. Had to go up to 100uF X5R before the noise was reasonable. HUGE improvement in sound, gone is the midrange harshness and detail is back with some velvet.
Added 2.2uF X5R on CS4398 pin 15 (FILT+) like above, across ground fill to the IC pin.
On the motherboard, I shorted the 4 electrolytics on the CS4398 outputs, 2 on the output RCA's and have only +/-3.3mV of DC offset at the RCA's with OPA627's.
The Vregs for +/-12V are the noisiest and 10uF tantalum caps from ADJ divider to GND reduce this.
Added 0.47uF X5R and 10uF tantalum decoupling for the op-amp's rails.
It's musical now, certainly worth the hassle of soldering SMT parts on a cheapo PCB.
The CS8421 ASRC up-sampler core is run off 3.3V when the design is 2.5V
Heat is usually why you keep the core voltage low. At 192kHz Icc is high at 80-100mA and the chip's lifetime will prob. suffer.
I found 47uF tantalums best at lowering noise on VC. 100uF electrolytics didn't cut it.
Is it possible to stack up and connect 2 different ldo voltage regulator together as double voltage regulator...
As an aside from the nuts n bolts of the thing, I'd like to ask if people have ordered from Hifi Diy--Amplifier,dac kits,hifi audio products and parts online shop, shipping to worldwide
I did, almost a month ago (Feb 10-ish). I haven't gotten the DAC yet. I asked them if they had a tracking number they could give me, but they said they didn't. As the item included shipping, I wasn't given a menu of shipping options when I ordered so I don't even know how they shipped it.
I'm just looking for confirmation that the DAC has in the past been delivered by this resource, and approximately how long it took to be delivered to the USA. I've heard that imports from Hong Kong have jumped up in delivery time due to increased security, but I don't know if that's true.
Still watching the mail... for all I know it could be there when I get home...
Thanks
I did, almost a month ago (Feb 10-ish). I haven't gotten the DAC yet. I asked them if they had a tracking number they could give me, but they said they didn't. As the item included shipping, I wasn't given a menu of shipping options when I ordered so I don't even know how they shipped it.
I'm just looking for confirmation that the DAC has in the past been delivered by this resource, and approximately how long it took to be delivered to the USA. I've heard that imports from Hong Kong have jumped up in delivery time due to increased security, but I don't know if that's true.
Still watching the mail... for all I know it could be there when I get home...
Thanks
CS8421 vreg fixes
I think you could fit a 2.5V reg on top of the existing 3.3V reg ("dead bug"); they're three pin SOT-223 vregs so two pins are paralleled.
I wouldn't run one reg after another 5V--> 3.3V -->2.5V instead 5V feed for both and 0.25W max heat dissipation. Or just use a diode in series to drop 3.3V to 2.6V
I think you could fit a 2.5V reg on top of the existing 3.3V reg ("dead bug"); they're three pin SOT-223 vregs so two pins are paralleled.
I wouldn't run one reg after another 5V--> 3.3V -->2.5V instead 5V feed for both and 0.25W max heat dissipation. Or just use a diode in series to drop 3.3V to 2.6V
Hopefully you used Paypal. They state clearly that shipping is 2 wks or less. If you can't get a satisfactory answer from the seller you should file a dispute with Paypal before you run out of time. Snow used to sell on Ebay, I don't know if that's really who you are dealing with though.
Last edited:
just my deal,It took 5 weeks to get mine here and the box was torn and parts were broken on the board,like rca plugs and opto in,I made the mistake,really no choice ordered the toroid with it,I think it bounced around inside the box,My heart sank when I saw that box in the guys hand,NO MORE from that guy in china. I hope you have better luck than I did !
from my understanding after the 3.3V, VD and VL shares the same trace for 3.3V and same as the crystal. In actual VD requires 2.5V, the suggestion for diode in series would be the simplest thing to do, but how it can be done with a such tiny space???
It might end up destroying the pin.....I've try it to other project and the result just as expected......
As I sold my Monarchy M24 to buy another dac I have to wait for, I took the dust of the CS4398 chinese dac with output transformers but stopped listening music soon because I found the mids too harsh.
I tried the mods proposed by Prairiemystic and i'm amazed by what it can do to this dac. As I have to wait 3 more weeks to get my new dac, i'm still tempted to play with the CS4398 dac again. Sometimes ago I thought I reached the end with this dac.