ES9038Q2M Board

Daniel,
The oscillator you linked to might not any better than the one that comes with the dac already, don't know. If you can't get a 100MHz Crystek 575, I would suggest to try an 80MHz NDK SD. Some people have reported pretty good results with that one. NDK NZ2520SD 3.3V 80Mhz 49.152Mhz 45.1584Mhz 24.576Mhz 22.5792Mhz 12Mhz Ultra low phase noise oscillator - DIYINHK

While a clock upgrade should help sound quality some, you may find the difference will be more noticeable after you have done AVCC and output stage mods. Also, I added a dedicated low noise voltage regulator to power the new clock on my modded dac board. It was attached to the ground plane side of the dac underneath the clock so as to make it physically close to the clock and on the same ground plane. Up to you how much you want to do. Dedicated voltage regulators for the various dac loads does help, but decreasing DPLL Bandwidth as much as possible is can probably make a bigger difference than just changing the clock. Problem with the DPLL bandwidth adjustment is that it takes an Arduino or other MCU and some pretty tiny soldering on the dac board, so some people didn't want to try it.
 
Last edited:
You can solder legs onto the NDK clock with pieces of bare wire-wrap wire. Then solder the wires to the PCB pads. One way to do that is get some gaffer tape and lay a piece of it upside down so the sticky part is on top. Then tape that piece down to your workbench with four strips of tape, one for each each edge of the upside down tape. Then stick the NDK clock upside down on the upside down tape. At that point you can put a tiny blob of solder on each NDK solder pad. Then solder on some bare #30 gauge solid wire, one short piece to each NDK solder pad. Now you have wire legs on the NDK clock that you can solder to the dac PCB clock solder pads. Make sure you don't short anything out by touching the bottom of the NDK clock to the PCB. A small strip of Mylar tape under it can provide safety insulation to prevent accidental shorts.
 
Last edited:
Hello,
I found a Es9038q2m board in China, it is 20 USD board with dual NE5532 opamps and 27.5 USD with OPA1612+AD8397 opamps.

My question is : Which alternative do you prefer to me?
20 USD package or 27.5 USD package?

Note : I will add Csr8675 Bluetooth 5.0 module (Aptx-HD and LDAC firmware module) to Es9038q2m Via I2S.
Board dimensions are : 68 x 30 mm, so it is enough small mobile usage for me.

Thanks for your answers...
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.
 
Hi Adrian,
The images you posted do not display where I am. Maybe you can provide links to the board(s) you found? Or, maybe try attaching pics as files appended to your post rather than putting them in the body of the post.
To add files that way, click on the 'Go Advanced' button, then scroll down to the 'Manage Attachments' button. Clicking on that will let you upload files as attachments. That way they will always be available to all readers.
 
Hi Adrian,

I am familiar that DAC board because I bought one to check out. I thought it sounded awful and I would not recommend it to anyone.

Nonetheless, OPA1612 is a very good opamp for use in dacs. Don't know about using AD8397, which is probably intended as a low cost headphone amp. Probably not the best choice for that, but unfortunately the best one can't be properly used with dip socket adapters (OPA1622).

What I think makes that particular dac board sound so bad is the overall design. Its not something that can be fixed by plugging in better opamps.
 
Member
Joined 2018
Paid Member
You can solder legs onto the NDK clock with pieces of bare wire-wrap wire. Then solder the wires to the PCB pads.

This is good useful info thankyou. I have tried soldering NDKs into Ian Canada adaptor boards to to avail but this gives me more options.

Did anyone try the Aliexpress Accusilicon AS318...note none 'B' suffix. They are same paxkaging as the NDK s....ie...small!

Wondered on their performance. I bought some but not tried yet.
 
Can you offer a good sound quality DAC board please?
I want to add Bluetooth module on it Via i2s.
If possible, 5V DC type please. My budget is 50 - 150 USD.
Many thanks.

Hi Adrian,
Don't know that I can suggest a good dac board for you. Its that I haven't found any in that price range that I personally like the sound of. Maybe someone else knows about one and will be willing to chime in with some helpful advice. Sorry I don't have a better answer for you.
 
This is good useful info thankyou...

Hi Jim,
Actually, I haven't found it necessary to solder wire legs onto small clock modules if there are small pads available.

However, I find there is a trick for hand soldering them that works for me. The first step is the same as if one were going to solder legs on the clock, only I stop after putting a little bead of solder on each of the clock's mounting pads. I try to make each bead exactly the same size too.

Then I put some solder on PCB pads where the clock will go, and then wick most but not all of it off using Chemwick brand solder wick. (Also, I'm using small diameter Kester 63/37 leaded solder.)

Having prepared the clock and the PCB with some solder on both, and being careful not to burn all the flux off, then all I have to do is hold the clock in place with (Hakko) soldering tweezers and heat the PCB pads. Its may be enought to get one corner of the clock stuck down just by reheating the solder I left on the clock and the PCB pads. Or I may have to add a tiny amount of new solder to get a good joint. Its not necessary to touch the clock with the solder iron. Just heat some solder on a PCB pad from the outside corner and push it towards the clock with the solder iron tip. When the little solder blob on the PCB pad being pushed along by the solder iron touches the corner of the clock, stop there. It should be soldered.

The whole problem with soldering those tiny SMD clocks is getting solder to penetrate under the clock pad, between it and the PCB pad. Since I already left solder in both those places, the penetration problem is avoided.

Hopefully, that all makes sense.

Of course, getting old now I can't see what I'm doing any more, so I use an optical soldering microscope for the small stuff. Also tried a digital microscope, but found that hand-eye coordination is better for me with my eyes in a fixed location at the optical microscope eye pieces.

EDIT: If trying the smallest size Accusilicon AS318 clocks, I would pay close attention to what the datasheet says about the location of a bypass caps (within 1mm of the clock power pad, and with a direct path the ground pad. Otherwise, the datasheet warns the clock may malfunction. The SMD bypass caps on the Chinese USB board with Accusilicon clocks here can be seen to be in the recommended location. Also, the caps look no bigger than 603 case code size, could be smaller. It may be the AS318 clocks in bigger packages already have some internal bypass cap in the right place.
 
Last edited:
Daniel,

Nobody has reported such a problem in this thread that I can recall.

Did your scope arrive yet? If so, you might want to take a look at the fiberoptic receiver output waveform when the problem occurs to see if it has changed from the waveform during normal playback. Either that or just try replacing the receiver. There isn't much else in that part of the circuitry. However, you could check with a DVM to make sure the receiver is still getting 3.3v power as it should.
 
Last edited: